EricL

THIS IS A FIRST DRAFT AND IS VERY, VERY LONG AND HAS NOT BEEN FORMATTED OR PROOFED. THE SOURCE DOCUMENT THAT WAS SUBMITTED TO THE NHTSA IS OVER 1000 PAGES AND IS OF MEDIUM QUALITY. (The scanning was done at low-res and the color became black and white; in addition, no OCR was done, so all of what is here had to be typed by hand and took way too long).

---------------------------------



ACURA/NHTSA DOCUMENTATION SUMMARY

Preface:

I have made an attempt to “distill” the 1000+ pages of documentation provided by Acura to the NHTSA into some summary data. The full 16+ meg pdf file is available online. It took me about one hour to download the .pdf file. The files are scanned as image data are not searchable (You can search by eye, put the pages through a scanner and try out a great OCR system, or dig in and find the “goodies” buried in all of the reports.


Page 468 of 1046:
Acura Tech line Bulletin:

Erratic shifting, flare at upshifts, wont move (5/02)

The roughness of the third gear clutch steel plates is prematurely wearing the third clutch disks in production and requires a remanufactured transmission

ATF leak at axle seals, roaring noise, erratic shifting (5/02)

An ATF leak at the axle seals, a roaring noise or erratic shifting can be caused by wear [cut-off in scan], which requires transmission replacement. When the differential carrier bearing get d [cut-off in scan].

ATF Leak – If the differential carrier moves around, the axles can exceed the range [cut-off].

Erratic shifting – Contamination from the damaged differential carrier bearings can [cut-off] lock-up valves. A common DTC that accompanies damaged differential bearings is [cut-off] system).

Transmission noise – Worn bearings allow the differential carrier to move and cause [cut-off] carrier-bearing contact. This often causes noise that has been described as roaring.


Page 471 of 1046
Warranty Extension 3.2TL and 3.2 CL Automatic Transmission

(This is what the dealers have in front of their nose in case they tell you something different.)


BACKGROUND

In certain vehicles, a higher than normal number of automatic transmission may have defects in material or workmanship that could cause premature wear or failure. To insure that customers have adequate warranty coverage, American Honda is increasing the warranty on the transmission (including solenoids) and the torque converter to 7 years or 100,000 miles whichever occurs first.

VEHICLES AFFECTED

2000-02 3.2TL: All

2003 3.2TL (except Type S):
From VIN 19UUA5…3A000001 thru 19UUA5…3A019556

2003 3.2TL Type S:
From VIN 19UUA5…3A000001 thru 19UUA5…3A019061

2001-02 3.2CL: All

2003 3.2CL (all models):
From VIN 19UYA42…3A000001 thru 19UYA42…3A005203

…



If standard diagnostics confirm an internal problem with the torque converter or the automatic transmission, replace the affected part. (EricL note – exactly what is standard diagnostics? Perhaps this is the source of much grief for customers with transmissions that are clearly ready to grenade but show no error codes!)



EricL note: The following is an excerpt of a list/document, provided by Acura, that details ALL of the corrective actions required to “abate” (and/or correct) the downshift problem:


CORRECTIVE ACTION(s):

Page 5:

The NHTSA asks the manufacturer for a list of modifications/changes to “fix” an “outstanding” problem (downshift problem). Note: they don’t care about how long the tranny lasts, only if it leads to safety problems (accidents, deaths, etc).

They [Acura] needed to supply a summary of with the following “points” for each “mod” (set of mods).

A. The date on which the modification or change was incorporated into vehicle production;
B. A description of the modification or change;
C. The reason(s) for the modification or change;
D. The part number(s) of the modified part and any previous parts(s);
E. Weather the original unmodified component was withdrawn from sale, and if so when; and
F. When the modified component(s) was made available as a service component; and,
G. Whether the modified or changed components can be interchanged with earlier production components.

Also, include the above information for any modification or change that Honda is aware of which may be produced, distributed, or made available within the next 120 days.


EricL notes: Acura supplied a list of modifications to “correct” the downshift problem. I put question marks where the “meaning” – at least to me – in unclear.

Modification #1

A. Jan 25, 2002 for Countershaft; April 5, 2002 for Oil Guide Plate.
B. Tolerance on diameter of hole in oil supply area of the countershaft component was changed. Tolerance on outside diameter of oil supply area of the oil guide plate was changed.
C. To increase the supply volume of lubricating oil to third clutch on countershaft.
D. Counter shaft: 23220-P7W (Previous Part 23220-P7W-020)
Oil Guide Plate 23225-P7W-000 (Part Number correct?)
E. Not withdrawn from sales.
F. August 2002.
G. Interchangeable


Modification #2

A. May 3, 2002
B. Hydraulic pressure timing characteristics at shift were changed to reduce the thermal loading of the clutch.
C. To reduce heat generation in the third clutch during hydraulic pressure switching at shifts.
D. 37820-PJE-A62 (Previous Part: 37820-PJE-A61)
37820-PJE-L62 (Previous Part: 37820-PJE-L61)
37820-PGE-A58 (Previous Part: 37820-PGK-A57)
E. Not withdrawn from sales
F. June 2002 for 37820-PJE-A62/L62; May 2002 for 37820-PGE-A58
G. Interchangeable


Modification #3

A. Nov 25, 2002
B. Clutch plate surface roughness standard was changed for all plates. Surface roughness evaluation standard was added.
C. Reducing the surface roughness of the clutch plates reduces the clutch friction disks’ wear.
D. 22650-P7W-023 (Previous Part: 22650-P7W-013)
E. Not withdrawn from sales
F. Dec 2002
G. Interchangeable


Modification #4

A. Nov 12, 2002
B. Embossed stoppers (spell/scan problem??) were added to the filter’s ring binder. Ring binder shape was changed.
C. To improve separation strength of filtering material in ATF fluid strainer.
D. 25420-P7W-003
E. Not withdrawn from sales
F. Dec 2002
G. Interchangeable


MODIFICATIONS PLANNED IN THE NEXT 120 DAYS:


A. Planned for Feb 2003.
B. For the counter shaft component:
Change tolerance on plug press fit depth

For the counter shaft plug:
Change the inside diameter surface chamfer dimension.
Lubricating oil passage sealing width was increased.

For C-3 gear:
Change thrust bearing inside diameter
Increase lubricating passage width

For the thrust washer:
Oil passage shape and tolerance on overall width were changed.
Increase lubricating oil passage width.

C. To increase lubrication rate of AT third clutch.
D. 23220-P7W-040
23220-P7W-030
23451-P7W-010 (Previous Part: 23451-P7W-000)
90520-P7W-010 (Previous Part: 90520-P7W-000)
90521-P7W-010 (Previous Part: 90521-P7W-000)
90522-P7W-010 (Previous Part: 90522-P7W-000)
90523-P7W-010 (Previous Part: 90523-P7W-000)


E. N/A
F. Unknown
G. Interchangeable


EricL note: Hard to know which transmissions, if any, received the modifications specified in the “to do” (FEB ’03) list.



EricL note: Here is where there is a letter to the Acura service managers and they should all be aware of the problems that are occurring!!!!


Page 473 of 1046:

“Dear Service Manager” (Dated OCT 11, 2002)

A blurb to let the folks know that there are problems and what to do relative to the warranty extension. A service bulletin #02-027 is mentioned. It clearly mentions that there is NO RECALL CAMPAIN.

It basically tells them to run “tests” (as required) and replace the transmission and/or torque converter. (The torque converter is replaced with a new one; the letter says the tranny is replaced with a remanufactured one. [More on this later, this is not always the way of the world and some people “appear” to be getting “new?” units])


Page 474:

Transmission repair bulletin dated December 17, 2002

(This covers the Warranty Extension and vehicles with VINS noted in a previous comment.)

The procedure in this bulletin makes note of changing the immobilizer key transponder (new ECU). A later note is interesting in that is mentions that inability to “flash the unit” via a dedicated port. Could it be possible that the CPU could be flashed (again) with the proper IC probe/adaptor?

A new fuel pressure regulator is added for those requiring one.


Pages 476+

Quality Improvement Sheets

The list starts with an “occurrence date” of 04/26/2001. A specific vehicle with only 3142 miles is listed (with VIN!).

01M ACURA CL SHIFTIG PROBLEMS INVESTIGATION (Burnt Third Clutch)

Customer Contention: SLIPPING
Dealer Contention: ERRATIC SHIFTING

Market INFO:
* EST. TOTAL WARRANTY CLAIMS = 51
* EST. TOTAL WARRANTY COST = $102,000
* AVERAGE MILES TO FAILURE = 12,970

Analysis:

… Found in the analysis of (44) 01M Acura CL, 01 & 02 TL, and 01 MDX Market Failures, 41% of these (?) has ‘Burnt third Clutches’. 39% were ‘No Trouble Found.’ The “Sport” models of the TL and CL make up the majority of all burnt third clutch failures (81.5%). (EricL Note: Hmmm, does this surprise anyone that the high-powered sporty cars are having the bulk of the problems? However, there is more info and it just keep getting better. There is a set of graphs that show that certain build dates are better than others, at least from a statistical and Acura investigation/conclusion standpoint)

New Request = 11/29/01 Find root cause for Burnt Third Clutches for the 01M Acura CL and 01 & 02M Acura TL.


Page 478:

Minutes of monthly meeting held at HRA in August 2001 (Dated below Sep 25 2001)

<01M-CL shifting problems>
- Through 01M monitoring, AEP (ANNA Engine Plant of Honda of America Mfg, Inc) received many reports alleging that the transmission had shifting problems (167 reports, 25% of all). On investigation, third clutch was found burning in 5 vehicles. The AH ‘Chet’ vehicle (a vehicle lent to Chet for monitoring purposes) has had the same problem twice. The transmission assembly, which failed after it had replaced the transmission failed first, will be sent to HGT for analysis. (The mileage was about 8,000 miles at the first failure and about 20,000 miles at the second failure. The grammar is courtesy of Acura)

Page 479:

A graph shows defect rate vs. model for all Honda Acura models. As a note, the Acura CL had the highest Defect rate of 0.80+ % (this was for 228 vehicles).

EricL note: Data for the TL and 02 TL are included in the summary, the graphic depicts the “Acura TL” as having a defect rate of approx. 0.30% (89 vehicles) and “02 Acura TL” with defect rate of 0.10% (37 Vehicles). This is “early in the hunt for problems” and the statistics could be argued and crunched in a variety of ways by armchair transmission mongers. Acura seams to have done a statistical analysis later that concludes and shows that certain build dates of CLs are better – or worse – than others. The fact that additional time and mileage builds up with earlier models should be self-evident. Also interesting that some of the earliest builds of the CL/CLS (the very first months) have very few problems. As will be shown later, quality control was not exactly great!

Page 480:

Problems and diag codes by category for US BUILT 01’M CLs

Page 481:

Problems and diag codes by category for Japan Built 01’ M CLs

(There is a fair amount of info and this should be considered a work in progress – OK?)

IMO, the types of problems are more similar than different. However, I hope no conclusions are drawn from this early morsel.

Page 482:
01/02m Acura 3.2TL/CL third clutch burned ATM failures

All of the cars had burned third gears. One car had a burnt Torque converter, another had a forth clutch with signs of slipping, and the “No third gear” complaint car has third, fourth, and fifth clutches burned (OUCH!). The mileage ranges from 13K to 38K (approx).

Page 483:
Report from HRA Tanaka, ACE on Analysis of Symptoms Developed by Chet EVP Vehicle

[Symptoms]

The clutch slipped only in third gear. (more)

(Comment: problems identical to original transmission [no kidding – hey?])

When the transmission, which failed first, was disassembled, it was found that the third clutch was excessively worn, the piston was forced out of position and the O-ring was broken, and clutch pressure was thereby allowed to leak out. It is assumed that the same had happened to the replaced transmission. … the cause of excessive wear of the third clutch was not determined (this changes later and many changes have been made. Are they enough? Don’t know…)

[Mileage]

The second failure occurred when the mileage reached 10,707 miles.
The transmission was replaced after the first failure that occurred when the mileage reached 8.572 miles.
This means that the replaced transmission failed when the distance the vehicle had covered with that transmission totaled only about 2,000 miles

[Condition of ECU]
The ECU had no connector for flash
The ECU was, when removed from the vehicle and checked, found carrying the part number label ‘37820-PGE-A51’, showing that the ECU was a mass-produced one.

Page 484:
HRA Tech Line Data

01M CL Type-S Complaints Classification by Symptom (…)

Clutch slipping 46
Excessive shift shock 34
Won’t shift 34
Engine flares up 7
P0730 5 (SHFT CNTRL PRBLM)
Late shifting 5
Clunking noise 4
P0780 3 (MECH PROB HYDRLC SYS)
Harsh downshift 2
…


Page 485:
Trend Monitoring Data on 01M CL

A list of problems and diag codes with rates by build date is shown.

EricL Note: Interesting that Mar-01 build date vehicles have very low problems. As mentioned earlier, Acura finally notices a trend that shows that some of the trannies are not built with tolerances in spec (Please, this is not the only issue; however, it does show that the clutch problem is critically related to quality control/build issues [this is talked about in detail later])

Page 486:
01M CL Tech Line Data by Type

(Note this is a table with codes and compares defects by category and total by vehicle make. The following vehicle types are compared:

CL (no NAVI)
CL With NAVI
TYPE-S
TYPE-S With NAVI

The Survey was done on vehicles sold by the end of AUG 2001 and the CL PREM total was 14570 and CL Sport (Type-S) was 21629. (EricL Note: The Grand total for trouble codes/defects.) The amount sold by subtype (NAVI vs. non-Navi is not present).

GRAND TOTALS (Code registered by each OBDII category are summed for total.)
-----------------------
CL 65
CL NAVI 11
TYPE-S 145
TYPE-S NAVI 84

Page 487:
Tech Line Contact Situation – Trend Monitoring Data on 01M TL

The types of problems (complaints) and complaints by build date do not seem to match that of the CL/CLS.

(This page provides a complete breakdown by complaint type)


Page 487:
Contains 02M TL data

Page 489:
See the page 486 title; this is for the TL and contains totals for codes by TL vehicle type. The Survey was done on vehicles sold by the end of AUG 2001 and the 3.2TL total was 14766 and 3.2TL Type-S was 15941. (EricL Note: Notice how the “number” of cars – non-type-s vs. type-s are very close in number as opposed to the CL data.)

GRAND TOTALS (Code registered by each OBDII category are summed for total.)
-----------------------
TL 19
TL NAVI 4
TL TYPE-S 26
TL TYPE-S NAVI 14

(EricL Note: What the heck does NAVI vs. Non-Navi mean? I sure don’t know.)


Page 490:
Page Titled “Proposal on Positive Discovery

* It is required that the causes of the third clutch burning be analyzed and countermeasures be proposed.

* Availability of failed products

Today AEP was requested via HRA to call in two failed transmissions from the market.

…


Page 506:
Results of Primary Analysis of CO-S (sic?) Transmissions Called in from the Market


3 models are listed and all have problems with third gear pack. One has hydraulic pressure problem on third; the others have burned third gear packs. #2 has burned out packs 3, 4, and 5, with the other having slipping on fourth too.

Transmission of Chet EVP monitor vehicle had no other problems then the third clutch problem.
Disk is completely worn and piston ring is broken. From this condition, it is impossible to determine whether is due to burning or premature wear. For this reason, analysis should proceed as shown on the next page to determine the course of action.

(EricL Note: This is a ‘rehash’ of an earlier finding, but the next page looks like the first (?) formal analysis/review process of clutch burning.)

Page 507:
How to Proceed with the Analysis of Causes of CO-S (sic?) Clutch Burning

Eval of clutch dist as to amount of wear and burned level
* Find (1) wear due to burning (2) problem due to premature wear
Dynamic characteristics of clutch piston
* To verify weather out of spec
Check return spring load
* To verify weather out of spec
Measurement of deformation of clutch return spring guide
* To measure profile
Dimensional inspection of third clutch guide
* To verify whether out of spec
Checking of control unit
* To check hydraulic control unit for any valve stuck open/closed
Checking of third clutch lubrication system
* To (1) check dynamic lubrication rate and (2) check lubrication of passage
Return spring guide-to-washer clearance
* To verity drawing as to whether disk wear (maximum) can result in interference with washer
Recalculation of clutch capacity.
* (BLANK)
Checking of Q/A while shifting
* To analyze shift mode and measure Q/A in Chet EVP driving mode
Set load of return spring
* To have a firm grasp of change in actual load in mass production
Strength of return spring guide
* To confirm the time (?) for a switchover from Japan supply parts to local parts and compare strength of Japan supply parts and local [US] parts.
Analysis of clutch return spring guide deformation mechanism.
* To measure load at beginning of deformation and identity variance between Japan supply parts and local parts by comparative study and conduct single clutch assembly test to verify whether guide deformation can result in interference.

Page 512:

Progress of Analysis of Causes of CO-S (CL Type S) third Clutch Burning
November 2, 2001

Page 513:

Tabular Summary of Up-to-date Results of Analysis

EricL notes:

A switch was made from Japan supply parts to local parts in Apr 2000. A note about, “Local parts < Japan supply parts is noted on strength of return spring guide.

The clutch return spring guide deformation mechanism contains the following note: “If the clutch wear advances, the coned disk spring gets out of place and interference is thereby encountered between piston and hub and as a consequence, the guide is deformed.”

Page 514:

A table shows the clutch clearances and “burn level” for various trannies with assorted mileage.

“Among the returned transmissions, there are transmissions whose third clutches are burned independent of mileage.”

“The symptom ‘third clutch burned’ is not a symptom peculiar to CO-S.”

Page 515:

Comparison of Transmissions Called in from the Market (because of Shifting Problems) in terms of Mileage.

EricL note: A graph with spring clearance vs. Mileage with each gear shown as a different mark is shown. The poor scanning does NOT show the third gear data (at least I can’t find it).

“Results of analysis of returned transmissions show that the symptom is not due to deterioration in durability.” (EricL note: Someone asked about quality issues and there are some issues that have a large impact on which trannies are good ones and or bad ones; it depends on some critical tolerances. There are also some temp measurements that are done later that show that high second-to-third shifts under 80+% load cause extreme heating in the third gear clutch pack. A humble conclusion: you can get a bad one and/or stay away from constant second-to-third shifting under load [think traffic and flogging the car with lots of gear changing])

Page 516:

ERICL NOTE: IMO, things start to get very, very interesting in this page!

Changes in lubrication rate and clutch clearance.

A graph with Clutch clearance (y-axis) vs. Lubrication rate (cc/min) (x-axis) is shown for various gears.

EricL note/comment: The clutch clearance and lubrication rate show a clear relation to how the clutch wears/burns!

“Clutch clearance limit to ensure no flare-up of engine = 1.4mm”

“CO-S needs lubrication to third clutch at the rate of 1700cc/min.”

Note: production issues (can you say quality issues) allowed for SLOPPY tolerances for some of the cars. The “SLOP” varies by build date and dice tossing. It turns out the clearance is critical along with some other things. The flow rate assists in removing heat from the clutch packs. To little and the clutch packs get over 360 degrees! (I did mention spot heating in the transmission)



Page 517:
Correlation between Countershaft-to-Oil Guide Plate Clearance and third Clutch Lubrication Rate.

Third clutch lubrication rate is affected by clearance between countershaft and oil guide plate.

(There is considerable variation is supply volume of lubricant to countershaft.)

* As a tentative mass production spec, set the maximum clearance at 0.65mm (and apply it from Nov 2).



Page 518:
Proposed Improvement in Lubrication Specification

* The problem-solving specification is designed to secure the necessary lubrication rate for each clutch even if the countershaft-to-oil guide plate clearance is increased to the maximum.

* If the above specification is adopted as a countermeasure, lubrication to third clutch at the rate of 1700cc/min is secured. (EricL note: THIS IS IMPORTANT, BUT THERE IS MORE)


Page 519:

Future Activity

RE: Verification testing for confirmation of durability … The idea is to test the theory and see how the car holds up over 4,000 miles (the shortest mileage among the transmission-failed vehicles called in from the market for verification.


Page 520:

N/A


Page 522
Clutch Q/A while Shifting

Table with temp rise and other stats for WOT second-to-third, WOT third-to-second (kick-down), and closed throttle 5-to-3 MANUAL downshift.

ON WOT 2-to-3 upshift, the delta T was 100 degrees (assume C)
ON WOT 3-to-2 downshift, the delta T was 23 degrees
ON Closed throttle 5-to-3 manual downshift, the delta T was 111 degrees.

“* Clutch Q/A is not too high and is judged to be at acceptable level.”

EricL Note: If I was a betting man, and you got a transmission that was hand checked for optimum specs, I’d stay away from repeated 2-to-3 upshifts and closed throttle 5-to-3 MANUAL downshifts. I’m not going to argue to promote a view that a transmission should have “how to baby me instructions”; I’m only suggesting that constant shifting in traffic in D5 could be a real problem for cooking the third gear clutch. Please think about some scenarios on the freeway where repeated 80% downshifts could occur. Perhaps a shift to second gear in SS with a loss in gas mileage could help.


Page 523:

Analysis of Mechanism of Interference between Clutch Piston and Gear Hub

Minimum piston-to-gear hum clearance 3.53 mm
Maximum initial piston stroke 1.70 mm
Allowance for interference 1.83 mm
Disk wear allowance/piece 0.37 mm

Interference between piston and gear hub and deformation of clutch return spring are secondary failures derived from clutch disk wear.
Page 524:
Verification of Strength of Clutch Return Spring Guide

EricL Note:
A graph showing deformation vs. piston stroke for Japan and local (US) parts is shown. The Japan sourced part keeps its deformation under control with increasing piston stroke. The US sourced part causes the piston to go out of place.

The clutch return spring guide is considered to have deformed with the spring in compressed condition, and such deformation is secondary failure derived from clutch wear.



Page 540:
CL Trans Failure (Third Clutch Burned) Dec 6, 2001

EricL note: Basically, clearance between countershaft and oil guide plate is decreased.

The dimensional change has been applied to countershafts for production on and after Dec 3.



Page 541:
Analysis of AH Chet EVP Monitor Vehicle’s Driving Modes

EricL note: A graph showing throttle opening vs. speed has circles drawn on various “shift paths” to designate “modes” frequently used while driving. I found circles on the 2-1 shift at WOT and 10-20 km/h, and in 2-3 shifts at 50-60 km/h & 70-80 km/h.

[Findings of analysis]

* Frequent re-acceleration from low-speed range. 3-1/4-1 kick down shift, decelerating from 20 to 40 km/h

* Not “full-throttle, constant acceleration” but “throttle opening (TH=6/8) at the maximum” on 2-3 upshift.

Page 542:

Work will proceed focusing on the following matters:

* Third clutch Q/A analysis
* Third clutch pressure allowance for slippage

Page 543:
Proposed Improvement in Lubrication Specification

[Purpose]

* Optimize low hold (L/H) clutch lubrication rate and review the entire clutch lubrication rate to ensure the optimal distribution of lubricant

[Problem-solving specification]

* Clearance between L/H clutch hub and secondary shaft is decreased
* One-way clutch (OWC) end bearing is added

** Concrete values to be specified in drawing are, together with production engineering, under consideration. (EricL note: Quality issues???)



Page 545:

Future Activity

EricL note: Schedule of tests for mods described above (page 543)

Verification of tests…

(Short Description)


Page 556:

Progress of Analysis of Causes of CO-S (CL Type S) Third Clutch Burning

Dated January 16, 2002

Page 559:

(EricL note: Don’t laugh when you see the comment from Acura. IMO, problem not fixed – DOH!)

It is judged that the necessary lubrication rate for FDD durability for 240,000km or equivalent has been secured by countermeasure taken [at] this time.



Page 560:

(EricL comment: this page has a graph of Q/A [heat flux] (y-axis) vs. Lubrication rate (cc/min @ 5000rpm. The graph has a couple of “dots” for “NG” (no good) trannies and other plot markers for those that are OK. The graph shows that lubrication rate is rather important and depending on the Q/A (amount of heat

The comment/conclusion from Acura is so blurred that I will have to revisit this… well, perhaps not!

).




Page 561:

Results of Analysis of Clutch Q/A in Chet EVP Driving Mode

A comment on the page is interesting: “In the verified Chet EVP driving modes, Q/A (heat) is too low to directly cause premature wear. The same will be checked in other modes.

(EricL comment: IMO, depends on how many 2-3 shifts are being done per unit time…)



Page 562:

Analysis of Transmissions Called in from the Market

“When the returned transmissions were checked for dynamic lubrication, a case was found in which mileage was relatively short and lubrication rate was so high that burning SHOULD NOT HAVE BEEN CAUSED. This result necessitated our conducting further analysis of causes.

EricL notes/comments: Cars with mileage of: 3415 and 21867 had no third gear! The car with 3416 miles was definitely the “Chet” test vehicle. The third tranny had harsh downshifting (35136 km on odometer). Problems were still cropping up (no kidding for are longtime members).


Page 563:

Analysis of Difference in Incidence according to Month of Manufacture.

EricL note: There is a note about checking into later years and models

EricL note/comment: The following is important. The graph shows 01M CL shifting problems by month of manufacture. Here are the general numbers of failures by build date:

Feb-00 2
Mar-00 3
Apr-00 20
May-00 22
Jun-00 16
Jul-00 5
Aug-00 16
Sep-00 15
Oct-00 37
Nov-00 35
Dec-00 32
Jan-01 21
Feb-01 21
Mar-01 2
Apr–01 0 (?)
May-01 3
Jun-01 2

A circle “surrounds” the build dates from Apr-00 to Jun-00 AND Oct-00 to Feb-01. The quote at the top is: “Shifting problems show a tendency to occur in transmission manufactured in certain months. The date and analysis should be looked at with a grain of salt, as it was “current” as of Dec 17, 2001.



Page 564:

“Future Activity Shedule”

There is a chart and here are some of the areas they were looking at:

Verification of third clutch pressure of returned transmission
Investigation of third clutch plate surface roughness and lining quality
Investigation of influence of second/third clutch piston characteristics on clutch drag
Verification of Q/A of returned transmission (with new disk)
Verification of Q/A of Chet EVP monitor vehicle
Dispatch of replacement transmission after measuring lubrication rate and Q/A
Analysis of factors causing difference in incidence according to month of manufacturer (EricL note: The failure by build date is important!!!!!)


Page 574:

“Progress of Analysis of Causes of CL-S Third Clutch Burning” (Report dated Feb. 1, 2002)


Page 577:

“CL-S ‘Third Clutch Burned’ Occurrence Situation:

EricL note: This is when they failed for 01 & 02 CL and TL. The early months have low numbers (haven’t been around long would be one factor).

The numbers from 4/1/2000 through 4/1/2001 are less than 5.
The numbers start climbing from 5/1/2001 to 1/1/2002.

The graph looks to be colored and when scanned to B/W, the ability to separate 01 CL from 02 CL was lost (sorry).


Page 578:

EricL note: The graph shows the failure rate by month of manufacture and has two y-axis figures: # of failures for a build month and the failure rate (the earlier charts do not show failure by # of vehicles built during that month). The earlier figures from the earlier graph has a marker pen surrounding the build dates, so they must think there is something to the build dates and the first models look great. However, when looked at by “Total failure rate”, Feb 00 peaks on the graph.

Build date, # failures, total fail rate
-----------------
Jan 00 0 0.0%
Feb 00 2 1.4% (Highest total fail rate per number built, but failures are LOW!)
Mar 00 3 0.1%
Apr 00 22 0.7% Spike in total failures
May 00 25 0.8% Spike in total failures
Jun 00 15 0.5%
Jul 00 5 0.2%
Aug 00 15 0.5%
Sep 00 10 0.3%
Oct 00 40 0.4% Spike in total failures
Nov 00 41 0.4% Spike “
Dec 00 45 0.5% Spike “
Jan 01 12 0.3%
Feb 01 22 0.8%
Mar 01 18 0.5%
Apr 01 0 0.0% (Zeros What is this about???)
May 01 12 0.4%
Jun 01 12 0.4%
Jul 01 0 0.0% (Zeros ??)
Aug 01 3 0.1%
Sep 01 2 0.1%
Oct 01 2 0.1%
Nov 01 0 0.0% (Zeros ??)
Dec 01 1 0.1%




Page 579:

“Changes in Manufacture (by FCC) of Third Clutch Pack”

Third clutch pack manufacturing (assembling location of FCC was changed from Japan to the US in April 2001, and localization of the parts colored in yellow in the time table on the right was started at the same time.

Changes in parts/manufacturing process related to third clutch pack thereafter

* Localization of GUIDE COMP in the US started in October 2001
* Disk bonding process transferred from FCC-Indiana to FCC-North Carolina in December 2001


Page 580:

EricL Notes:
Has graph/table showing dates with changes to software in Base model (CL?) and Type-S. The table also has a “hardware change area, but nothing is done until Jan 2001. The note is: [IQ (01.M Acura TL) Improved on 5-2, 5-3 KD 5X orifice changed.]

The software part of the diagram has notes on changes related to shift shock and firth gear clutch burning (approx Mar 2000 date).



Page 581:

EricL Note:

A diagram with causal factors of third gear clutch failure and verification results is contained on this page. A number of items are looked into (insufficient lubrication, Set Q/A too high, clutch clearance, second clutch drag (dueling clutch packs of a sort), lining/plate defective (this becomes interesting later), HIGH Q/A INPUT (DRIVER ABUSE) [please see my note about driver abuse], Q/A too high.

EricL note: A comment about the High Q/A input (Driver Abuse) is reported as: “It was proved by verification in Chet EVP driving mode that the Q/A was not too high. (They put it in their notes, I didn’t. This doesn’t mean it is impossible to “cook the tranny” with certain usage, but they are saying, IMO, that this is NOT the root cause (old news around here).



Page 582 through 585:

EricL Note: The following pages detail BGFA Transmission Third Clutch Facing Excessive Wear. There are a number of tables with detailed measurements of clutch plate friction properties and wear.

EricL Note: P582: “The third clutch disk was excessively worn (and the core plate was exposed) in the six transmissions returned from the market”


P583: “* Facing of the frictional face on the E/P side tends to remain unworn.”

P584: “* Verification of frictional properties of the forth and fifth clutches of the transmissions returned from the market revealed that many of them had suffered deterioration in [friction property/friction coefficient]. (Verification of frictional properties should be continued obtaining more vehicles from the market).


P585: Has pictures of surface of #1 Plate (between #1 & #2 Disk in third gear clutch pack). (Too bad I can barely make out the damage due to the scanning.) Here is the Acura “comment” about what they are “seeing”: “* When the #1 clutch plate (the one next to the end plate of the No 1004257 transmission was checked, it was found that both the rounded side and the burn side were badly pitted.



P586: “Clutch Plate Surface Roughness”

A table shows BGFA “designated” trannies in the following categories (one each per row): Returned tranny, new transmission made in US, and new transmission made in Japan. The returned tranny has a burned clutch with a note: “* Surface roughness measured at an angle of 90 degrees to sanding direction.”
The specs between the US and Japan build trannies (new) are different – but is not explained in detail just yet. The table shows the new US trannies with average surface roughness in the 0.4 to 0.5x range vs. the Japan trannies with average roughness of close to 0.1. The peak roughness for US trannies was around 4.x with the Japan trannies having a value of 1.x. (There is a large difference in the surface properties between US and Japan trannies!!!!)

“* Clutch plates made in US compared poorly with clutch plate made in Japan in both average surface roughness (Ra) and maximum surface roughness (Rmax)

EricL note: The surface roughness specified in drawing is 0.5 (design error? If they spec this value in the engineering drawings, then one would expect specs that are close to design spec – hey?)

P587:

“Clutch Plate Surface Roughness (Surface Roughness Waveform)”

EricL note: When I look at the returned graph of the US plate, the surface roughness looks BAAADD. The Japan plate has an even level of “grit” when compared to the US product. The uniformity looks great for the Japan built parts, and not nearly as uniform for the US parts.


P588:

“Results of Duplication Test on Individual Clutch Plates”

<Purpose> To verify influence of surface roughness on amount of wear.

<Result>

* The clutch plate returned from the market and the US-made new clutch plate compared poorly with the Japanese-made new clutch plate in both amount of wear and surface roughness (Ra/Rmax).


EricL note: A test apparatus was built or a tranny was set up (I’m assuming this from the pictures on the page) to check third clutch wear with different plates (Japan new, US new, US returned from market). There is a comment about the test being run for 3000 cycles and the words “SAE xx tester” is used. The wear amount for the Japan plat is only 0.001 MAX (the Ra=0.093/Rmax=0.904). The wear amount for the returned plates were up to 0.11 (11 times as much). The wear for the NEW US-made plates were 0.07/0.08 (depending on side); this is about 8 times greater wear than the Japan plates. US plate friction spec: Ra 0.397/Rmax 4.281. (There is quite a difference in WEAR and surface properties!)



P590:

“Duplication Mode Analysis ~ Auto Shift ~ ”


EricL note: There is a table with shading for various gear change modes and I couldn’t find a “block” or “item” that pointed to a “nasty” area of operation. There are some blocks that say, “Mode negligible without calculation”

* Auto mode at ordinary temperature had no problem


P591:

“Duplication Mode Analysis ~ Tiptronic (Sequential) Mode ~”


* Tiptronic (sequential) mode had no problem either


P592:

“Duplication Mode Analysis ~ Analysis of Influence of Low Temperature ~”

Q/A [HEAT] exceeded the requirement on 2-3 upshift and 3-4 upshift at low engine coolant/ATF temperatures.

* FOR CO [CL/CLS], consideration will be given to implementation of the following countermeasures:

1. Change the engine speeds at which shift point occurs after cold start. (This change needs adjustments with 34D) ~ 6/Feb

2. Review the transient hydraulic characteristics at shift change when low temperature MAP (shift schedule) is selected to shorten the time required for shifting, thus keeping Q/A down (EL note: HEAT inside clutch pack) ~ 6/Feb

* Verify the influence of low temperature on high Q/A by single unit durability test. ~ 6/Feb

EricL note: The 2-3 shift exceeds a design spec for heat limit during 2-3 upshift during cold ATF/coolant operation. Do be careful with your shifting in the AM (my opinion). The TLS folks were talking about this at one time – it’s not the only issue; IMO, it’s one of many! There is a note about “countermeasures”, but I’m not sure I follow. There is a note about “drive by wire” (DBW). [Wonder why there is DBW in the Honda Accord … well, there are other reasons too.]



P593:

“Duplication Mode Analysis ~ Kickdown Shift/Upshift Repeat ~”

95-cycle repeat mode
* Test started using new US-made disks and plates

* After the “trick” test in kickdown shift/upshift repeat mode, the third clutch [was] worn but its frictional properties suffered no deterioration (see the next page for details). Verification will be continued as to the amount of wear.


EricL note: I’m pretty sure that this is not the final word and a test later on digs deeper. IMO, 2-3/3-2 shifting repeatedly in traffic is not a great thing. I keep the car in second gear using SS to keep the clutches from heating up. (You do what you want – it’s only my opinion)


P595 (IMPORTANT READING HERE!)

“Result of Analysis of the Failed Third Transmission of Chet EVP Vehicle”

When the distance the Chet EVP vehicle had covered totaled 2,560km, the engine flared on 2-3 upshift.

Disassembly of the transmission revealed that the third clutch was NG [no good]; 3 out of the 5 clutch disks were so worn that the core was exposed

Driving mode was reviewed again:

2-3 upshift with at least 6/8 throttle 1 minute after starting engine (ATF = 34C)

Repetition of 3<=>2 shifting (ATF = 87C)

When Chet EVP vehicle was used for commuting:

- 3 <=> 2 shifting was repeated 2 or 3 times, twice a day in the morning and evening. (There was no repetition of kickdown shift at high speeds)

- 2-3 upshift with at least 6/8 throttle after cold start was once/day.


Results: Repeat mode and cold (without retard) shift mode were discovered). Verification of these modes should be conducted.

EricL note: IMO, this could be something to watch for in “some” vehicles when driven cold. There was some banter about “ECU” changes for TLS to “counter” cold start wear problems, but driving “easy” until warm probably wouldn’t hurt. IMO, if the car is changing gears A LOT, put it in a single gear and avoid the clutch transitions (changing). YMMV

EricL note #2: The graphics with details data were scanned at a resolution too low to allow a detailed examination. I have some general guesses about what’s happening. I have no control over the document scanning.


P596:

“Measure to Stabilize Lubrication to Countershaft/Oil Guide Plate Area”

EricL note: getting lube to the clutch packs is important to get fluid to a localized area to remove heat. If you don’t have something moving to transfer heat away, it doesn’t matter how big an ATF cooler is mounted on a tranny. It would be possible to have the tranny sensor reading “normal”, but have ATF and clutch disks inside the third clutch pack cooking at 300+ C. (My opinions again)


* When the oil seal type to provide complete sealing was used, the third [gear] lubrication rate measured 2080 cc/min at 4000 rpm (EricL note: HIGHER IS BETTER!!!)
When the standard type was used, the third lubrication rate measured 1540 to 1740cc/min at 4000 rpm. (The graphs show that low flow – which gets lower with wear/time – causes clutch overheating failure. Note: the graphs are on earlier and later pages.)


P597:

EricL note: There is a schedule to test the following (some may have been done, some may be repeated):

Q/A in normal shift mode at ordinary temp.
Q/A in Tiptronic mode
Influence of coolant temp
Influence of low coolant temp (without [ignition] retard)
Influence of ATF temperature
Trick mode (repeated up & down shift) [Comment: This is not a great thing]

ECU Improvement
Duplication tests

Stabilization of lubrication rate
- Confirmation of effectiveness
- Confirmation of negative aspects

Manufacture-related verification

* HTM
- Changes in manufacturing process
- Changes in parts (localization, M/L [Manufacturing. Location] changes)
* ??? (of failed/healthy vehicles)
- Checking of hydraulic control unit
- Analysis of plate, disk and piston (EricL note: this refers to clutch pack parts)









P622

“Progress of Analysis.. Third Clutch Burn” Dated Feb 18, 2002


P624:

EricL note: Another wonderful graph of failures by month of occurrence.

Graph now peaks at Jan-02 with Aug-01 through Feb-02 being pretty high. EricL note: More cars on the road – with increasing mileage and time -- and people are becoming aware of “the tranny problem.”

The Acura conclusion:

* The undesirable increasing tendency of occurrence is continuing.




P625:

CL-S ‘Third Clutch Burned” Occurrence Situation by Month of Manufacture
* Both the number of failures and failure rate are particularly high among the transmissions manufactured between April 200 and June 2001.


Failure rate of transmissions of HTM make is rather high compared to that of transmissions of (Japanese) make. (EricL note: the graph has US and Japan CL/TL models in graph. The Mss (sp? [Japan]) make have a 0.09% failure rate vs. HTM rate of 0.22%.

EricL note: The lack of color in the graph prevents a “good read” of the data (low res B/W prevents conclusions – well, at least for me)..






P631:

Interesting 3000 cycle studies of clutch wear (second clutch plate), Q/A = 60

* It has been found by durability test of single full-size clutch plate that the amount of wear of facing is influenced by the surface roughness of plate. The poorer the surface roughness of plate, the larger the amount of wear of facing (particularly in case of initial wear).


P632:

Same as above, but for third clutch plate under high Q/A, low ATF temp.


EricL notes:

The initial wear is highest regardless of JP or US manufacturer. (JP is better)
The JP parts have very, very low wear during “steady-state conditions [cycles 500-3000] relative to the US parts. Depending on how one reads the graph after “break-in”, the JP parts “can” be ~4 times better for wear (4 times less wear).








P633:

“Changes in Manufacture of Third Clutch Pack”

Third clutch pack manufacturing (assembling location of FCC was changed from Japan to the US in April 2001).

Changes in parts/manufacturing process related to third clutch pack thereafter:

* Oct 2001:
Localization of PGH in the US started
Localization of GUIDE COMP in the US started
* DISK bonding process transferred from FCC-Indiana to FCC-North Carolina in December 2001
(PGH disk bonding process is scheduled to be transferred in late Feb. 2002)

No other changes in M/L, material, equipment, etc since HTM launched upon mass production in April 2000.

* The transfer of the third clutch pack manufacture (assembly) to the US factory and the beginning of the increasing tendency of occurrence synchronized with each other.

EricL note: There is yet another graph that shows clutch disk failure by build date and has graph lines for TL/CL (very hard to read graph “markers”). The 01 CL has highest clutch pack failures between Oct-2000 and Dec-2001. Note, there are other peaks, but that is the highest with the graph going through Dec 2001.



P634:

EricL notes/comments/summary: A graph of the third clutch plate surface roughness by month is all over the map by build date. The spec was Ra = 0.5 Max; here is the Actual for Mar ’00 to Dec ’01: Average: 0.31, Maximum: 0.48, Minimum 0.16.

There is a location in the graph from 5/30/00 to 6/23/00 where the graph looks “uniform” and close to Ra=0.2x. The impression is that quality control – with time -- is not very impressive. The graph has a lot of triangles connecting dots (extreme variation in adjacent build properties). It’s hard to read the failure graph, but there is a relative low in clutch failures in the 5/30/00 to 6/30/00 time (this is my own opinion from hard to read graphs).





P635-636:

EricL note: Shows tests / process change to improve US disks to Japan spec. This is process related.

- By replacement of sanding belt, surface roughness will be better than the present level.
- Unstable surface roughness immediately after replacement of sanding belt can be solved by re-lapping and as a consequence, the overall level is further raised.

* Replacement of sanding belt can bring about an improvement in surface roughness of US-made clutch plate to the extent that it is comparable to the surface roughness of Japanese-made clutch plate.


EricL note: Can any here doubt that there are a number of quality control issues that could easily account for some people getting “perfect” trannies and others getting total duds? (I’m not leaving out subtle driving style that would not be part of “common sense” – in fact would lead to some rather uncommon-sense conclusions [my opinion here].)




P638:

“Analysis of Influence of High Q/A at Low (Engine Coolant/ATF) Temperature”

EricL note: A test of US vs. Japan made clutch plates during 2-3 upshift mode (50 cycles) was performed to check wear. The surface of the clutch plate makes a difference!


* In the high Q/A mode (50-cycle durability test, the amount of facing wear of US-made clutch plate was larger [greater] and more influenced by clutch plate surface roughness than that of Japanese-made clutch plate.






P639:

Low temp shift test, etc…

EricL summary: Change to ECU for lower 2-3 shift change when low temp MAP is engaged in conjunction with change in transient hydraulic characteristics lower the temp in the clutch disks/plates to an acceptable level. None of the models have a severe problem, but better quality clutch disks + ECU change + valve body (my guess) should drop stress and temp on clutch disks/plates.



P641 (IMPORTANT!!!!):

“Analysis by Repetition of Kickdown Shift/Upshift”

WOT 2-3 UP + 3-2 KD (Wide open throttle 2-3 upshift + 3-2 kickdown shift) 200-cycle repeat mode

EricL note: Notice that this is a different test with 200-cycles (more than previous tests); the conclusions are different and you can see more work is needed to get the trannies to be reliable under “normal” use. The graphs do show that the Japan made clutch plates don’t seem to help the facing wear (there is wear on each side of the clutch plate and the wear is different depending on position [OUT or IN facing side] and position in clutch pack [nos. 1 through 5] ).


* The amount of facing wear of US-made clutch plate was larger and more influenced by clutch roughness than that of the Japanese-made clutch plate. However, even the Japanese-made clutch plate, which was good in surface roughness, met with considerable amount of facing wear when tested in the “tricky” full-throttle kickdown shift/upshift repeat mode

CONCLUSION: -- > Countermeasure is required (EricL note: DOH!!!)



P642 (IMPORTANT AND SCARY):

EricL comment: I had mentioned “spot heating” in various contexts when the transmission cooler was being discussed and it is important to note that areas can create a lot of heat due to friction and not be able to transfer the heat out even if the “general fluid” temp is “normal.” The temps inside the third gear clutch pack are getting close to 400-degrees C (Man, that is way too hot). BTW, the test is run with the “NEW” clutch disks with better quality control made in the US.

EricL suggestion: Try and minimize the WOT (or even 80% WOT) upshift and downshift operation. If you think you are going to “hunt” for a gear in “maniac mode” during heavy traffic, try to stick with one gear (second or third gear using SS) and STAY OFF THE REPEATED UPSHIFT from 2-3. This is only a humble suggestion that will waste gas in traffic, but it seems rather clear from the graphs that the temps and clutch pack loads are getting a beating during shifts. (Only my opinion and YMMV)

Here are the basics of the test:

Repeat cycle 200 times
Each cycle consists of alternating 3-2 WOT kick downs with 2-3 WOT upshifts
The first set of cycles consists of 3 up/down shifts and each 2-3 upshift gains about 130 to 170 degrees C of temp. Note, the math won’t add up as the flowing ATF cools the peak temp during each “dwell”/”coast-down period (between upshift and downshift cycles). Hey, you have to let the car slow down to repeat the cycle – hey? The 3-2 kickdown only causes a small temp spike, but the temp increase – from the 2-3 WOT upshift -- is so brutal, that the “test cycle” waits for approximately 30 seconds before repeating each set of 3-WOT kickdown/upshifts. The graph/cycle is finished at approximately 120+ seconds. (This 120+ second set of 6 [2 x 3] cycles is repeated 120 times.) The velocity (speed) varies between 80 KPH (~50MPH) and 125 KPH (~80MPH). The kickdown occurs at the lower speed (~50 MPH).

Here are the temp changes for the second set of three kickdown/upshifts (according to the graph) and each cycle lasts for approximately 13 seconds (with the exception of the previously mentioned “30 second time-out” between pairs of 3).

3-2 KICKDOWN followed by upshift (think about the auto mode doing a “normal” 2-3 upshift)
~140 – 275C (#1)
Temp drop with coast/deceleration.
3-2 KICKDOWN w/auto 2-3 upshift

~200 – 250C (#2)
Temp drop with coast/deceleration
3-2 KICKDOWN w/auto 2-3 upshift

~210 – 380C (#3)
Temp drop with longer coast till next cycle of 1,2, & 3. (Approx 30 seconds)


Acura conclusion:

* Clutch plate temperature will climb up to 380-degrees C (as calculated) in the “tricky” full-throttle kickdown shift/upshift repeat mode. In order to keep the heat built up in the plate at not more than 300-degrees C, it should be so arranged that the RTD (retard) is increased only on 2-3 upshift (this should be implemented for MO/B (???)) and 3-2 kickdown shift limit should be reviewed for effective countermeasure.






P643:

“SUMMARY”

CL-S third clutch problem arose in the market is summarized as follows:

1. High Q/A at low temperature and poor surface roughness of clutch plate combine together to accelerate wear of facing

2. Wear of facing is increased in case of repetition of kickdown shift/upshift.

3. Clutch clearance is increased, and engine flares up.

Eventually the third clutch is burned.



This problem should be solved by restraining initial wear through (1) improvement in surface roughness of clutch place and (2) review of high Q/A on 2-3 upshift at low temperature” and by “(3) keeping clutch plate temperature from rising when kickdown shift/upshift is repeated.”


EricL note: Please note the last word in the previous paragraph: “REPEATED.” IMO, there was a questionnaire done by another member on the board related to shift styles and the results seem counterintuitive regarding use of SS and “gate shifting.” I’ll have to check again, but I believe that gate shifters and people using SS had fewer problems – at least according to a “rough poll.” However, there are other quality issues that could cause a tranny failure (birth defect) that are a result of other “quality” issues that are mentioned earlier (diff carrier, torque converter, etc).








P672:

“MM Report on CL-S Third Clutch Burning” March 1, 2002


P673:

Fault tree diagram broken up as much as possible…

-Third gear clutch burn
+- Premature burn

+-- Spec toughness
--- Lube rate - Lubrication rate improved
--- Set Q/A too high - Set Q/A of TL/CL re-verified and OK’d
--- Insufficient third clutch capacity – Trannies returned from market were check and hydraulic pressure and capacity (OK’d)

+-- Manufacturing related
--- Clutch clearance too wide – Trannies manufactured in certain months in which most of the failed trailed transmissions were manufactured were checked by HTM and OK’d

--- Third clutch drag – Clutch piston and hydraulic pressure of trannies returned from market were checked and OK’d

--- Lining plate defective – (1) Surface roughness of plate of returned tranny was found POOR! (INCREASE IN INITIAL WEAR)
--- Lining plate defective – (2) Parts made in US compared poorly with parts made in Japan in surface roughness. (INCREASE IN INITIAL WEAR)


+-- Usage
--- Plate temperature too high – 3-2 kickdown shift/2-3 upshift repeated heat build up mode. (EXCESSIVE WEAR)

--- Q/A too high – Verification in high Q/A mode. Q/A on 2-3 upshift at low temp: 180 J/cm^2 (EXCESSIVE INITIAL WEAR)



+- Deterioration in durability

+-- Third gear lubrication rate too low to maintain toughness – Spec to increase lubrication to provide enough toughness to prevent deterioration in durability (APPLICATION ONLINE APR 02)

+-- Lack of toughness against engine flare-up – Compared with other models’ data at development and OK’d



P675:


This page shows before and after graphs of the “Repetition of Kickdown Shift/Upshift” test/calculation of Built-up Heat and Proposed Countermeasure”

EricL notes: The old graph seems to be identical or close to the one I mentioned that has the temps climbing to 380-degrees C in the 3-2/2-3 test. The graph to the right now has a lowered peak temp (limited to 300-degrees C). I believe that the retard has been increased during the upshift to “spare” the clutch disks (they were smoking). Note: the word “calculation” makes me a bit “nervous” as it could mean that the “model” being used to confirm corrective behavior is not being “verified” with an internal sensor. I just don’t know (YMMV).


* Clutch plate temperature climbed up to 380 degree C (as calculated) in the “tricky” kickdown shift/upshift repeat mode. As a countermeasure, hydraulic pressure characteristics on 2-3 upshift were reviewed and RTD (retard?) on 2-3 upshift was increased, which proved effective in keeping the clutch plate temperature (heat built up in the plate) at not more than the target temperature of 300 degrees C max.




P676:

* Owing to the countermeasures taken with regard to surface roughness of US-made clutch plate, ECU data, and lubrication, the amount of facing wear was reduced by about 30 microns (um)/piece and the clutch clearance was reduced by 0.27mm (decreased from 0.57mm to 0.30mm).





P678:

“MM Report on CL-S Third Clutch Burning” April 5, 2002


P679:

EricL note: more graphs and a histogram of failure by mileage. Boy, can someone say “OCR” please!

Notes:

The average failure is at 20K miles.

The other two graphs show failure by build date and by occurrence date. The scanning was done in B/W of color graphics, which make it very difficult to get ALL the data. Add in the low-res scan, and if enough people ask, I’ll try to revisit this later.

Acura conclusions:

* Occurrence of failures began to increase from around June 2001
* Failures increased sharply after transfer manufacture to HTM in April 2000
* 85% of failures occurred after vehicles has been in service for not more than 35,000 miles, indicating that the subject failure is “early failure” (EricL note: Hey, Acura said it in their very own docs to the NHTSA)


P691:

Summary (EricL notes – I put in a few clarifying notes)

1. The cause lies in the clutch plate surface roughness which took a turn for the worse
Modes discovered as the factors lying behind the concentration upon third clutch of CL/TL: (1) Repeat mode (2-3/3-2) (2) High Q/A mode (temp inside clutch pack)

2. Clutch plate surface roughness improvement
Manufacturing/process changes from Feb 7 2002 to all US models
Manufacturing/process changes from Mar 20 2002, beginning with V6-5AT.

* Clutch plate surface roughness standard: Ra 0.5 - - > Ra 0.5 + Rp 0.85 max
(New drawing issued Apr 1, 2002


3 Increase in toughness

* ECU data fixing (retard?/cold remap?) on 18 Mar 2002
* Improved ECU application. Application to TL/CL on the line from the end of Apr 2002.


[Future activity]
* HTM to have a firm grasp of surface roughness distribution in mass production to make an accurate forecast of occurrence. By the beginning of Apr 2002



P696:

“Actual Surface Roughness Values in Mass Production as Compared with New Standard Values”
Acura comment conclusion regarding disk quality (statistical analysis – bell curves, etc shown in two graphs):

Both clutch plates mass-produced in Japan and post-improvement clutch plates of FCC-Indiana make are well meeting the standard value of Ra 0.5 max.

In addition to the above, Rp 0.85 max is set up as a new standard value. HOWEVER, clutch plates of FCC-Indiana make do not meet the standard value whereas clutch plates made in Japan do. Further improvement is required in this regard so as to be comparable in level to Japanese-made clutch plates.

EricL note/comment: Well, looks like there is still work to be done and the friction material from the US is still not UP TO SPEC!!!!




P710:

“MM Report on CL-S Third Clutch Burning” Apr 11, 2002


P711:

Contents of Report:

1. Recurrence Prevention (EricL note: does this talk to people who got 3 trannies)
2. Concrete Control Technique Employed by FCC-Indiana to Ensure Rp < 0.85
3. Three-dimensional Evaluation of Surface Roughness.


P712:

1. Recurrence Prevention

(1) Repeated up & down shifts (90km/h [55MPH] 3-2 KD & WOT 2-3 UP)

- Changes were made in hydraulic characteristics on 2-3 upshift and RTD (retard) was increased by –2 degrees, which proved effective in keeping the plate temperature (heat built up in the plate) at not more than the target temperature of 300-degrees C max.

EL note: Graph shows previous CL-S with max temp of 350-degree C (Hmmm.. ); MD with max temp of 270 (MDX??); and CL-S after countermeasure of 296-degree C (The Target Max temp inside is 300C)

(2) High Q/A at low temperature (WOT 2-3 UP [upshift])
- Compensated hydraulic pressure at low ATF temperature was reviewed and change was made in hydraulic characteristics on 2-3 upshift, which proved effective in decreasing the Q/A below the target Q/A of 130 J/cm^2 max.




P715:

“Three-dimensional Evaluation of Surface Roughness”


EricL note: A method and notes describing 3-d method to surface roughness (Ra & Rp)



P724 (DAMAGE COUNT AS OF JULY 18, 2002 – REAL NUMBERS):

CL/TL Clutch Wear (July 18, 2002)

Consideration of Market Action to Take

P725:

EricL notes:
The page contains two graphs; the top is total third gear failures by month (an almost exponential rise?). Months Nov 99 thru Apr 01 are almost at the bottom of the graph’s y-axis with the average around 10 (best guess). By Nov/Dec 01, there are 150 for each; Jan 02 has 250; Feb 02 has 300; Mar/Apr 02: 350 each and May 02 is highest with 450!!!

The second graph has count of failure (defect rate in percent) by model and “service months” The 01 and 02 CL (assume non-s and Type-s) have respective totals of 607 and 820. The defect rate stays low for about 13 months and then starts uphill!

“Known Third Clutch in Data Base”
* Total third clutch failures = 2394


P726:

EricL notes:

Two more graphs with third clutch failure by mileage. The US build avg mileage to failure is 25,520 and the Japan Built Avg is 36.047 (no Japan CLs – the number is for CL and TL 01 + 02 built in Japan and US)

The third clutch failure by Build month is a bit hard to read as the scanning was done without color and the stacked bar graphs are hard to divine in black & white. Argh!

What I can comment on:




The total failures for build dates of Oct – Dec 2000 are high for TL/CL.models (Sorry, I just can’t separate out the 01, 02, CL, and/or TL).




P730 (OK – COUNT OF HIGH SPEED DOWNSHIFT EPISODES)

“What is the Problem? – Customer Contention.” Dated 6/3/02

A graph shows the total = 74

Once again, due to the lack of color scanning (out of my control), I’ve got to “aggregate some areas of data”

Here goes:

00M TL
01M CL
01M TL - - > All three 25 + 2 = 27

02M TL (14)
02M CL (8)



P731 (MORE TROUBLE):

“Other Concern”

“Loss of ATF strainer’s filtration function due to separation of filtering material”

“Strainers which did not meet spec for separation strength of filtering material flowing onto the market. (EEK!)”


LOOK AT THE FOLLOWING STATEMENT:

“Having a firm grasp of what is happening with the strainers in the field, consideration should be given to whether to include replacement of ATF strainer in the market action.”

P740:

“QAH Top Problem Investigation”
“5AT Third Clutch Burning”
Date: Oct 3, 2002

A whole agenda is given here.



P742 – 744:

EricL note:

The following pages contain graphs of the wear of the third clutch pack plates. The clutch disks and plates consist of a stack of five disks and plates. There is an “end plate” (at the top of the picture/schematic) and a “disk spring” at the bottom of the “stack” of disk plates/disks.

Acura concludes that the disks near the clutch piston (bottom and near the disk spring) have higher wear/burning levels than the endplate. There is a “Due to?” note in there!

Graphs for TL/CL, MDX, and Odyssey are included (the MDX is blank).


P748:

Another chart with # of “Sudden Downshift Breakdown” by type (TL/CL) that was in color and was scanned into B/W. The total at this time is 89 (according to internal Acura documents. (DATE 9/11/02)



P750:

“Firth Clutch Engagement (NORMAL PRESSURE)”


A diagram of the hydraulics shown the “modulator pressure” to overcome spring pressure which permits gear engagement and hold above 45 MPH.

P751:

“Second Clutch Engagement (ABNORMAL PRESSURE @ SUDDEN DOWNSHIFT)

EricL note: Another diagram showing the modulator and hydraulic schematic shows that low pressure can cause second gear to engage.

“LOW MOD. PRESSURE IS UNABLE TO OVERCOME SPRING PRESSURE THEREFORE MISSION ENGAGES SECOND GEAR. THROTTLE OFF WILL DROP MOD. PRESSURE BELOW SPRING PRESSURE DUE TO CLOGGED FILTER SCREEN.”


P754:

“MQ Analysis Report Summary Info (Sudden, Self-Downshift Problem)”

Failure Mechanism:

1 Clutch disk wear - - >
2. Clogged ATF Strainer (broken) -- >
3. Much contamination throughout ATM (Clogged VB Screens) - - >
4. Low Line Pressure (Esp. @ low RPM’s) - - >
5. Low Modulator Pressure (Esp @ low RPM’s) - - >

RESULT: Abnormal Shifting (5 --> 2 or 4 --> 2); Sudden, Self-Downshift

Deceleration rates expressed at delta G over delta T (deceleration over time) (Note: they didn’t put in a conclusive comment here on exactly what deceleration rate is occurring )

0.5 G/Sec 0.1 G / 0.2 sec
2.46G/Sec 0.6 G / 0.2 sec (math??)
4.25G/Sec 0.85G / 0.2 sec (Comment: “Tires will chirp”)



P756:

“5AT Third Clutch Failures – Current Situation and Seven Yr. Extended Warranty Impact Forecast” (Date 9/23/02)


EricL note: There are two graphs and here is the conclusion:

“The largest majority of the claims cause for failure is third clutch (except for 00-01M TL’s and CL’s had some failures due to the Taper Bearing Problem [Diff carrier bearing problem])”

“In general, the newest models are failing at a much higher rate than the older models”

P758:

“Total Failures vs. Third Clutch Failure Comparison”

Total Failures = 6604 (AS of 7/29/02)

“The third clutch failure trend shadows the overall failure trend.”


Total third clutch failures = 3039 (7/29/02)

P759:


“Total Failure Build Month”

Note: the graph was color, but copied in B&W. The stacked bar is impossible to use to isolate CL from TL and 01 from 02 and Japan from US build (in case of TL).

Note: There is a note in the graph: “ TC Sand Contamination” for APR-000 to JUN-00. I suspect the peak at May-00 is what they are showing (not sure).

Note: Oct-01 through Jul-02 is circled with note: “Confirmed Oil Pump Seizures 03M TL & CL”


P763:

“Build Month Total Failures vs. Third Clutch Comparison” and “Build Month Sudden Downshift”

The graph has peaks at Jun 00 = 9 and Dec 00 has a peak (including adjacent months) of approx 12-14. The incidents drop by Dec 01 and go to zero by Mar-02 build date.

P764:

“Model Breakdown Sudden Downshift”

Note: I thought there weren’t any 01M CLs made in Japan (Hmmm… typo?)

00M TL (US) 6%
00M TL (Japan) 4%
01M CL (Japan) 4%
01M CL (US) 33%
01M TL (??) 31%
02M TL (??) 22%



P776:

“02M MDX Third Clutch Failures”

Note: They had problems too! The claim rate is low, but they have failures in the 0.010 to 0.060 % range [depending on build month].

P787:

Notes about change to ECU to the hydraulic pressure characteristics. They changed the ECU for different shift characteristics during (1) low temp startup and (2) during repeated shifting in 3-2 kickdown and 2-3 upshift modes. The retard is increased by two degrees during the shift. The resulting plate temp is kept below 300 deg C or less [the old temps were > 360 degree C!].

“By changing ECU data setting for shift, the repeated shift mode and low water temp mode have become OK”

P788:

ECU change info (COLD):

Previous shift when cold was 6200 RPM on 2 to 3 shift. Modified (new) version is now shifting at 5700 RPM. (THIS IS FOR COLD OPERATION [ATF COLD])

G-force went from 0.100 (pre-mod) to 0.112 (post-mod)

ATF temps around 30 deg. C

P789:

ECU change info (WARM)


Previous shift, when warm, was 6426 RPM [measured] during 2 to 3 upshift. Modified ECU now has shift at 6369 RPM [measured] during same 2 to 3 upshift.

G-force went from 0.110 to 0.130.

ATF was 83 – 85 degrees C.


EricL note: I don’t think the figures point to a loss of power with the increased G-Force from the ECU replacement/mods! (YMMV)


P791 – P798:

Strainer failure analysis:

Acura made changes to the filter/strainer to keep it from “bursting” and causing a “sudden downshift.”


P812:

Lube pressure Data.

Conclusion: High claim months have low lube pressure


P880 – 881:

Mentions failure in valve body screens and strainers due to contamination and deterioration

P887:

“Mission #8027653 EMCS DATA ANALYSIS CONCLUSION.”

RESULT:

* When ATF strainer and each filter is blocked due to the wear of third clutch causing contamination through out mission. THE PL(MOD) pressure decreases due to the state of the hydraulic circuit on FIFTH. Shift valves A,B,C enters the state of turning off this becomes second hydraulic circuit. Result arrives at the incident of 5-2 kick down (EricL note: the bad grammar probably the result of a “translation” from Japanese to English and seems to be common in the documents.





P900:

A look at the shift solenoid screens…

*Disassembly revealed contamination – NG –
*Modulator pressure to solenoids restricted




P901+

There are “Disassembly Analysis Results” of various transmissions with VIN numbers and build dates.

In general, the gears were in good shape, but the third clutch-burning problem was considered a “root cause of failure.” The clutch failure (third gear clutch pack) caused friction material to clog the ATF strainer causing it to fail, and allowed contamination to enter the hydraulic control circuits (OUCH!)

Some of the models had other types of failures (or discoloration and or heat damage) and forth and fifth clutch packs were also found to be worn. (Some other data indicated that once third went, the other clutch packs would start to “fry”). A note about one of the trannies that had been brought in before slipping was too bad, had 1,2,4,5 clutch packs in A1/A2 shape (very good condition).


P958:

G Force during “Sudden Down Shift” was -.85G and occurred at 56 MPH (this is a confirmation/duplication of a 01M TL failure with 32,595 miles).

P961:

A simulated G-Force of –0.96 G was calc’d for a TL @ 60 MPH



P970:

G-force on 5-2 down shift at 64 mph was 1.029 G (another episode/vehicle).


P971:

There is a graph with comparison of peak deceleration vs. “Sudden BRAKE at 5 MPH.”

The brake application generated a force of –0.9G, with the peak value of approx –1.1G occurring at 68 MPH (auto tranny TL/MDX).

EricL note: I guess the implication is: “If the G-forces don’t pull the eyes out of your sockets, your ok to hit the road.” (Sorry, I just couldn’t resist…)


P982:

It seems that the “downshift problem” occurs mostly when the car is at a low engine speed (low pressure from ATF pump == low modulator pressure).


P988:

“5-2 down-shift occurrence ratio = 84/681,103=0.01%”

Note: total warranty claims = 9,234; total third clutch failures = 5375

LiQiCE

Hey thanks for taking the time out to type all of that EricL! Its very interesting.

Do you think that cars manufactured after February 2003 with all of the mods and planned mods (and trannys replaced after February of 2003) will still experience this problem? It sounds like even after Mod #1 and #2 are applied that the problem seems to be reduced, although not completely gone.

EricL

Guess/think are the keywords here...


I 'think' the newer ECU/clutch pack/"other mod" set of "upgrades"/"manufacturing process improvement" should keep the clutch packs cooler. The updated specs and process for the clutch disks "should" make them last longer.

IMO, if you are talking about the "spontaneous downshift issue (4-2 / 5-2)", you are probably going to be OK with the new mods.

IMO, if you are talking about having a tranny that will basically last for 200,000 miles, then I have my doubts. I would NOT classify the current transmission as a "build like a rock" design that will last for years and years.

I believe that someone could "trash" the third gear clutch pack by doing constant 2-3/3-2 WOT up and downshifts (over and over again until the temp built up inside the third gear clutch pack). The test that is/was performed by Acura has a "break" between each set of 2-3/3-2 up and downshifts under WOT. The graph showed each temp spike from the WOT 2-3 shift getting higher and higher. They let the unit "cool" (a small time-out is allowed) between sets of 3 cycles. IOW, if you take the car to a track and it happens to have a set of curves that calls on the car to make NONSTOP 2-3/3-2 transitions under WOT, it will probably fail after one or two sessions.

Bottom line: If you have a "newer" box that has ALL of the new mods and the car is given a break (some time to cool down) between upshifts (the real heat maker), it should hang in much longer. I just don't see it lasting for 200,000 miles if someone really "bangs on it" all the time. As for the dying at 2000 miles or 20,000 mile during "normal" day-to-day driving, it should be much more reliable. Not having an electronic throttle limits how much power can be cut during each gear change.


I'd love to give a very simple yes/no answer, but I have a gut feeling that the design changes were limited to keep costs contained and allow the tranny to last during the 7yr/100k extended warranty period.

Non-limited changes could have been:

1. Completely revised tranny case halves (need room for more clutch disks).
2. Much higher ATF flow for all clutch packs.
3. More clutch plates for all the clutch packs.
4. A system/design mod to allow high ATF flow rate even at low RPM (while in 5th gear "loaf" mode). (The downside is more power being used to drive the ATF pump and this would mean lower gas mileage; IOW, this would not be something Honda/Acura would be doing with the current design. An electric ATF pump that used stored energy – similar to a hybrid gas/electric vehicle -- could make for more HP at the wheels with better cooling)
5. External ATF filter that was easy to change.
6. Electronic throttle to completely remove the power during gear shifts (See new Honda Accord).
7. Additional external cooler to INSURE that ATF fluid temp remained constant, but would allow for "additional" cooling during high demand usage.

(Acura would probably say, "You go build a race car auto tranny ... and you can fund it too...")

BMWETR

We should all be very grateful for the time and effort you expended in reading and summarizing the ACURA/NHTSA DOCUMENTATION SUMMARY. It answered many questions for me and not surprisingly raised a few new ones. Since you seem to have analyzed the data extensively, perhaps you could comment on them for me.

My car is a 2001 CL-S, with a build date of February 2000. Currently it has 37,000 miles on it and the transmission has not as yet required service. The car has never been to the track and probably has endured less than a dozen WOT 2-3 upshifts and half that number of 3-2 KD followed by a 2-3 WOT upshift. As I’m sure most people on this Forum realize, the 2-3 WOT upshift occurs at about 70 mph. Between traffic and Troopers, there are rare opportunities to be at WOT at 70 mph. I use Sport Shift about 20% of the time.

1. The chart for page 563 indicates a relatively low incidence of problems in cars with a Feb. 00 build date. That, and the fact that is was a “Japan” transmission made me somewhat optimistic.

2. The chart at page 578 indicates that the Feb. 00 build date cars had: “Feb 00 2 1.4% (Highest total fail rate per number built, but failures are LOW!)”. Could you explain how you can have the highest total failure rate and still have LOW failures? Also, I thought the "Japan" transmissions had fewer failures. This made me pessimistic and confused.

3. IMHO, the worst possible scenario is the unexpected 4-2 or 5-2 downshift while cruising. It sounds like this is caused by a clogged/burst filter from debris emanating from the third gear clutch pack. Would it not be in Acura’s and our best interest if, as part of their extended warranty campaign, they provided periodic replacement of the filter with the new upgraded part? The necessitated fluid change couldn’t hurt either. Although this would not come cheaply to Acura, it would be far less costly than replacing or rebuilding all of the potentially affected transmissions and the potential lawsuits and press which will inevitably result when someone is seriously injured or worse.
4. The article answers why those who use SS frequently or exclusively seem to get at least equal if not longer longevity from their transmissions, even though as a group those people probably drive more aggressively.
5. Your point about using SS in heavy traffic makes sense. In heavy stop and go traffic I actually lock the transmission in second using “2”, since SS still permits the transmission to wander between first and second.

I really love my car. It's a shame that the joy of ownership is marred by possibility that the transmission will some day drop into second gear at seventy MPH. I don't even want to think of the result.

Any input you could provide would be appreciated.

BMWETR

P. S. I called Customer Service last week. I had yet to receive my Extended Warrenty notification letter. They told me:

1. My car is definitely covered.

2. They did not have access to the letter.

3. They would send me some kind of confirmation.

True to their word, I received service bulletin #02-027 in the mail. Any interest?

EricL

Yes...

darrinb

thanks eric,

i have a 03 auto but i am still covered under the 100k waranty.

so... i have prolly gone WOT on the 2-3 shift like i'd estimate 200-250 times in the 9 months i've had the car, 150 of those w/ the intake and headers on. so is it safe to say that my tranny will not go casue i have already beat the fuck out of it and it hasnt gone yet??

EricL

It's the repeated WOT 2-3/3-2 shifts that fry the clutch packs. (Emphasis on repeated).

IMO, the more you beat on it, the faster it will go.

Short reply: Sorry, I just don't know...

darrinb

my tranny is bullet proof, i've made like 20 1/4 runs in less than a 2 hour span

i know poeple that have never beat on there car and there tranny went

EricL

1. 20 1/4 runs in less than 2 hours is not the test that was used to "fry" the new trannies. The test that Acura used to "destroy" the third gear clutch pack required "constant" 2-3/3-2 WOT upshifts and downshifts (there were small breaks). If you run down a quarter mile, you are going to be cooling down the tranny just heading back to the starting line (even if you had the track to yourself).

2. Some of the problem boxes were so messed up, that they were predestined for failure (irrespective of treatment).

Look, if you get a tranny, and it's one of the poor Q/A ones, it will fail.

I general consider "bulletproof" to mean: indestructible (just my opinion and one of many interpretations).

EricL

The graph displays two values per build date. If you got a car built for the US market, you have a US built tranny. A lot of document has duplicates in Japanese (Kanji, etc), and the translation is not always "wonderful."

If you built relatively few cars in a given month (compared to another month), you could have low incidences (2-3 are shown for that Feb-00 period), but the "Total failure rate" could be relatively high at 1.4%.

This is what is very sad about statistics and human nature: there is some comfort in finding out that you "got the good month", however, I don't think it is helpful or comforting to the person that gets the lemon that month. IMO, as far as they are concerned, that stats are not of great import (being burned is being burned). IMO, the stats only point to trends.

Would you prefer to have the stats for Dec 00? Dec 00 had approx. 40 failures, with total failure rate sitting above 0.5%.

Also, consider that the Feb 00 CL/CLS was on the road LONGER than a Dec 00.


You'd have to ask Acura about that.

IMO, once the clutch packs start going, you will know it and you have to MAKE a stupid dealer replace the box. Some dealers will happily replace the box once it starts acting funny. The NHTSA rep. commented on how he “believed” that the “spontaneous downshift” was preceded by slipping, flare, and other issues.

IMO, if you start seeing dark/burnt ATF, be concerned.

IMO, and yes, it would be nice to have a filter that was easy to inspect/replace.

There is some debate as to how easy it is to get to the filter/strainer (if at all). (Hey, I don't know for sure...)




YEP...

I would look for any signs of flare, slipping, or other symptoms that "normally" occur before the filters burst.

Some Acura dealers and regional representatives stonewalled members/owners who complained about flare, slipping, noise, etc. Some had to fight to get them to replace the tranny.

Watch for signs of very dark/burnt ATF.

As for the joy of ownership, I buy Sony TV's, but had to replace my primary tuner within a year and another friend found all of the power supply transistors unsoldered (they were hanging in by the crimps alone -- no solder used). Both have been running along with no problems (well, after the one-time repairs).

darrinb

thanks again for all the info Eric!!!

I hope my tranny dont go

hurricanegms

My car is a TL-S w/ Navi 2002. Build was sometime around June 2001. My transmission failed badly at 30K mi. It was "replaced" at Acura of South FL after much hassle. They denied knowing anything about a problem. I dealt with Acura Central here and they were not much better. I am very disappointed in Acura. My trani failed again this week but Acura said it was a computer problem and after it was reset it appears to be OK. However, I am driving around just waiting for it to fail again. Needless to say, I am looking into a nice new Toyota now. Oh, Acura Roadside Assistance took nearly 4 hours to get to my house and I am NOT in a remote location and it was during the day!

ffearless

My 2nd tranny just went out at 88k. Acura is replacing it without a problem however I am concerned about whether it will last. I put about 50k miles a year on. Should I just sell the car and get something else? I have enjoyed driving it very much but really need reliablilty... and fun. What else should I consider?

showgunz

EricL, can anyone access this document? I'd like to read the Honda Accord version of it to see their tranny failure roots. Thanks!

EricL

Yes, anyone can access the docs.

I had to get the document at night and it took about 5 hours to complete the download. I have high-speed access, but the NHTSA site is very, very slow -- at least it was for me.

I don't think you will find any docs for the Honda Accord. The downshift issue can be found by looking for the 2001 CL/TL/CLS/TLS.

italiaCLs

so your saying that some of the auto trannies are better then others?

stonebridge

Here is a link to Gears Magazine article about Honda 5 spd transmission issues. It has several pictures and some explanations.
http://www.gearsmagazine.com/2003-01/2003_1_44.pdf

Here is a link to an article about Honda's 4 spd.
http://www.gearsmagazine.com/2002-08/2002_08_46.pdf

beestea

Wow thanks for taking all the time and doing all the work to get that summarized. It is a very interesting read.

Edward'TLS

Can someone explain what the heck is "closed throttle 5-to-3 manual downshift" ? I thought the auto box will downshift from 5-4-3 by itself when the SS is tapped back twice manually.

allenn

My CL-S trans went by by on 8/15/2003 while I was commuting home on Georgia 400 at 5:00 P.M. I have 46000 miles on the odometer, no hot roding, no up and down shifting, used the manumatic a few times when it was new. I drove it slipping and bucking to the dealer who checked the CL's computer which reported a lot of transmission error codes.

I am concerned that the replacement will not last. What warranty is Honda giving on the replacement trans? It seems to me that Honda had problems in the 80's with their automatics for similar reasons such as over heating fluid. What is the differences between the CL / TL tranny and the one in the Honda Accord V6? My wife's '99 Accord V6 has not had any problems.

DeezNutz

probably dropping it from D5 to D3 gate-shifting.

SiGGy

http://www.acura-cl.com/forums/showt...ghlight=tranny


(sorry couldn't resist)

Sucks man, I'm on #3

allenn

Great to hear from someone from my old home area, K.C. Too bad about the tranny, mine went at 40K +. I figure it will go kaput again before I reach 100K. The sad thing is that no one will buy a CL when it is time to sell.

Jay CL Type-S

Excellent research and efforts to investigate the matter. Many thanks!!

My 01 CLS transmission failed last week and is now being replaced. 40k miles. At 20k I thought there was something wrong when it seemed like it needed more gas to keep constant speed than it normally did. They check the computer, no codes, test drive, they said it was normal as the computer was adjutsing. It then cleared up after 1-2k miles.

But then about 3 weeks ago, I noticed occaisonal hard 3rd gear shifts (thought that it could have been a problem, but I let it ride). Then last week, in the morning erractic shifting, slipping gears; made the appointment at the dealership, they were surpised no lights from the indicators; that day on the way to the dealership idiot light central started 5 minutes from the dealership.

Dealership told me the transmission and the computer module would be replaced. Will these be re-manufactired or new parts and what now is the predicted reliability? I normally drive in D5 auto, but frequently shift down with the SS at red lights; I never use it for upshifting---do you think my SS downshifting made the problem worse?? Almost was deciding to sell, but I still plan to keep it. Your thoughts?

allenn

My guess is the transmission has a major engineering flaw which Honda probably has corrected for the new TL and newer models.
All that Honda wants is to get you over 100,000 miles then the problem is yours.

I doubt your downshifting has anything to do with the tranny's early demise. I drive in auto five too and never use the SS up or down. The funny thing about this whole episode is the fact the tranmission is a Japanesse source part.

sgriffin

Nope, it's not made in Japan.
Even the "fix" was originated in the U.S. by their outside vendor or transmissions.

allenn

It was my understanding that the 5 speed auto transmission was sourced from Japan. In fact, that was the only major part that was
from Japan.

RichardA7

After replacing my transmission and after my injuries, I received a notice that I should replace my PCM. I want to be sure to keep the replaced part so it can be reviewed by experts on the consumer's side. After my tranny was replaced an attorney friend of mine tried to get the damaged part and was told I gave them permission to throw it away. I never recalled doing that, but perhaps it was in the fine print of something I signed. This time I want to be sure to keep the PCM Anyone else being told to replace the PCM? Contact me at richarda7@yahoo.com

sgriffin

Glad to have provided the correction for you. Use the search function on this sub-topic for lengthly details.

EricL

1. The 2-3 upshift and 3-2 downshifts do accelerate the wear of the clutch packs. What the Gear magazine article neglected to mention, was the failure mechanism that was documented and known (by Honda/Acura) inside the transmission due to the clogged filter/strainer (filled with clutch pack material). Excerpt follows comments #2 and #3 (this is from the summary/condensation):

2. There were (are still?) well-documented quality control issues with the trannies!

3. I'm sure Honda/Acura "focused" on doing fixes that did NOT require a major redesign and focused ONLY on dealing with "obvious" problems documented in their internal documents that were provided to the NHTSA. (I'm quite sure that you wouldn't not see them thinking about making a new case for a completely redesigned tranny with more clutch disks, electronic throttle, modified ATF cooling, better filtering, etc, etc...




'“LOW MOD. PRESSURE IS UNABLE TO OVERCOME SPRING PRESSURE THEREFORE MISSION ENGAGES SECOND GEAR. THROTTLE OFF WILL DROP MOD. PRESSURE BELOW SPRING PRESSURE DUE TO CLOGGED FILTER SCREEN.”


P754:

“MQ Analysis Report Summary Info (Sudden, Self-Downshift Problem)”

Failure Mechanism:

1 Clutch disk wear - - >
2. Clogged ATF Strainer (broken) -- >
3. Much contamination throughout ATM (Clogged VB Screens) - - >
4. Low Line Pressure (Esp. @ low RPM’s) - - >
5. Low Modulator Pressure (Esp @ low RPM’s) - - >

RESULT: Abnormal Shifting (5 --> 2 or 4 --> 2); Sudden, Self-Downshift...'

allenn

Thanks for the additional information. In my case, downshifting is not an issue; because its rarely if ever done.

pdxmac

Has anybody thought about trying a heavy duty aftermarket tranny cooler? ala towing package type add-on. Seems like that would help a bunch

EricL

Yes, some people have put on B&M coolers that are added to the little hockey-puck sized intercooler the factory installed. Comptech sells a kit that is basically the B&M with brackets, hoses, etc, w/manual.

It will help to some extent. However, there is localized (spot) heating in some areas of the transmission -- this means that only so much ATF can get INTO the clutch packs. Depending on the particular transmission unit (a quality control issue), it is possible that a given unit can have reduced flow – into a clutch pack -- and that can make it a "choke point" for the ATF. Since the ATF is used for COOLING and LUBRICATION, an external cooler can only help to a limited extent when the supply of cooling fluid (ATF) is limited in volume to the 2nd and 3rd gear clutch packs and other locations.

It will help pull down the “overall” temp to keep it from degrading over time – especially when used for spirited driving, stop-‘n-go driving, uphill driving, etc, etc.

YMMV

pdxmac

Thanks, My tranny is toast. Car's in the shop having the tranny replaced as we speak with a "new unit" from the factory. 24350 miles! woo hoo

sgriffin

I'm getting 25K per transmission.
#2 died today at 49,700 miles

pdxmac

Well good thing we have the extended warranty. I'll just have them change the transmission every other oil change. " 10w-30 synthetic please, and oh yeah, change the transmission too, K?"

jkatt

hi,

i just bought a 2003 Acura CL - automatic.
has about 55,000 miles on it, should i be concerned on
this transmission problem? it does have a loud roar
on take off if you punch the gas hard, havnt had any
other issues with it.

thanks,
JC

wytman

A friend who has a 1998 Honda Accord V6 has his transmission replaced last fall. My CL-S '01 trans just died last week - 46,000 miles. I unquestionably have the same concerns - most notably because my problem was not between 2-3 as the long but entertaining post from EricL showed the problems to be, but between 3-4. I was driving to work last Thursday morning, when the car shifted out of third, and couldn't get into 4th for about 1/8 mile. It was litterally in neutral. It got worse and worse - when the dealer insisted on a test drive, we almost ended up having to push my baby back to the shop. The fact that my problem occurred at a different clutch assembly (apparently) has me very concerned that they really either don't know what the real problem is, or that they don't want to pay to fix it.

Eric

EricL

You're talking about a third gear clutch pack slipping (flare).

That's exactly the clutch pack that is the one that is most commonly trashed; a problem can easily manifest itself as slipping from 3rd to 4th. It's actually the THIRD gear clutch pack that gets the most severe trashing in all of the tests. And, it doesn't EXCLUDE problems in other clutch packs. (See earlier part of first posting.)

ACURA's test that goes from 2-3 and 3-2 is primarily designed to "trash" the 3rd gear clutch packs; that's the one they did the temp studies on.

One thing that may not be clear is that the test they did to "abuse" the tranny -- going from 2-3 and 3-2 – was to look for overheating. I probably should put some bolding and extra comments that point out that Acura is measuring the temperature INSIDE the third gear clutch pack.


Just one of many excerpts from original posting:



Acura owner's link excerpt: