Diameter

This question has been bugging me for a while. I did a search, and found that the instrument panel used QAS caps, but I could not find if the ABS and TRAC computers also used the QAS problem caps.

Does anyone know if they used QAS caps in 89631-30080 (TRAC) and/or 89541-50060 (ABS + VSC) modules?

I apologize if this has previously been asked and answered.

89OLDMAX

I have only dealt with QAS caps in Toyota and lexus ecu's and instrument clusters in the older models. I have no knowledge for the 89631-30080 or 89541-50060

Diameter

Behold the wall of text (sorry)

I need help with my car. Here is my horrible story (1994 LS400 w/130k miles):

Had a strange flooding / no start leaving work in January. It flooded, spun really fast, and refused to start or even fire. At the time I resisted the notion of flooding so I went down a lot of troubleshooting dead ends. I tore it apart and checked a lot of things and tinkered with many things in an attempt to find "the problem", but I never found anything obviously wrong, so I re-assembled it, and it started OK after a lot of cranking and some obvious excess fuel. Seemed to run OK, but I was still curious, so I checked my ECU, and found some of that QAS fluid under C105 and nowhere else. AHA! So I read a lot of this forum topic, ordered a bunch of low-ESR caps, and did the capacitor replacement. I also repaired the trace that was etched through right by C105 with a bodge wire. After the repair, I was really stoked when it started right up and ran great. I was actually amazed that it worked because of the complexity of the device and the fact that the board itself had sustained damage from the C105 leakage. But this amazement only lasted a couple months, when it started running odd again. This time it was not a flooding issue. Now it just wanted to die while idling, or driving (low-mid speed), but I could generally drive it so you may imagine it was easy to assume something else had gone wrong. I still cannot prove there was no other problem than the ECU, but whenever it would act up the diagnostic port was unresponsive, so diagnosing it was made difficult. I did not come to believe until very recently that an unresponsive diag port is sort of a smoking gun. FWIW the only code I ever got (only one time) was for the igniter. I swapped the igniters out and the code disappeared, but the sudden intermittent dying problem did not.

So I procured a totally stock '93 ECU unit (89661-50141) and kept that in the car for when it might act up and I could do a swap to see if that was my problem. It was a couple months of driving before I got a chance to swap them out. This time, again, I was leaving work and it started, then died instantly, then flooded again like the very first time back in January. Except this time I held the throttle WFO so it would start, just to have it die again. Immediately I swapped the '94 ECU out for the '93 and it started and idled perfectly right away. Just to be sure, I swapped them back and of course it wanted to die with the '94 unit. So I swapped the '93 unit back in and left it in there. It ran fine for a while, so I declared that my '94 unit was either still faulty, or I had botched the capacitor replacement. Either way, I had a running car and I was generally happy. This lasted maybe two more months.

Then the problems started again. But they were different and mild. This time it was running rough, getting terrible mileage. Full throttle acceleration was fine, but that legendary LS400 smoothness was still missing from the midrange and at idle. The diagnostic port on the '93 ECU unit was still responsive, but not throwing any codes whatsoever, even in the hi-speed diagnostic mode. So I read a lot more forum posts and tried my best to troubleshoot the phantom problem with no diagnostic markers. I checked the TPS (again), checked the coils, plugs, plug wires. I replaced both of my upper temperature sensors. Still ran like crap, no codes, no smell of fuel in the oil. I still had no clear indication what the problem could be, but the whole time I was suspicious of my injectors since the original flooding way back in January.

So out of sheer frustration I ordered a set of 8 "rebuilt" injectors off of ebay. (I have come to regret this decision.) Within a week it was running on what felt like 5 or 6 cylinders, intermittently. The strange thing about it was it would run OK if I started it, drove it to where I was going, and parked it. If I started it after sitting about 30 minutes, it would start up like it had one or more dead cylinders. If I let it sit for several hours it would start right up and run fine as long as I wanted, until I parked it again. That behavior lasted for maybe two or three weeks, and it was getting me to work and back so I kept driving it. Suddenly one day it just spontaneously started running on less than 8 cylinders. After that day It could not be coaxed into running OK no matter how long it sat between starts. So I drove it home that way, and knowing that I have at least got a constant problem to diagnose, I pulled the plug wires one by one and when it got to #6 there was no difference. So that was the smoking gun, or so I thought. I swapped out injector #6 with one of my originals, and it ran OK for another week or so. Then the problem came back. That is when I set aside my weekend to go through some injector wiring diagnostic routines. This included the Injection/ECU diagnostic routines in the Factory Service Manual (FSM), and putting noid lights on the injectors and cranking it. The noid light fired just fine on every injector. But while checking voltages and ground integrity at the ECU I noticed a very small difference in the voltage supply to injector #7 (something like 0.2 or 0.3 volts less than the others). I tested injector #7 resistance and it was completely open. This was the other smoking gun, and the second ebay injector to fail.

I did some cursory research on reman injectors, and ordered another full set of 8 injectors from a reputable injector remanufacturer. Knowing the work involved with injector installation, and worrying about what I had missed along the way, I cleaned the throttle body and got that ready, then I tested the compression while I was installing the reman injectors. Incidentally my compression was excellent on everything except #7, which was in fact a normal value, but slightly lower than the rest. A few drops of oil brought it up to within 6% of the others on the second try. At the time I attributed the difference to that cylinder having no fuel for a time prior to the test. I could be wrong on that. I also put new spark plugs in, and tested every single plug wire resistance while I had it apart.

Seemed to run OK with the clean throttle body and reman injectors and new plugs, but it had a strange, subtle starting problem. It always started, but about 1 in 4 starts it misses a little and then belches to life rather roughly. The rest of the time it starts about normally. Didn't seem to matter what the temp was. And it still ran a little bit rough in the off-idle and mid-range. But it was running on all 8 cylinders, it wasn't dying intermittently, it didn't emit rich exhaust, and it got decent fuel mileage.

Not completely satisfied with my results, I open the '93 ECU and I notice some of that QAS fluid has leaked out, this time under C108, and this time a lot less than my original '94 unit. No visible board etching or damage, just a little puddle right under C108. So I order a "repaired and tested and guaranteed" ECU off of what seems a reputable seller on ebay. As soon as it gets here, I put it in, and it runs just OK. It shifts pretty smoothly, it makes decent, smooth power, and I get about two good tanks of gas before it starts to degrade. So I pop the ebay ECU open and what do I see? All the capacitors have been replaced with regular grade caps. They did a decent job of it, but they used inferior capacitors. And there is a huge patch of damage down by C832 where a lot of the protective coating was etched off by the leakage at some point. One damaged trace has been repaired with a little solder. Kind of ugly in that one spot, but it tests out OK so it should work. And it does, for a while.

Meanwhile, this occasional rough starting is still really bugging me. So last week I got to thinking about what might confuse the ECU when it is starting. Having eliminated what I think is everything else, and replaced my coolant temperature sensors, I tested my ECT/electric fan switch and discovered it was faulty. I am not convinced that caused my problems, but it was bad so it had to be replaced. So I ordered a new ECT switch, and lots of Toyota red coolant. Yesterday I drained the coolant and installed the new switch. After the coolant refill and new switch it still had the starting problem, and did not run any better. But I still felt a little better about it because it had fresh, clean coolant and a working fan switch. No new issues to declare, but some kind of mild phantom is still lurking inside there. Drove it about 10 miles without any major problems.

Then, to my horror, when I started it this AM it almost didn't run at all. It was obviously misfiring, I had to throttle up to get it to start, and blowing tons of smoke out the exhaust. It smelled like pure fuel. It was horrible. I warm it up a little and check the coolant (I dont know why I check anything at this point LOL) and it looks OK. Needless to say I am not in the best frame of mind at this point, so I drive it around the block, see if I can get it to clear up. Yes, I gave it an Italian Tune-Up. AKA WFO. Amazingly, it clears up within one block, and it runs really strong. Making lots of power. So I park it, and come back to this forum, remind myself that the capacitors need to be low ESR, and soak myself in the regret of buying crap off ebay once again.

So I prep for another cap replacement (this time on the '93 unit that has been in my car since maybe March/April). I am thinking maybe that '93 unit with superficial leak damage is still salvageable. Since I still have leftover caps from the first job, I spent part of my afternoon doing another capacitor replacement. I finish it up, plug it in, and it runs like crap right off the bat. Did not want to start or stay running. I try to "teach it" to idle with my throttle and let it warm up a little. Then the idle stabilized so I take it for a spin. Dies in the middle of the road at 20 mph, full normal temperature. Sure, it starts right back up, but obviously it has issues that remain. And I have no idea what to do now.

Again I apologize for the giant wall of text, but I need your help with this problem, and I can't see stringing readers along with a cat and mouse diagnostic Q&A over the course of a dozen posts.

This has gone way beyond frustrating and well into the realm of humiliation. Any help is appreciated. Even if all you have to offer is a similar story, I'd like to hear some feedback on where to get a decent ECU.

Thanks in advance.

dicer

First off you of course are going to need a known "good" ecu, to rule that out. Next thing cam and crank sensors since you mentioned igniters. At least I think you did, I rarely read long posts and don't want to do it again. At least this post was interesting. Something else to check is what Yodaone has said about all the grounds. I also recommend cleaning the contacts for the fusible links where the wires connect to them. Not sure if the 94's have the cold start valve but if it does and they leak it would cause some problems of running too rich and that is a separate injector. Other things that will cause problems would be the idle control motor, and egr system problems. Then of course other normal stuff like the airflow meter and if you tried to clean it you then ruined it so that could cause some grief.
You've had lots of hoses off, any leaks?

And you know if you have TRAC you need a TRAC ecu, and if you don't you don't want a TRAC ecu.

Yamae

I only had a chance to read the initial part of your post above. What I assume from above is that the leaked QAS liquid damaged not only the traces of the surface but also the internal traces. The board is multi-layered and there are traces internally.
It is often too late to replace capacitors when they are already leaked. I recommend people to replace those QAS capacitors before the leakage starts checking the ripple voltage. There is a saying,"No use crying over spilled milk". I'd say, "No use crying over spilled liquid".

gymguy79

I'm not sure if the ECM boards are double sided or have additional internal layers (Yamae probably knows). If board has internal layers with plated through holes then replacing caps and repairing surface traces may not be enough if the internal layers are not connected. I would love to see a real schematic of the boards so I could determine possible results of damage caused by the caps. Replacing damaged caps may be required but as has been said may not be sufficient. A known good board may be the only permanent solution since other components on the board may be approaching end-of-life (as all of us do also) even if good caps and repaired traces are done.

89OLDMAX

Diameter, I did a 1990 ls400 cap change on a 89661-50010. It was good for about 6 months than started running extremely rich one day and would not keep running. I rechecked all the soldier joints in the computer and they looked ok. I solved it with a different computer. Of the 3 93-94 ls400's I have bought they all either have already rebuilt computers or the computer in them was bad. Just one trace corroded through and the computer is bad. Sounds like classic bad computer is what you have. Good computers are getting hard to find. I suspect a high percentage of original computers for the 93 and 94 ls400 are bad by now. I only buy used computers from cooler climate states.

Dannyjmck

If I wanted to replace my ECU instead . Are the aftermarket ones unacceptable? If some have potential to be an okay replacement does anyone know which ones?

gymguy79

My 1993 LS400 computer just died. After testing all possible causes except the ECM I removed it and opened it to find apparently pristine boards. Luckily I found a junk yard with a wrecked 93 with the front end smashed up to half the engine compartment but with the passenger compartment intact. It had been there for about 2 weeks so I took a chance that it must have been running before the accident and I removed it. It had warranty stickers on it so I knew it had been repaired and I bought it for $48. Plugged it in and now the 93 runs like new. I will replace the caps (on order from Digikey) and try it in the 93. I plan to get another for a backup for my 94 and repair it if necessary

TySoCal

Many thanks for this useful and very informative thread!! I just wish I had found it before I re-capped my ECM. Would have saved me some of re-inventing of wheels and lots of troubleshooting. Now here is my info for the statistics that I believe Mr. Yamae is still collecting.

Car is an SC300 (2JZ-GE) model year 1995, with automatic trans and TRAC. 161K on the odometer, Arkansas USA delivered, so no California emissions, just Federal. I got the car in 1998 with 50K on the odo and brought it home to Cali where it has been ever since. It has always been garaged at home but has been driven to my office almost every day since 1998. At the office it parks outside. In the summer the interior temperature easily reaches 45 degC or more with fairly low humidity.

Had typical symptoms: at first, driveability issues that resembled bad TPS, but no codes. As I was troubleshooting these issues I disconnected the battery several times for various reasons. At some point I re-connected the battery and found I had no CEL when ignition was on but not yet started. Engine would crank and briefly run but die due to lack of fuel. Jumper B+ to FP on DLC1 and it ran fine. No codes, CEL would not light at all with TE1 jumped to E1. Hmmm. After many hours work with the meter and o'scope I determined that the ECM was telling the fuel pump to run while cranking (FPC = 4.5VDC) but telling it to stop when running (FPC = 0V). Injector pulses, igniter pulses, various inputs coming from distributor, all looked fine, fuel pump ECU working properly and giving good status (DI) back to ECM, so no valid reason for ECM to be cutting off fuel.

The interesting part is that if the battery was disconnected for 12 hours or so the ECM would "recover" and work completely normally. But if I disconnected the battery for only a minute or two, the ECM would go back to cutting fuel, unresponsive to TE1, no CEL at all, etc.

Anyway. Re-capping the ECM fixed everything. Also the overall performance of engine and trans is significantly improved. I didn't even realize I had a performance problem - it had been deteriorating so gradually that I didn't even notice it.

Not having seen this thread, I selected the replacement caps on my own and decided to use Nichicon UPW series. I see at least 1 UPW cap in Cowboy's list on page 1 so I guess I chose okay. I based my choice on low impedance, high temperature rating, and supposed long life.

The ECM is 89661-24350. The capacitor complement is:

2 pcs. 10uF 50V
2 pcs. 15uF 35V (actually used 63V due to availability)
1 pcs. 47uF 63V
1 pcs. 100uF 10V
1 pcs. 220uF 16V

The old caps look fine at first but when inspected under magnification and good light there is slight leakage around the leads on two caps.

Now I'm thinking I should open up the ECM on my 1996 Dodge truck and see what sort of caps are in there. Unless they're all tantalums (doubtful) they are probably degraded due to age.

Yamae

Thanks for the report, TySoCal. So the thread was useful for a 95 SC300 owner like you. I'm pleased that your car runs well again replacing those QAS capacitors.

Your analysis is something super, I'd say. Using an oscilloscope and finding the root cause of the fuel pump issue can't be done for average DIYers. You must be a skilled engineer, I can easily assume.

Regarding your 1996 Dodge truck, if it uses regular electrolytic capacitors and not QAS capacitors, I don't think you can expect a lot changing those to new ones. Of course electrolytic capacitors loose the capacitance as aged in accordance with the Arrhenius equation and there may be some improvement changing those but it wouldn't be as big as your SC300, I bet.

BigGs300

Hello I have a 99 Gs 300 and it sounds like I have some of the problems that you have in your post. My question is where is my ECU located on my car that your talking about fixing the capacitors. Is it in the engine bay or is it the ABS and VSC ECU under the driver side steering wheel?

Spartan536

Hello, I own a 1992 Lexus LS400 with 86,500 original miles (all garage kept, all service work done by my local Lexus dealer as the car never even left my state). It was a 1 owner car (until my purchase) and is a rare find from what I understand.

That being said I have had a few issues that have developed shortly after my purchase.

1. Alternator died (replaced with Denso Reman)

2. Battery died (was 5 year old battery, replaced with Interstate Mega-Tron+ 810 CCA)

3. Front cooling fans not turning on, replaced the ECT sensor on the radiator, fans now work.

4. Strong exhaust smell (not sure if this fits the sulfur smell on the ECU checklist), but you can definitely tell it's strong.

5. Poor performance or performance not matching YouTube videos of other Stock Gen 1 1990-1994 LS400's, especially acceleration, and it has gotten worse as I have driven it.

6. Recently, erratic idle with low idle (low idle being 150 RPM), erratic idle being 150 RPM - 1000 RPM in "surges", problem gets worse as the vehicle is operating over a short time (appx 15 minutes), once left to "cool down" things smooth out again until that operating window is hit.

7. Climate Control system turns off and back on again at random (A/C compressor functions when turned on, cold air is blown when operating). I had my Climate Control dash sent to Tanin Auto Electronix and they found no issues, so I had them change out the LED color for the backlighting to White, kept the OEM green indicators. Tanin Auto took 2 weeks longer than expected but that was due to employment issues they were having at the time, their customer service was excellent however, they were very understanding and I pestered them quite a bit, they never showed any hostility, very professional would recommend.

I am currently sending my ECU out for repair to LScowboyLS on Facebook, which I am to assume is the well respected LSCowboy on these forums, friend of Yamae. I will be writing up a summary of what was fixed and how performance has changed when the ECU is returned and reinstalled. If it really is the same LSCowboy, I feel really lucky to have someone from these forums do the work and know what they are doing with amazing detail... He also apparently worked for NASA's Marshall Space Flight Center, in buildings 4727 & 4708 at Redstone Arsenal near Huntsville Alabama, so rocket scientist?

datarecvry

I recently bought a 1995 LS400 with 110K miles. Was sold because the car would severely stumble / lurch during shifting. I thought well might be the throttle body, or it could be the tranny mount (it certainly had that characteristic rumble). Finally started doing it every day, so i knew it was time to diagnose it.Chose the ECU first, because sometimes the OBD scanner would connect to the ECU and sometimes not. It was an ACE rebuild from 2010. Lasted 7 years.
Here's what I found when I opened it up.
I replaced these two with recommended LOW ESR and the shifting has greatly improved. Shifts smoothly, firmly and more quickly at lower speeds than before so if you have similar issues, please PM me or reach out to me

CELEXUS84

Sorry to bug again.

But i found a used 89661-50234 and upon doing a bench test on it the check light was solid and opening it I found 1 cap that has black crusty under it. I also have a computer that one of my team mates brought to me thinking it would work but it does not, a 89661-50150. On the same bench test I can get it to go into diagnostic mode and produce check codes (link to video ) So I believe that proves my bench test is wired correctly. This computer im assuming is wired differently as far as ignition etc goes because it just cranks in the car and no spark etc.

Now I have emailed one rebuilder asking if they "test" the ecu for functionality and they basically called me a idiot and said there is no way to test a ecu out of the car...Thoughts?