Bylan

1998 LS400, 196k miles, maintenance records from Lexus/Toyota from 1999 to October 2018
First off, as always, I thank you for your time reading this, and for any advice you have. This forum is amazing for helping us all enjoy our LS's. Now onto the oddities that my LS is experiencing. I bought this black/black 98 LS with Nakamichi sound system around 2 months ago. Ive done the whole timing belt and water pump kit, replaced valve cover gaskets, drained and filled the ATF once (more drain/fills will follow as I am able to put some miles on the car), and replaced some of the basic stuff like temperature sensor, PCV valve, all the vacuum hoses, and the transmission mount with OEM. When all is well, this car is butter smooth and kicks you back into the seat, I have an obsession with these cars. I also have a 97 LS that I bought 4 years ago and I've done most of the same maintenance and little fixes to it.
With THIS car, it randomly refused to start after stopping at a friends house for 10 minutes. It acted like it had a dead battery, that fun click*click*click when I tried to start it. NOT a single click, indicating the starter. I went through the basics, swapped the battery with the Lexus OEM out of my rusty trust 97. No dice. Cleaned the battery terminals, and replaced the positive terminal. No dice. Checked every fuse in the car, all good. At this point, I replaced the starter. Not a terribly difficult job, and it made no difference. Still click*click*click*click. I walked away from it for a while, took a vacation to Miami, and then had a lot of work and summer classes to contend with before I could jump back into it. On a whim, I reattached the battery and tried to start it before I tore off the intake manifold and checked the wiring to the starter. It started right up, no Check engine light, no weird noises, nothing. As i was refilling the coolant and bleeding the system, the car shut off, and a Check Engine Light for code B2797 came up on my OBD reader; B2797 Communication Malfunction No.1 (immobilizer). I cleared the code and the car started right back up, and my suspicions grew. I replaced the key fob battery immediately, although I'm quite positive the key works off RFID and the fob battery being dead would not effect the starting of the car, since the valet key I have has no such battery and is just a solid key. I knew these issues weren't over with, so I've refrained from driving the car more than just around town and 5 minutes to work. Yesterday, the car developed a misfire, very obvious with such a silky smooth engine. I took it for a short drive and after 5 or so minutes, the Check Engine Light began to flash and I checked the codes with my OBD reader. misfire on cylinder 6. I inspected the coil connector to cylinder 6, clean and dry and fully seated to the coil. It was late at night, like 1am, so I left the investigation for today. I started up the car to drive to work in the morning and check for a misfire and it was running silky smooth again, so I drove it to work. At the 1st of 2 stoplights, it began to slightly misfire at random times, and the idle would drop to 400rpm or so. It did NOT display a check engine light, and my OBD reader found no stored codes. I parked the car after driving very gently to work, and drove it home on my lunch break to grab my other LS. Again, after a few blocks and stopping at a stoplight, it would randomly misfire, and the stereo cut out for a brief second, something I would not think is related to the misfire, but more so an electrical issue. I've experienced odd electrical blips and stereo cut-outs in my LS because it had the infamous leaking P/S pump and killed my alternator. This 98 however has no P/S leaks, and the pump and alternator were replaced at a Lexus dealership in 2016, they look to be in great shape and the battery shows over 14v while running at all times. After work, I let the car idle in my driveway and get up to operating temp, but it refused to misfire. Occasionally I would feel slight vibrations from the engine, but no bucking like at stoplights. After idling for about 10 minutes, the idle would slightly fluctuate, at one point rising to 1000rpm as if the A/C compressor had just turned off. It was hot out, so I had the A/C running. Then the car shut off quite suddenly and I expected that code B2797 to come back, but my OBD reader showed no codes. It has no problem connecting to the ECU and displaying all the information the Torque+ has to offer (awesome app, highly recommend) and the car has no CEL. Again it restarted with no problem, and I started to test all the other electrical systems after this, and everything seems to be fine, although after repeatedly locking/unlocking the doors with the key fob, it began to refuse to respond. I tried using the door-mounted lock/unlock switch, and after a few cycles all the door locks got sluggish and stopped responding. I suppose I could contribute that to the lock motors not wanting to run that many times in a span of 15 or so seconds, but ALL the locks did the same thing. This culmination of strange happenings has me looking at the infamous ECU capacitors failing. I know there are threads upon threads all over the web about this, but I'm always interested in hearing what others have to say about it. I pulled the ECU out of the dash today to inspect it, will open it up probably tomorrow. I will of course also be checking my spark plugs (probably just replace because they're of unknown age and mileage) and testing the resistance of the ignition coils, specifially on cylinder 6. I have a rather thorough set of PDF shop manuals for all this stuff, I will not just be throwing parts and money at the car. Its my 2nd LS after all. The culmination of symptoms has me looking at the ECU or some other subsystem that is causing mayhem somewhere. I will absolutely update after I peek at the ECU, and probably send it out for repair if anything looks sketchy (or attempt a DIY repair myself, I know how to solder and am not afraid to make a mistake and pay for it). Thanks a lot for reading, I know it's long, but I also know you need a lot of info to draw any conclusions. May we all help each other enjoy our 20+ year old leather-lined-boats!

Fit1too

I just looked over the symptoms list of a bad ECU and you do have several. I'm interested to learn what you find out when you open yours. It has to be that, a loose ground, or a short circuit of some sort. We'll work it out!

Bylan

I have removed my ECU, and disassembled it the other day. No signs of leaking capacitors, but I'll be looking at professionals to replace my caps anyways. I know they don't always visually fail, and being 21 years old and living in an all black interior, they've definitely been cooked. Weather has been either 95F or raining the past few days so I haven't removed coil packs and spark plugs, but my service records from Lexus show the last plug change almost 100k miles ago so I'll be replacing them once I have $ to throw at it. If my random misfire, AC seeming to shut off for a few seconds here and there, and lack of a CEL for the misfire, and the car randomly shutting off parked and idling are NOT the ECU I'm going to be really stumped at where to troubleshoot. Cars are fun right?

Yamae

Your symptom "click*click*click*click" is nothing to do with the ECU since the ECU does not control the starter motor at all. It's simply driven by the battery via switches, a relay and a solenoid with wires that connect those.

If I were you, I would check the ripple level of the ECU first before suspecting it. Some people quickly jump to the ECU but in case of 98-00, the initial symptom of a failing ECU was always the AT shift problem in my past experience.

Bylan

Im so glad you you took the time to read of my troubles and comment. Means a lot, you are a master on these forums. Whats really strange is that the starter click seemed to solve itself. I know it really didn't, but I didn't touch anything at all and it started fine after sitting for a few weeks. I cleaned up the terminals even more after it started, and it has started dozens of times since then without a single hiccup. The new starter I installed is also quieter than the old one and turns the engine over with a lot more energy. I have read LSCowboy's post on testing the ripple level of the ECU. It requires "a meter that can measure up to 300KHz or higher. and measures the AC voltage (~) between "OP1" and "E1" in series with a film capacitor 0.47uF". I'm comfortable with the mechanical side of things, but electrical beyond volts and amps confuses me. I understand that this test measures how much voltage is running between these ports, and that you want a read out (engine warm) of 40mV RMS or less. What do I need to look for to know my meter can read very low voltages accurately? If there is another ECU test thread that may explain differently, or if you'd be willing to explain the testing procedure a little more, I'd appreciate it.

Yamae

So your starting issue is OK now but I worry that your starter's solenoid contacts are not fully OK. The larger contact (connected to the battery side) tends to wear out faster than the smaller one as is shown below and your symptom may be the initial indication of this problem. I'll post about the ripple issue after finding a photo of an actual waveform of the ripple.

Bylan

Just to be clear, I already replaced the starter, when all other troubleshooting didn't lead to any changes. Since I figured I needed to inspect the wiring from the battery to the starter, I bought a Denso re-manufactured starter off Amazon. I know re-manufactured parts these days are hit or mostly miss, but getting a hold of an OEM starter is just way out of my price range, and the starter I removed that was installed by Lexus in late 2016 was also a "Denso Re-manufactured for Toyota" starter. My inspection of the wiring to the starter didn't yield any signs of corrosion, damaged wiring insulation, or mangling from past mechanics. The car was serviced by Lexus it's whole life with every little thing being done by the dealer, even a sticky rear seat cup-holder mechanism. I am familiar with the starter solenoid contacts often being the failure point, and not the motor itself. Perhaps one day I will take apart the starter I removed and rebuild it, and replace the solenoid contacts so I have a backup or make some money off the otherwise good core. Just to reiterate, I replaced the starter and inspected all the wiring I could access and found nothing wrong, and the car still refused to start. It simply decided to start after sitting disconnected from the battery for a few weeks. Since then, I've only ran about a tank of fresh gas through it when it began to exhibit the random engine shut-offs and then the misfire on cylinder 6.

I read through the ECU test a bit more and It seems quite simple, I'm just unfortunately not well versed in AC electrical terms. I think I have a good enough understanding of how to run the test, and will be sourcing a .47uF capacitor and taking a look at the multimeter I have to see if it is capable enough to get an accurate reading for the test. Weather and life haven't been cooperating with my hobbies, so I haven't had a chance to get out to the garage and do some wrenching/troubleshooting. Hopefully tomorrow. I would still be very interested in learning more about the waveform of the ripple, thanks again Yamae.

Yamae

So you think your starter should be OK. I tell you what. It was a alternator and not a starter but a friend of mine once had a bad experience using a re-manufactured Denso alternator. The core was the Denso alternator but parts used were non Denso's ones. I hope your starter is re-manufactured using genuine Denso's parts.

Regarding the ripple level, attached below are the actual waveform when I measured my 98 Celsior's OP1 and E1 last October. The ripple level is around 15mVp-p and I can confirm that there exists 2.4kHz energy and others. The best way to measure is to use an oscilloscope like I did. The ripple here means the AC (Alternating Current) component. In other words how much voltage swing is there. If those capacitors are good and the internal impedance is zero, there will not be any voltage swing. But there exists some impedance and the voltage swing will not be zero. Aged capacitors have higher impedance and the voltage swing (ripple) can't be absorbed by them and the ripple remains. The 5.0V DC voltage is used as the reference voltage of many AD converters used for sensor lines and bigger the voltage swing, more often the error occurs.

The reason to use a film capacitor 0.47uF is to eliminate the DC component and only AC component passes through enough. The highest frequency of the ripple is higher than 10MHz but major components are much lower than that and I selected 0.47uF in accordance with the result of experiments. If you can't use a multi-meter which can't detect higher frequency like a few hundred kHz, you may use a cheap one but in that case the limit level would be smaller than 40mV.

Bylan

Thank you Yamae for the info on the ripple effect of the ECU. Im going to have to drop some $ on a more accurate multimeter to perform the test, but at this time i DO NOT think my problems are ECU related. I believe I have solved the random engine-stalling and misfire, some reddit users pointed me in the direction of the crankshaft position sensor. I tested the resistance of mine and found it to be 2180ohms, far above the 835-1400ohms that it should be. On this very hot day, the car stalled out after idling for less than 60 seconds. I swapped over the Crankshaft position sensor from my 97 LS (identical parts) and tested its resistance; 1100ohms on a 95F day so it is within spec. The car now runs and idles fine, I let it run for 15 minutes and get thoroughly up to temperature. I will be purchasing a Denso OEM crank sensor and installing it to ensure I don't experience another failure like that. Will definitely post updates as the next issues occur, knock-on-wood that they are few and far between now. I also got offered a position at my local Lexus dealership, so I'll be gaining some inside knowledge and get to work with their master LS tech. I thank you all for your help, and not being shorthanded and defensive about someone trying to learn and fix their car themselves. That feeling of accomplishment after getting something to work is a lot more rewarding than paying someone $120/hour to do the same thing.