Overboost code 34 when not at high pressure?

[underscore]

1991 GTFour RC, it's having an intermittent problem where it will trigger fuel cut and throw a CEL for code 34 when the car isn't actually making much boost (if any). It's happened a couple of times when doing longer climbs up hills at 3-4psi and also once while turning at a traffic light. All just moderate throttle situations. After it happens the car seems to be ultra sensitive until I restart it. My boost controller has a very conservative boost cut function so I unplugged the factory MAP sensor on the firewall and the factory boost gauge is now pegged at max. I figured this would take care of the problem for now but it just happened again and now I'm at a loss. I was figuring the sensor was defective but now I'm thinking maybe the ECU is being flakey? How else can it be throwing a code for overboost without a pressure sensor?

 

[Gert]

Perhaps time to open the ECU and check for bad capacitors?

 

[93celicaconv]

Perhaps time to open the ECU and check for bad capacitors?

I agree. If the ECU hasn't ever had a capacitor problem, consider yourself fortunate, but in most cases, that is one area to check on. When you unplugged the factory MAP, did you also get a check engine light to come on?

 

[underscore]

No CEL with the MAP unplugged, which from what I've learned is strange because there should be a code 33 for that?

The caps in the ECU were replaced a couple of years ago, but they could be defective/incorrect/I messed up the assembly. I have a second ECU that got the same treatment that I can try putting in. After I made this post I drove about 1000km over the weekend without issue which was strange.

 

[93celicaconv]

I believe starting an engine with the MAP disconnected immediately turns on the check engine light with DTC 35 (Open or short detected for 0.5 seconds or more in MAP sensor signal circuit PIM). If your check engine light did not come on (assuming your check engine light works and it illuminates when starting your engine and stays on for a second or two after the engine has started before turning off), I would say that there is either a problem in the wiring between the MAP and ECU (so that there is some kind of reading being made on the PIM circuit by the ECU) or a problem with the ECU itself. Given your boost gauge pegs to maximum when your MAP wiring harness is disconnected (which it should) makes me feel the wiring circuit is likely OK. But I don't know how an ECU responds when it is no longer in control of the turbo boost VSV or fuel cut outputs (given your add-on boost controller now takes over with that). Something isn't right.

 

[Gert]

Indeed very strange.
Is your ECU tuned once, perhaps?
Back in the days, I did sent my stock ECU to an company called G-Force and they made changes in the ECU, like fuel and ignitionmaps and they could alter boost settings too, like fuel cut level, etc, etc.

Otherwise, I really don't know. Our cars are getting old and the electronics too. Time for an programmable ECU I suppose

 

[underscore]

The CEL does seem to work, it flashes out codes, it comes on with the other lights when the key is turned to "on" but when the engine is started it turns off immediately. I double checked that I unplugged the right thing and the part number on top is for the "SENSOR ASSY, VACUUM" so that seems right. Looking at the vacuum diagram it was hooked up to the wrong spot on the manifold (one of the charcoal canister ports) which I'll correct soon, but that still doesn't explain the weirdness with the codes.

I didn't see any markings or modifications to the ECU when I had it out but who knows what the previous owners did in Japan. It does have an external top speed delimiter so I'm assuming that wasn't tuned out at least.

As far as going to an aftermarket EMS, I'd really rather not. I've had enough friends install them and have all kinds of headaches caused by them. If anything I'd maybe switch to the ST205 system but at this point those parts aren't much newer and I imagine they're pretty pricey now that the cars are US legal.

 

[underscore]

After a year of fighting with this and failed searching for answers I managed to this thread on MR2OC: https://www.mr2oc.com/threads/ecu-rom-code-exposed-overboost-fuel-cut.196813/

I'll copy the first post here in case the other forum wanders off or breaks the link

I have been spending quite a bit of time lately reverse engineering the 3SGTE ECU ROM code. There is a lot of interesting stuff here and some day I might give a complete high level overview of what the code is actually doing but for now I will post on sections that I found noteworthy.
Here is the pseudocode describing exactly how the fuel cut mechanism is implemented in the ECU ROM code (warning, serious nerd content here, you should skip to my comments at the bottom if you do not have a computer science degree):

Of note is that there are two 2D maps that drive fuel cut. The first is a flow vs. RPM map. This guards against a faulty turbo pressure sensor (PIM). This map has only two points to describe a line (via interpolation). A flow of 84 is required at 3200 RPMs and a flow of 76 is required at 6000RPMs. Flow is calculated roughtly as the AFM sensor value divided by RPMs. Once fuel cut occurs once, any flow above 80 initiates fuel cut immediately until the ECU is powered down and restarted.
Then the second 2D map is used to initiate fuel cut based on the turbo pressure. This map has three points, 221 at 3800 RPMs, 215 at 5000 RPMs and 205 at 6000 RPMs. This means that fuel cut is not initiated at the same exact boost level across the board, but that slightly less boost is tolerated at higher RPMs.

So it turns out it can also be triggered by the AFM, which explains a heck of a lot. I'll swap in my spare and see what happens.

 

[Gert]

Very curious now... let us know.

 

[underscore]

The original AFM tested fine according to the BGB but I swapped it anyways. It'll take a while for a conclusive answer since this problem was intermittent to begin with but the car can actually run now. On a short test drive it was back to feeling the way it did before the problem first started happening so I'm hopeful.

 

[underscore]

So this kept happening even with the spare AFM. I went back to the original (the car ran very rich on the spare) and I've been able to experiment with this enough to be able to consistently trigger it. If I fully open the throttle below 3000rpm in a higher gear, when it hits ~3200rpm I get fuel cut. If I'm easy on the throttle til I pass ~3700 or so, then go wide open, no problem. Still weird but at least I know when I need to chuck it down another gear to avoid problems.

 

[RedCelicaTRD]

Have you swapped the map sensor? If you have an oscilloscope have you monitored the map signal?

I’ve ran over 20 psi on the stock ecu (not recommended!) and used a homemade FCD to handle boost cut. I never had an issue with boost cut with an FCD. I’ve run over 20 psi at 3000 rpm with a modified ct26, so I don’t really buy that the afm is causing fuel cut. I don’t think that has ever been verified by anyone but Ricky. I would assume the afm value is used after fuel cut has been initiated, not as something that can trigger fuel cut.

If you’re trying to bypass the map sensor, don’t unplug the electrical connector. Just unplug the vacuum hose to the sensor, or add a check valve to that line. The ecu will view either open or a short as something that should trigger fuel cut.

Of course yours is an RC. The afm and ecu are different.

 

[underscore]

I haven't swapped the sensor as I don't have a spare but I'm not sure if that would do anything. Unfortunately no scope on hand either.

I've run it with the MAP sensor capped and with it unplugged and there's no change. No CEL with it unplugged either. My boost controller/gauge logs peak boost and it'll do this without getting above 0psi. Once it happens it becomes extremely sensitive and you have to be very gentle on the throttle or it'll trigger the cut again. I'll restart the car while rolling to clear it and then do a pull over 16psi no problem.

Now that I think I can do it at will I'll see if I can get a video of it happening. When I get a chance I'll try hitting the AFM wires with a meter and see if there's anything weird between the plug and the ECU.

 

[RedCelicaTRD]

But every time this happens the CEL comes on (until restart) and it’s always code 34?

 

[underscore]

Every time I've had the chance to pull the code it's been 34.

Fuel cut will hit, the CEL comes on and it then becomes easier to trigger again. It's hard to explain but you have to be very easy on the throttle or it'll fuel cut again. I think it behaves differently than when I hit fuel cut for going over 17psi, but that was on the old motor 7+ years ago now so my memory could be wrong.

 

[Gert]

I am lost here... 34 is triggered with the map sensor only and with no boost line still code 34?
That cannot be related with this sensor then, yes???

Lookes like the trigger signal gets lost or something like that. That one is coming from the distributor. If that happens, she hesitates too, but that does not explain the 34??

 

[underscore]

I am lost here... 34 is triggered with the map sensor only and with no boost line still code 34?

Yes. Which is why it's so confusing.

That cannot be related with this sensor then, yes???

That's what I would think, because how can the sensor trigger something when it's not even plugged in?

 

[RedCelicaTRD]

You should have the sensor plugged in electrically, but the vacuum hose disconnected. The ECU can sense if the sensor is disconnected and will do fuel cut randomly if the sensor shows an open or short.

If the sensor is plugged in electrically, it could be failing and randomly going short or open, which will trigger boost cut. Or the wiring is damaged between the sensor and ECU and could be shorting out or going open.

 

[Gert]

You should have the sensor plugged in electrically, but the vacuum hose disconnected. The ECU can sense if the sensor is disconnected and will do fuel cut randomly if the sensor shows an open or short.

I thought he did this, but maybe I did read it wrong?

I cannot test it with my own car, since the stock ECU is long gone.

Haven't you tested it with an zener diode?

cheap fuel cut defender

www.mr2oc.com

 

[underscore]

I did try that but I think that was on the spare AFM. I'll try it again with the original one since it seems to run better on it.

I haven't tried a diode yet but that'd be an easy thing to try too.