Engine stalls at 3lbs of boost, but only sometimes?

[solei]

Alright, I'm completely out of ideas, and I need help.

I've never owned a car with forced induction before, so most of this is completely new to me. I cannot comprehend what would cause this to happen, and what I do not understand I can only dream of fixing.

Currently, when I drive my Supra, It's perfect in almost every way. The engine purrs all the way up to 5,000 RPMs (I'm sure it would purr farther then that, but I don't like to push an engine that hard), and it comfortably does over 100MPH on the highway... but as soon as I push more then about 3 lbs of boost through it, it starts to stutter, or half the time, will completely die as if the fuel was cut (until I pull my foot off the gas, and let the turbo spool down). Once the turbo has spooled all the way down, it immediately continues to drive normally- as if nothing had ever happened in the first place. The thing that completely blows my mind is, sometimes, I'll floor it, and it'll hit 5 pounds of boost without a single glitch, as if nothing was wrong (though this phenomenon is becoming more and more rare).

I've checked all the lines off the turbo, they all seem to be in good repair- and I've got a hose techniques kit coming from Suprasport, just to be sure the problem isn't there..

My limited amount of research makes me think either the ECU is freaking out for some reason (air sensor?), or there's something wrong with the waste gate, but I'm not even sure where that is.

Could anyone confirm my suspicions, and tell me at least if I'm on the right track? Pointers would be greatly appreciated.

 

[Nick M]

Do you have a big cloud of black smoke with the stall?

 

[oscolivar1]

out turbos are external. Its your can't miss the wastgatge....you see it when u look at the turbo.
http://cgi.ebay.com/ebaymotors/Used...742QQitemZ170088667819QQrdZ1QQsspagenameZWDVW

I just googled this to show you the wastegate. It's the one in black.
Have you checked for CEL (check engine light) even if it's not lit up you can still have codes popping up.

I thought it may be a boost leak, but once u said it does it once in a while... hmm i dunno..... check all your IC pipes, especailly the one connecting to your turbo.

Any smoke thats visible?
If so whats the color of it?

 

[1988SupraDreams]

Throwing this out there: How much fuel is in your tank when it stutters?

 

[solei]

Do you have a big cloud of black smoke with the stall?

I have yet to notice a black cloud behind the car when it stalls- I was talking with a friend who knows more then a little about turbos, and he asked me the same question. If there's a cloud coming out when it kicks back in, it's so thin I don't notice it.

I just googled this to show you the wastegate. It's the one in black.
Have you checked for CEL (check engine light) even if it's not lit up you can still have codes popping up.

Thanks for the information- I'll poke at that tomorrow. I also keep forgetting to check and see if it's throwing codes- I'll also do that first thing tomorrow.
EDIT: Code checked- it's 21.

I thought it may be a boost leak, but once u said it does it once in a while... hmm i dunno..... check all your IC pipes, especailly the one connecting to your turbo.

That was one of my first thoughts, too... but I figured if it was a boost leak, the problem would be consistent. I checked the Intercooler hoses, and all of them appear to be intact as well.

Throwing this out there: How much fuel is in your tank when it stutters?

That was my first suspicion- I noticed it started happening when I was getting close to empty- I brought it to a gas station and filled it up, only to have the same problems continue throughout the day.

 

[Nick M]

Do you have access to a DSO? That is a fancy name for a hand held scope. There are a couple of things you could check.

 

[solei]

Do you have access to a DSO? That is a fancy name for a hand held scope. There are a couple of things you could check.

I don't think so, unfortunately.

 

[solei]

I just checked to see if it was throwing a code: It is: 21.

According to the TSRM, that's an oxygen sensor- I'm either running too lean, or too rich. From what I gather, if I were running too rich, I'd see black smoke when it kicked back in, right?

I had dinner with a friend who owns a Conquest tonight- he said the way his injectors worked was he had two in the throttle body- one for regular driving, and the other which kicks in when it needs the extra fuel- it switches from one to the other. The high performance injector was bad, and that caused his car to produce the exact same problems mine is producing now- are our injectors similar in any way? Could that have anything to do with it at all?

 

[Nick M]

No, ignore his setup.

The oxygen sensor won't cause a stall.

What else do you notice. When you hit 3 psi, WOT only, what about RPM? Same place everytime?

 

[solei]

What else do you notice. When you hit 3 psi, WOT only, what about RPM? Same place everytime?

I can rev the car all the way up to redline without any glitches whatsoever, so long as I don't push more then 3 lbs of boost. I used to notice it only happened in the higher RPMs (4k+), but now it's all over the place.

 

[Nick M]

So you can go half throttle in high gear, make 3 psi, and the car stalls?

 

[solei]

So you can go half throttle in high gear, make 3 psi, and the car stalls?

It only appears to do it at full throttle, but I only say this because I can't recall a time it's happened when I wasn't flooring it to see if it would happen.

I worked with the kindest person I've ever met today, who helped me do a bunch of tests- we determined the problem is likely in fuel pressure- though the code it keeps throwing is a 21- which I understand is a bad oxygen sensor? The thing I don't get is, it passed smog 2 weeks ago, and with some testing in emissions, we determined the oxygen sensor should be working just fine. Very strange...

 

[Nick M]

Did you monitor fuel pressure while driving? I wonder if you have problem in the 9-12 volt switching relay. Maybe you are going from 9 volts to 0 volts instead of 12.

I wasn't originally thinking that, but since you seem to have a load problem, and not an RPN problem, that is a good place for you to check. I thought maybe you were losing the pickup for a moment. A code should be set with it, but a scope lets you look at the actual signal.

 

[jetjock]

Code 21 is set when the O2 signal fails to meet either the minimum or maximum amplitude (or both) for a certain period of time. Can be due to an open, short, or sensor degradation with age. The code can also be set when the heater circuit exceeds a high or low resistance threshold ie; an open or short. Start by checking the sensor wiring and connector for damage and be sure the bonding wire from the head to the firewall is connected.

The stumble does sound fuel related. The TCCS stays in closed loop during mild boost at less than WOT so it could set an O2 code under those conditions but in that case I'd expect a 25 or 26. Verify the stumble doesn't happen when boosting without being at WOT. You can also try unplugging the sensor although I doubt it'll help. Expect a code 21 while doing it and don't run that way any longer than it takes to get results. Clear any code afterwards.

I'm with Nick about fuel pressure. A snap throttle test with the regulator vacuum line disconnected is a quick way to check for problems in the fuel system. If pressure dips more than a couple of psi there's a restriction somewhere.

 

[Nick M]

One thing that is different is the stalling at one specific point. *Usually* that has been pickup related in my experience. I also think you should sweep the TPS, but that is not possible without a scope.

 

[solei]

Did you monitor fuel pressure while driving? I wonder if you have problem in the 9-12 volt switching relay. Maybe you are going from 9 volts to 0 volts instead of 12.

I wasn't originally thinking that, but since you seem to have a load problem, and not an RPN problem, that is a good place for you to check. I thought maybe you were losing the pickup for a moment. A code should be set with it, but a scope lets you look at the actual signal.

You were completely right in this guess!

I noticed after playing with it for a week that when I drove with the fuel pump jumpered in the check terminals (in the same box as the terminals one jumpers to get a code readout), that the problem went away entirely. So a friend of mine and I went in and tested the electrical, working back from that point, and found the relay was bad.

We took the thing apart to see why it had failed- there's a copper strap that goes from the main structure of the relay to the actual switch peice, which carried the current from one to the other. There was also a small spring holding the switching piece back against one side when there was no current flowing through it. The copper strap had broken off the piece, and the spring ended up carrying the current in it's place, which made the very thin metal of the spring melt, which killed the whole thing. The strap must have been making partial contact until we took it out!

 

[kapkaniMD]

its weird, my car is doing the exact same thing and i replaced the fuel pump relay with a bosch type 40A relay and the problem is still here, I havent tried jumping the terminals I will probably give that a try now though..