Bored Legacy Posted January 4, 2019 Share Posted January 4, 2019 Where to start...We built a 1998 Legacy to race with Champ (Chump) car and WRL and we have raced it on and off over the last 5 years. It was an automatic and we swapped to a 5MT. No problems with the swap. We built a high compression Frankenmotor, and ran it several races with no issues. We built a 2.2 bottom end with EJ25D topend and turbo with an AEM FIC. We had tuning issues but the car ran and we were headed in the right direction with it. Then, I got divorced. My good friend and I had done all the work on the car. The team took the car and another guy rebuilt the transmission and the 2.2 bottom end. He also totally removed the ABS system and the entire combination gauge cluster, along with a bunch of wiring. When they tried to race, the car had a rev limit of 4000 rpm. I now have the car back. We built a circuit to simulate the 5 volt signal on pin 83 to the ECM for the VSS. We also jumped pin 82 to ground so it does not think it’s in neutral all the time. We still have the 4000 rpm limit, and it’s running rich. Does the ECM need some signal from the ABS controller? Any ideas? I’m at a loss at this point. Any help is appreciated. Link to comment Share on other sites More sharing options...
1 Lucky Texan Posted January 4, 2019 Share Posted January 4, 2019 cam/crank timing OK? Link to comment Share on other sites More sharing options...
GeneralDisorder Posted January 4, 2019 Share Posted January 4, 2019 Subaru's of that era have no software imposed rev limit. Even in neutral. In fact Subaru has never really put in a soft touch rev limiter. It will hit a hard rev limit but that's certainly above 5500. Probably around 6k. GD Link to comment Share on other sites More sharing options...
jonathan909 Posted January 4, 2019 Share Posted January 4, 2019 (edited) I really don't know $h!t about $h!t, but I'm kind of curious about this. Assuming that by "We built a circuit to simulate the 5 volt signal on pin 83 to the ECM for the VSS." you mean you built an oscillator that feeds a pulse train to the VSS input, why would you go to all that trouble instead of just connecting the transmission's speed sensor output to it as usual? (I was also briefly confused by this, as "VSS" has a totally different meaning in my world - it's a power supply rail, usually ground.) And is p.81 grounded to tell the ECM that it's connected to a 5MT? Edited January 4, 2019 by jonathan909 Link to comment Share on other sites More sharing options...
Bored Legacy Posted January 4, 2019 Author Share Posted January 4, 2019 Thanks for the replies. We verified the signals from the cam and crank sensors. We have not checked the timing and they did install a new timing belt during the rebuild. We will do that. From what I understand stock this car was speed limited to 120 mph. If so there is some limiter in the software. I had read that if the VSS signal is missing it will default to a 4000 rpm limit. We built the circuit to give the ECM a speed signal. We do not have a gauge cluster, since they removed it. Yes pin 81 is grounded for the 5MT. I will pull a gauge cluster and an ABS module and plug everything back in and see if that works and check the cam timing. Would the cam timing being off cause it to pull easily to 4000-4500 and then fall flat? Link to comment Share on other sites More sharing options...
jonathan909 Posted January 4, 2019 Share Posted January 4, 2019 (edited) 4 hours ago, Bored Legacy said: From what I understand stock this car was speed limited to 120 mph. If so there is some limiter in the software. I had read that if the VSS signal is missing it will default to a 4000 rpm limit. We built the circuit to give the ECM a speed signal. We do not have a gauge cluster, since they removed it. Yes pin 81 is grounded for the 5MT. Oh, I get it. You're feeding it a pulse train slower than normal to make the ECM think the car's groundspeed is lower than it is. I can't speculate on what other implications that might have. Any idea what normal is, in Hz/mph or something? What do you think your top end speed should be? [Edit] That's interesting - the xmission's speed sensor output doesn't go straight to the ECM. It gets fed into the speedometer circuit (in the gauge cluster) which in turn feeds the ECM VSS input. That could either just be a wiring "convenience", or the speedo circuit is conditioning that signal somehow. Seems to me the easiest approach to debugging this would be to unplug your VSS hack and put a gauge cluster in to see if it behaves differently. Edited January 4, 2019 by jonathan909 Link to comment Share on other sites More sharing options...
Bored Legacy Posted January 4, 2019 Author Share Posted January 4, 2019 That’s what we plan to do. The sensor in the transmission is a transducer. It takes the mechanical movement of the gear in the transmission and converts it to an AC waveform that goes to the combination gauge cluster. The circuit in the cluster converts the signal to a 5 volt DC square wave used by the ECM. Our frequency generator operates randomly from 30 to 40 Hz. And replaces the signal that should come from the gauge cluster which we don’t have. I’m an industrial electrician, controls guy. You apparently have an electrical, electronics background? Link to comment Share on other sites More sharing options...
1 Lucky Texan Posted January 4, 2019 Share Posted January 4, 2019 (edited) do the axles still have tone rings? maybe you could create the proper signal from one of the old abs wheel sensors? once you discover if it needs some kind of scaling.... - it would be speed (though, not rpm) proportional too. Edited January 4, 2019 by 1 Lucky Texan Link to comment Share on other sites More sharing options...
jonathan909 Posted January 5, 2019 Share Posted January 5, 2019 7 hours ago, Bored Legacy said: That’s what we plan to do. The sensor in the transmission is a transducer. It takes the mechanical movement of the gear in the transmission and converts it to an AC waveform that goes to the combination gauge cluster. The circuit in the cluster converts the signal to a 5 volt DC square wave used by the ECM. Easy enough to confirm with a scope whether in fact the speedo circuit is just doing a simple waveshaping/level shifting without otherwise altering the signal. 7 hours ago, Bored Legacy said: Our frequency generator operates randomly from 30 to 40 Hz. And replaces the signal that should come from the gauge cluster which we don’t have. I’m an industrial electrician, controls guy. You apparently have an electrical, electronics background? Unlikely, I think, that it's "random". Random is actually pretty hard to do, and I don't see how it would be useful in this case. More likely that it's variable or adjustable or something that's easy to implement with a 555 timer or similar. You wouldn't happen to have the schematic for it, would you? Might be helpful at some point. And yeah, I'm an electronics geek, mainly embedded systems (like this one). I'll be interested in hearing the outcome of the test with the VSS fed through the speedo. Link to comment Share on other sites More sharing options...
GeneralDisorder Posted January 5, 2019 Share Posted January 5, 2019 (edited) That's all cute and academic and so forth. But the reality is there IS NO REV LIMITER @ 4k on the stock ECU. Ya got that? Jeezuz. Look - I understand the frustration but I build these cars for a living and one thing we do after every water pump, or HG job (like three a week for many years now) is rev the engine to clear air pockets in the cooling system once the oil temp come up. We rev them past 5,000 RPM's about 10 times while in neutral on the lift. So I can tell you with 100% Absolute authority that there IS NO REV LIMITER on a 98 Legacy @ 4k RPM. GD Edited January 5, 2019 by GeneralDisorder Link to comment Share on other sites More sharing options...
Bored Legacy Posted January 5, 2019 Author Share Posted January 5, 2019 Ok....lol thank you for the information would the cam timing being off cause it to not rev above a certain RPM? I am frustrated, since when I had the car I knew what was done to it. It’s hard to troubleshoot it after so many things were changed and then it’s handed back to me. Link to comment Share on other sites More sharing options...
1 Lucky Texan Posted January 5, 2019 Share Posted January 5, 2019 the thing with cam timing is, it 'could' act like floating valves. so I'd think 1 or maybe 2 teeth off somewhere might get you a car that sorta runs up to a point, then flops. low fuel supply might do it too but????? I'd even check voltage at higher rpms if the above didn't find an issue. start with the easy basic stuff. Link to comment Share on other sites More sharing options...
jonathan909 Posted January 5, 2019 Share Posted January 5, 2019 GD, I heard you and believe you. I'm just keeping my mind open at the moment to the possibility that this faked-up VSS signal is otherwise confusing the ECM and causing some other unpredictable behaviour unrelated to rev-limiting. Link to comment Share on other sites More sharing options...
GeneralDisorder Posted January 6, 2019 Share Posted January 6, 2019 That's possible I suppose. Don't hookup goofy electronics to the ECU. I've seen a mis-wired secondary O2 sensor cause the ECU to freak out and randomly idle at 3000 RPM with no codes or other reasonable explanation as to why. Honestly it took months to figure out what that idiot had done wrong. It was hidden and made no sense. GD Link to comment Share on other sites More sharing options...
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now