legacycentral bbs

You could get a Link ecu... pretty much the same or more work though...

log1call wrote:You could swap the wires, and the TPI would then read the other way around, but what you need to know first is whether your ecu is expecting to see a rising or a falling voltage.

I don't see how you can swap the wires and reverse the TPS reading. For one, the TPS is just a variable resistor and resistors don't care what way current flows. If you can explain it differently, I'm all ears.

Secondly, you can't swap around the wires because the idle switch in the TPS utilizes the same ground as the main throttle portion.

Like I said, if you can explain it differently and how changing the wires would reverse the voltage and not interfere with the idle switch, I'm all ears.

Me too. I've read from two others on nasioc that say you can swap the wires to reverse it.

kimokalihi wrote:Unless there's a wiring issue I highly down its the idle switch. This is a brand new sensor and I've already had two other used ones and the behavior is the same albeit much worse with the new one and the last used one.

Kimo,

If you haven't properly adjusted the TPS, then it doesn't really matter whether it's new or past TPS's did the same thing. There is ONLY ONE adjustment that can be made with the TPS and that is the idle switch. If you follow the instructions I previously posted and adjust the idle switch, then you've properly setup the TPS.

I'm not saying it's going to 100% fix your problem, but if you have never adjusted the idle switch, that's what you need to do next before doing anything else!

kimokalihi wrote:Me too. I've read from two others on nasioc that say you can swap the wires to reverse it.

Again, I'm all ears to how exactly that works. Plus I'd be curious to know whether the NASIOC people are talking about the 3-wire or 4-wire TPS with the built in idle switch.

Here's some logs I did while driving today. Just normal acceleration and cruising and the last one should be a pull in 3rd gear.

https://skydrive.live.com/redir?resid=9 ... ZxkXk_rFB4

https://skydrive.live.com/redir?resid=9 ... shFX47gEuc

https://skydrive.live.com/redir?resid=9 ... Af3DOQmmR8

https://skydrive.live.com/redir?resid=9 ... mWtV41qk8I

I will check the idle switch hopefully tomorrow.

IF your using a 4-plug ECU (you are)

and

IF your using the EJ22T TPS (you are)

and

IF everything is still pinned correctly at the ECU and the TPS (I assume you checked)

then, your TPS voltage should not need to be swapped.

However, if you feel like you need to try then you can swap the "ends" of the resistor in the diagram. swapping 2 with 3 will reverse your TPS voltage from low to high. It will likely throw off your idle switch too.

EDIT: typoed the wire numbers on the first round.

If you want to reverse the OPERATION of that resistor, you need to swap 2 and 3.. the centre contact will start with it's contact at the reverse polarity then, and move in the right voltage direction... As you say though, the switch will then be a problem.. not unsurmountable...

By the sound of it though, you are certain that he has the right bits..

Kimo, in that last log, at line 14... that low reading from the TPI, that is meant to mean throttled off on your car right? Look at the air flow and IPW, they both go the wrong way. They would be good on my car, when the TPI reads a low voltage that's what I expect to see. On your car, it should be the opposite.

What's happened on that line is that you have floored it, the air flow has gone up, you are accelerating, the load is high, the IPW is high.. everything is as it should be, except for your car, the TPI should have just shot up, not down.. right?

Also, some of your readings are not working rihght.. but if that knock correction is right, you have a problem with that.. the knock correction, if it's working, should only show upp for a second and then it should have got rid of the knock, and the reading should go back to zero..

That might be a bad reading, or a symptom of the TPI... keep an eye on that.

If you want to get good easily readable data you really need to go on a deliberate road test, on a quiet bit of road where you will experience all driving conditions, so full throttle right through the rev range, some idling and gentle but prolonged acceleration to see if O2 feedback is working, ign and knock all show up for at least a few readings in a row.. A nice bit of hill country road is good, otherwise, open empty road and use the brakes to put a load on for a while at a time...

That other skydrive is my old hotmail one and I can't access that any more, but here is a log called Burma Rd.. a local windy hill... it's of a 90 2 litre auto...

https://skydrive.live.com/?cid=4ca3c345 ... 51DD%21106

See how the times on or off or steady on the throttle are a bit longer than yours... it makes it easier to read... see how there is a lag (because of the logging rate), between throttling off say, and the knock or timing adjusting... That lag is always there and you need steady readings to make it less noticable..

OK, so do some normal driving keeping throttle steady and then some slow accelerating and brake with throttle to create load? Are you suggesting that last part to keep the speed down for safety?

Yes, and to sustain the load for a while without needing a long stretch of empty roau.. I live in the country so it's easy for me. I've worked in the middle of cities though and you have to get a bit tricky..hold up the traffic to get a bit of room ahead, and then drone off in second with the brakes on....haha and a stream of cars behind you..
Did you see that your TPi was working like mine, when it shouldn't be, and that the air and injector readings behaved like mine though, where your tpi was in opposite result mode...

Another tip.. when you open the spreadsheet, click on the second row in the left border/number column, so the second row gets highlighted, then go up the top and on the window tab click the freeze pane ... then when you scroll the names at the top stay in place...

Oh, to clarify that earlier post about your log...
the TPi is controlling the engine in the right way. Despite your local models working in the opposite way, that one works correctly for that ecu..

Thanks. I'm new to excel and as soon as I started using it I wondered if there was a way to keep the names of the categories on screen when scrolling down and if there's a way to view it as a line graph. My evoscan gives me an error when I try to view the logs as a graph.

Marnix over on NASIOC says: Only 3-pin TPS sensor (MY97-98) has the opposite values.

Looked in my FSM and the idle switch test was not part of the TPS tests I did before. It's a seperate code (42). The gap between the idle/stopper screw and the linkage needs to be 0.7mm with 0.? ohm resistance and 0.9mm gap needs to be at least 1 M ohm. Both my throttle body's with seperate TPS, the old one (dirty one) had gaps closer to 1/4" which is 7 to 8 times the gap it's supposed to have when the idle switch it disengaged. This was causing the crappy throttle response and engine cutting out at very light throttle during cruising.

What doesn't make sense to me is why was it only throwing the TPS 31 code? Why not the Idle Switch 42 Code? And why the hell doesn't it tell you that you need to adjust that gap in the TPS tests?? The reason the gap was so large was because I had adjusted the TPS to get it closest to the fully closed spec of 4.4-4.8v which meant adjusting it all the way in one direction creating a massive gap.

This is where I need to give credit to Josh. Seems he was right all along. I was mislead with the misinformation and confusion of NASIOC. Those guys never said there was two different JDM GC TPS's. They just kept referencing the 97-98 sensor. I have one of those and it's got the wrong connector and it's totally different. That's why I had to use my EJ22T throttle body and TPS. Anyways, thanks Josh. Hopefully this problem is behind me forever.

I adjusted it to spec and took it for a drive and the throttle responds instantly now instead of lagging and on a couple short drives it appears to be fixed. No more stalling at cruising speeds (50-55mph). The code 31 was still in the ECU so afterwards I disconnected the battery and pumped the brakes. Hooked it back up and the code is gone, even after another short drive it did not come back. I wonder why it stayed in the ECU so long though. I drove probably 15 miles doing some datalogging on evoscan before coming back home and finding it was still coming up when I checked it.

Dirty old throttle body with Idle Switch gap.






Current EJ22T throttle body with brand new OEM Subaru TPS gap before adjusting Idle Switch.








Here's my test setup and a couple pics of the feeler gauge to show you how small of a gap is needs to be.

Here's the link to the new logs from today. Did some country road cruising.

https://skydrive.live.com/?cid=9bbcb90a ... sjAiJQuM_8

kimokalihi wrote:... I was mislead with the misinformation and confusion of NASIOC. Those guys never said there was two different JDM GC TPS's. They just kept referencing the 97-98 sensor. .......

Yeah, to anybody who doesn't know your set up you need to just tell them it's a '92-'96 V1 or V2 EJ20G for electrical problems. It's only your mechanical hardware that is like the '97-'98 EJ20K. Many on NastiCok will even doubt you got your later engine running on earlier electronics.

Glad to hear it's fixed! Now get her back on the road!

Its not insured. Waiting til july. I'm dealing with an insurance claim on my new brakes with the post office. They broke the box in shipping and damaged the paint job and they have my brakes and they're taking weeks to file the claim.

kimokalihi wrote:What doesn't make sense to me is why was it only throwing the TPS 31 code? Why not the Idle Switch 42 Code? And why the hell doesn't it tell you that you need to adjust that gap in the TPS tests?? The reason the gap was so large was because I had adjusted the TPS to get it closest to the fully closed spec of 4.4-4.8v which meant adjusting it all the way in one direction creating a massive gap.

It is in the EJ20G troubleshooting section for code 31. See below. I think the reason why it doesn't show up as a code 42 is because the ECU did see the idle switch changing (meaning the switch is functional) but it was happing at a point above some MAF limit check. That would tell the ECU that car was certainly not at idle even if the idle switch had not tripped and thus the fault that the TPS sensor was not adjusted properly.

Here is troubleshooting section for the EJ20G ecu;

Quite different than my 91 legacy FSM procedure. It doesn't say anything about removing the ECU or checking the idle switch at the ECU end of the harness. It is also adjusting it by voltage rather than by resistance.

Likely the difference between the Hitachi and JECS procedures. They both end up with the same results, but different ways to get there.

kimokalihi wrote:Quite different than my 91 legacy FSM procedure. It doesn't say anything about removing the ECU or checking the idle switch at the ECU end of the harness. It is also adjusting it by voltage rather than by resistance.

Well you are running an EJ20G ECU now, so you need to follow that manual

PhyrraM wrote:Likely the difference between the Hitachi and JECS procedures. They both end up with the same results, but different ways to get there.

You are correct it is likely the difference between the two ECUs. The JECS is a better ECU all around (one of the reasons I stopped working on the 22t ecu tune), but I don't think it would be the same result between the two. If the legacy FSM doesn't say to check for proper switch operation for code 31, then I would guess the ej22t ECU would not flag this situation as an error 31. It is a subtle difference, but I think it was this difference that let Kimo chase his tail for so long.

-Rob