ECMlink Switch from closed to open loop causing misfire

[OldGregg]

Hey guys, I’ve got an odd misfire P0304 that only happens on the switch from closed to open loop at idle after getting to operating temp. I’ve already ran through the list of normal misfire troubleshooting- check plugs, swap plugs, check wires, swap wires, replace coils, swap injectors, etc.. could the openloopmax table be out of wack and causing a misfire? I’ve read on ecmlink wiki that you can set the threshold longer to allow more time for the “learning” to happen in closed loop that will transfer over to the LTFT, is that correct and does anyone have any experience with this? Any help or ideas would be greatly appreciated. Thanks

 

[We're on Boost]

Switching from closed loop to open loop at idle sounds backwards.
If you are talking about how it goes after a cold start, it will start in open loop, then at some point it will switch to closed loop as it is warming up.
Once it is in closed loop, the log should show a CombinedFT (Combined Fuel Trim) which is the total fuel trim at the moment, combining short and long term trim.
Anyway, if you post a log of a cold start showing the cranking and warmup to 180 degrees or so (more is fine) we could take a look at the fuel trims and some other things. If the combined fuel trim is too far from 0 it could be the problem.
You post a log with the "Add Photos" button.

You must be logged in to view this image or video.

 

[steve]

As always, please post a log of the problem, and if you can't capture it as least post a log of your current config and a mods list.

 

[OldGregg]

Thank you for taking a look. Log is attached. The beginning of the log is making idle adjustments and calibrating AF gauge. The first misfire occurrence is at 545.9 sec when I had pulled out of driveway and up to second stop sign. When i turn the car off and turn back on, it goes away. You can tell when it happens as the CombinedFT will hit zero and it switches from Closed to Open Loop.

Current set up: 2.3l Stroker, shep transmission, ID1050x Injectors, E85 -pump and lines, GM 3" MAF, AEM Wideband- simulated, bastard 20g.

 

[Stapl3]

It's caused when you blip the throttle which sends the ecu into a different spot that it thinks it can go into closedloop. I think you have some wrong settings all adding up to this problem.

Your LoadScale is set for a 2.0 engine and must be changed for a 2.3.

If you are having to add ~20% airflow sliders, just zero out the silders and add the 20% via global fuel or base fuel pressure. You are using airflow sliders as fuel sliders and that is the wrong method. The less airflow adjustments you do, the better you are. These ECUs are so reliant on correct airflow settings and ironically that's what everyone messes with first and does so rather aggressively.

TPS is not calibrated correctly and the idle switch should be checked to be simulated.

 

[We're on Boost]

The first misfire occurrence is at 545.9 sec when I had pulled out of driveway and up to second stop sign.

I don't see any of these driving events in the log. Did you post the wrong log? All I see is 0 mph all the way through with it idling at about 1100 rpm and 0% throttle. There is one little throttle blip at 1530 which lasts for less than 1 second. During that blip your MAFRaw gets up to about 830 Hz. The rest of the time MAFRaw is down around 550 Hz. So the entire range of MAFRaw in the whole log is only from about 500 to 830 Hz. That's not enough to get much information from a MAFCompAdjust (CombinedFT) which is why they talk in the wiki about what kind of test drive to take in the car, getting a wide range of LoadFactor numbers (up to a little over 1) and along with that they talk about temporarily extending the LoadFactor range where your car will stay in closed loop. Because MAFCompAdjust will only give you suggestions that are valid for the closed loop region.
That wiki is here.
That wiki has a good example setup using a GM 3" MAF, 850cc injectors, and E85 fuel.
You have told the application that your injectors are 698cc so I guess you used ECMtuning's formula for using a fake injector size in the fuel tab calculator when your fuel is E85. Same thing they did in the wiki.

I’ve read on ecmlink wiki that you can set the threshold longer to allow more time for the “learning” to happen in closed loop that will transfer over to the LTFT, is that correct and does anyone have any experience with this?

They talk about extending the operating range of closed loop, like I mentioned above.
In Step 4 in the wiki they say that when you make a large change to a fuel setting, you might want to then speed up the re-learning of LTFT by deleting the previous LTFT. "So if you're making fairly large changes to MAF compensation, you should consider resetting the long term fuel trim values using the “Reset fuel trims” menu item in the ECU menu." (ECU, Reset fuel trims).

As far as the misfire - this log doesn't show any DTC's at all, and I don't see anything that looks like a misfire in the log.
Your InjOn time goes to 0 for half a second right after that throttle blip at 1531 but that is ok, that is just because after the throttle blip, while your rpm's are elevated but your foot is off the gas, the fuel will cut out until rpms get back down to about 1500, then fuel comes back on, which is what it did.
BTW you should log RawThrotPos to see the low end of your throttle range. Your Raw throttle pos gets down to 20. When you simulate idle switch from TPS like you are doing, raw throt pos values of 34 or less tell the ecu that you want it to idle (simulates a closed idle switch). Usually we adjust to get that raw value to 32 or 33 when foot is off the gas, when simulating isle switch. You do have "enable idle switch" turned on, so that is good.

You should take if for a drive with actual non-zero mph and a range of LoadFactors, rpms, throttle%, and so-forth. No heavy throttle of course. Just granny driving. See what the fuel trims do. In this log the Combined fuel trims are very close to 0 which is good. But you are really only showing idle so far.

Oh yeah, I assume you are using the GM MAF cable that ECMtuning used to sell? You should mention that in your configuration, because hardly anybody has one of those cables. Usually what we see in here is people with an old MAF Translator.

It does look like you used MafComAdjust to set you Maf Comp values, because the suggested values you get from running it in this log coincide with your table exactly. But the only valid values are the ones between like 500 and 800 Hz, because that's the only data there is in the log.

 

[OldGregg]

It's caused when you blip the throttle which sends the ecu into a different spot that it thinks it can go into closedloop. I think you have some wrong settings all adding up to this problem.

Your LoadScale is set for a 2.0 engine and must be changed for a 2.3.

If you are having to add ~20% airflow sliders, just zero out the silders and add the 20% via global fuel or base fuel pressure. You are using airflow sliders as fuel sliders and that is the wrong method. The less airflow adjustments you do, the better you are. These ECUs are so reliant on correct airflow settings and ironically that's what everyone messes with first and does so rather aggressively.

TPS is not calibrated correctly and the idle switch should be checked to be simulated.

Thank you! I have made these adjustments and redone the idle to reflect, still misfire after a short go around the block

I don't see any of these driving events in the log. Did you post the wrong log? All I see is 0 mph all the way through with it idling at about 1100 rpm and 0% throttle. There is one little throttle blip at 1530 which lasts for less than 1 second. During that blip your MAFRaw gets up to about 830 Hz. The rest of the time MAFRaw is down around 550 Hz. So the entire range of MAFRaw in the whole log is only from about 500 to 830 Hz. That's not enough to get much information from a MAFCompAdjust (CombinedFT) which is why they talk in the wiki about what kind of test drive to take in the car, getting a wide range of LoadFactor numbers (up to a little over 1) and along with that they talk about temporarily extending the LoadFactor range where your car will stay in closed loop. Because MAFCompAdjust will only give you suggestions that are valid for the closed loop region.
That wiki is here.
That wiki has a good example setup using a GM 3" MAF, 850cc injectors, and E85 fuel.
You have told the application that your injectors are 698cc so I guess you used ECMtuning's formula for using a fake injector size in the fuel tab calculator when your fuel is E85. Same thing they did in the wiki.


They talk about extending the operating range of closed loop, like I mentioned above.
In Step 4 in the wiki they say that when you make a large change to a fuel setting, you might want to then speed up the re-learning of LTFT by deleting the previous LTFT. "So if you're making fairly large changes to MAF compensation, you should consider resetting the long term fuel trim values using the “Reset fuel trims” menu item in the ECU menu." (ECU, Reset fuel trims).

As far as the misfire - this log doesn't show any DTC's at all, and I don't see anything that looks like a misfire in the log.
Your InjOn time goes to 0 for half a second right after that throttle blip at 1531 but that is ok, that is just because after the throttle blip, while your rpm's are elevated but your foot is off the gas, the fuel will cut out until rpms get back down to about 1500, then fuel comes back on, which is what it did.
BTW you should log RawThrotPos to see the low end of your throttle range. Your Raw throttle pos gets down to 20. When you simulate idle switch from TPS like you are doing, raw throt pos values of 34 or less tell the ecu that you want it to idle (simulates a closed idle switch). Usually we adjust to get that raw value to 32 or 33 when foot is off the gas, when simulating isle switch. You do have "enable idle switch" turned on, so that is good.

You should take if for a drive with actual non-zero mph and a range of LoadFactors, rpms, throttle%, and so-forth. No heavy throttle of course. Just granny driving. See what the fuel trims do. In this log the Combined fuel trims are very close to 0 which is good. But you are really only showing idle so far.

Oh yeah, I assume you are using the GM MAF cable that ECMtuning used to sell? You should mention that in your configuration, because hardly anybody has one of those cables. Usually what we see in here is people with an old MAF Translator.

It does look like you used MafComAdjust to set you Maf Comp values, because the suggested values you get from running it in this log coincide with your table exactly. But the only valid values are the ones between like 500 and 800 Hz, because that's the only data there is in the log.

Let me know if the attached represents better. I just came back from going around the block and it triggered the misfire again after sitting there for a moment- misfire occurs at 543.65 sec, ( threw misfire code on car but didn't instantly show in DTC on ecmlink)

 

[We're on Boost]

Let me know if the attached represents better. I just came back from going around the block and it triggered the misfire again after sitting there for a moment- misfire occurs at 543.65 sec, ( threw misfire code on car but didn't instantly show in DTC on ecmlink)

Better. Go less granny next time!
Only looked for a minute but I see you changed load scale to 87 which is good. (2.3 liter)
I see the range of MAFRaw going up to 1100 and that's about it. LoadFactor highest was about 0.5, MAF highest about 160 Hz.
So for reference, your normal closed loop range will usually go up to a load factor of about 0.9 and I think MAF there will be around 500 Hz, not sure about that. That's without "extending" closed loop at all from the normal values.
Looks like your RawThrotPos at idle is now 32 which is good for simulating idle switch.

Oh I see, at 544, right after the misfire, the LinWideband starts to go lean and it stays lean for the rest of the log. So there is definitely something going on there. Looks like a misfire alright.

 

[OldGregg]

Oddly enough, I had the “non 95/96 CAS inversion” selected under misc tab, when I unselected it (had the wrong setting) the misfire moved from cyl 4 to cyl 2.

Better. Go less granny next time!
Only looked for a minute but I see you changed load scale to 87 which is good. (2.3 liter)
I see the range of MAFRaw going up to 1100 and that's about it. LoadFactor highest was about 0.5, MAF highest about 160 Hz.
So for reference, your normal closed loop range will usually go up to a load factor of about 0.9 and I think MAF there will be around 500 Hz, not sure about that. That's without "extending" closed loop at all from the normal values.
Looks like your RawThrotPos at idle is now 32 which is good for simulating idle switch.

Oh I see, at 544, right after the misfire, the LinWideband starts to go lean and it stays lean for the rest of the log. So there is definitely something going on there. Looks like a misfire alright.

I reset the fuel trims and took it for another drive that was less granny , even got into some knock under low boost. Obviously i've got a bit to adjust in timing and haven't extended closed loop yet but this is what ive got so far. Important to note, after this quick drive, pulled into parking lot and after about a minute the misfire code threw again when it switched into open loop from closed loop at idle. Seems to only happen after the car is warmed up.. Vacuum leak? coolant temp sensor? CAS? A/F super off?

 

[We're on Boost]

Oddly enough, I had the “non 95/96 CAS inversion” selected under misc tab, when I unselected it (had the wrong setting) the misfire moved from cyl 4 to cyl 2.

Well I guess that tells us that it isn't the plug wires, plugs, or injectors. Beyond that, it must point to something but I don't know what. Somebody in here I think will know what things it must be, given that info. And the fact that it only happens after the car is warmed up.

When my power transistor went bad it would work for several minutes after a cold start, then I'd start to get a heavy misfire for a few seconds, then the engine would completely quit running. After letting it sit for about 5 minutes to cool off, it would restart and run fine again for a while. I ran and logged the car a few more times before replacing the power transistor. Never got a DTC in any of those logs. So I don't know, maybe the power transistor is one possibility.

I can help you find a decent way to buy a new one if you want. My experience doing that was with my 1g though. What year is your car?

 

[We're on Boost]

Next time you log you should turn on Edit, Captured Values, for Battery and ISCPosition.
On the Misc tab, you should probably turn on "Lock comm in ECMlink mode". Although it probably won't make any difference. Read the ? help about what it does.
Are you using the GM MAF cable that ECMtuning used to sell?
I don't see any knock in the log (KnockRet). Where is it?
I see the misfire though - LinWideband going lean and cycling wonky from 544 to the end.
I don't see any boost either. Highest LoadFactor I see in the log is 0.50 which is at 489 sec.
It did eke lean a little bit right there, right at 488.691 the LinWideband was 16.3 which is lean, so it's likely that as your throttle goes more open from that level you might be going lean.
Looks like you don't have a MAP sensor to measure boost. When I was using a GM MAF I had a lovely AEM brass 3.5 bar MAP sensor Teed to the vacuum hose that goes to the fuel pressure regulator. I still have it actually, sitting in a box in the house.
Do you have a boost gauge?

Other than right at 488.691 your AFRs look pretty much perfect. In the 14's just like they should be in closed loop.
Your timing numbers would be too high for gasoline, but on E85 they are reasonable. You are on E85 right?

A/F super off?

No. Not in this log anyway.

coolant temp sensor?

No.

 

[OldGregg]

Well I guess that tells us that it isn't the plug wires, plugs, or injectors. Beyond that, it must point to something but I don't know what. Somebody in here I think will know what things it must be, given that info. And the fact that it only happens after the car is warmed up.

When my power transistor went bad it would work for several minutes after a cold start, then I'd start to get a heavy misfire for a few seconds, then the engine would completely quit running. After letting it sit for about 5 minutes to cool off, it would restart and run fine again for a while.
I ran and logged the car a few more times before replacing the power transistor. Never got a DTC in any of those logs.
So I don't know, maybe the power transistor is one possibility. I can help you find a decent way to buy a new one if you want. My experience doing that was with my 1g though. What year is your car?

Thank you for your help. It’s a 97’ Talon TSi

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

Next time you log you should turn on Edit, Captured Values, for Battery and ISCPosition.
On the Misc tab, you should probably turn on "Lock comm in ECMlink mode". Although it probably won't make any difference. Read the ? help about what it does.
Are you using the GM MAF cable that ECMtuning used to sell?
I don't see any knock in the log (KnockRet). Where is it?
I see the misfire though - LinWideband going lean and cycling wonky from 544 to the end.
I don't see any boost either. Highest LoadFactor I see in the log is 0.50 which is at 489 sec.
It did eke lean a little bit right there, right at 488.691 the LinWideband was 16.3 which is lean, so it's likely that as your throttle goes more open from that level you might be going lean.
Looks like you don't have a MAP sensor to measure boost. When I was using a GM MAF I had a lovely AEM brass 3.5 bar MAP sensor Teed to the vacuum hose that goes to the fuel pressure regulator. I still have it actually, sitting in a box in the house.
Do you have a boost gauge?

Other than right at 488.691 your AFRs look pretty much perfect. In the 14's just like they should be in closed loop.
Your timing numbers would be too high for gasoline, but on E85 they are reasonable. You are on E85 right?


No. Not in this log anyway.


No.

Thanks for taking a look. Knock was at 1037.65 & 1184.2 sec, yes I do have the ecmlink gm maf cable, don’t have a 3-4 bar map sensor yet but do have AEM Af/boost gauges- looks like boost gauge is T’d into the FPR line, and yes running E85

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

 

[We're on Boost]

Thank you for your help. It’s a 97’ Talon TSi

If you want to try a new power transistor, the best place I know of to find a good brand of new one is Rock Auto.

Check this Rock Auto page out, showing a new power transistor for 1997 Tsi.

They call it an Ignition Control Module. There is only one offered, but it is a brand that should be ok (Wells) and it costs $95 which is only half what they cost for my 1990. Wells is a pretty good brand that was bought by NGK about 10 years ago.
Pick the "Info" button on it and look at that page. On that page they give you the OEM interchange part number as MD189747.

Then, for extra research, you can take that number, google "Mitsubishi parts online" and bring up 2 or 3 of the Mitsubishi dealers that you find that way. Paste in the MD189747 and see what they got. In this case, they show you the right part and they say it's correct part number for your '97 but it is Discontinued. . By doing this just now I found out that my usual Mitsubishi online parts dealer, which was "Mitsubishi Parts Now" is apparently not existing anymore, although the actual dealership (Capital City Mitsubishi) is still there. So I don't know what's going on with them but I checked a couple other Mitsu dealers and got the "Discontinued" from them.

I also checked JNZ and couldn't find it there which is surprising, maybe I was using their web site wrong.

Anyway, looks like it's the Wells part, from Rock Auto, if you want to try a new one. BTW the tests we have for the power transistor don't really work for this intermittent type failure (failure only when hot).

BTW the pics they show you in Rock Auto are usually the actual part, not "similar" or "representative".

 

[We're on Boost]

Thanks for taking a look. Knock was at 1037.65 & 1184.2 sec, yes I do have the ecmlink gm maf cable, don’t have a 3-4 bar map sensor yet but do have AEM Af/boost gauges- looks like boost gauge is T’d into the FPR line, and yes running E85

Ok.
The log you uploaded only goes up to 562 seconds though!

I'm not actually very optimistic about the power transistor fixing it. Because usually a problem with either the power transistor or its electrical connector would cause a misfire in 2 cylinders, not just one, and it would be a heavy misfire. It would be pretty hard just to keep the car running. Yours seems to keep running even with no added throttle, it just drops about 150 rpm in idle speed.
So hopefully someone will have a more likely suggestion.

 

[We're on Boost]

CAS?

I notice in the DTC list I posted up in post #2, that only the crankshaft position sensor is mentioned, not the camshaft position sensor. When you are talking about the CAS on a 2g, you are talking about both of those sensors. The factory manual definitely has lots of mentions of both sensors, so I think they would have mentioned it in this list if it was a possible cause.
Have you downloaded a copy of the 2g factory manual?
It is searchable, which is awesome! You can type in camshaft position or crankshaft position in the "Find text" blank in a pdf reader and read all you want.

On my own car, a 1g, they are both in the same gizmo and I've never had one fail. So I don't know much about them, especially on the 2g.
But I notice that good new ones are cheap for the 2g. And you can get NTK brand which is usually the best brand for this type of stuff. NTK is the same company as NGK. Standard and Wells should be good too, and maybe Walker.
Here's the Rock Auto page for these 2 things. Should come up showing both camshaft position sensor and crankshaft position sensor fields opened up for your car. Aw it comes up with only the crank sensor field open but the cam sensor field is right above it.

Here is the page in DSMtuners with a link to the 2g factory service manual (1996 actually). It's a link to the file on Tim Zimmer's google drive.
Just save it to your computer and use a pdf reader to search and read in it.

 

[OldGregg]

Ok.
The log you uploaded only goes up to 562 seconds though!

I'm not actually very optimistic about the power transistor fixing it. Because usually a problem with either the power transistor or its electrical connector would cause a misfire in 2 cylinders, not just one, and it would be a heavy misfire. It would be pretty hard just to keep the car running. Yours seems to keep running even with no added throttle, it just drops about 150 rpm in idle speed.
So hopefully someone will have a more likely suggestion.

So weird that it keeps uploading the wrong log. This should be the one-

 

[OldGregg]

If you want to try a new power transistor, the best place I know of to find a good brand of new one is Rock Auto.

Check this Rock Auto page out, showing a new power transistor for 1997 Tsi.

They call it an Ignition Control Module. There is only one offered, but it is a brand that should be ok (Wells) and it costs $95 which is only half what they cost for my 1990. Wells is a pretty good brand that was bought by NGK about 10 years ago.
Pick the "Info" button on it and look at that page. On that page they give you the OEM interchange part number as MD189747.

Then, for extra research, you can take that number, google "Mitsubishi parts online" and bring up 2 or 3 of the Mitsubishi dealers that you find that way. Paste in the MD189747 and see what they got. In this case, they show you the right part and they say it's correct part number for your '97 but it is Discontinued. . By doing this just now I found out that my usual Mitsubishi online parts dealer, which was "Mitsubishi Parts Now" is apparently not existing anymore, although the actual dealership (Capital City Mitsubishi) is still there. So I don't know what's going on with them but I checked a couple other Mitsu dealers and got the "Discontinued" from them.

I also checked JNZ and couldn't find it there which is surprising, maybe I was using their web site wrong.

Anyway, looks like it's the Wells part, from Rock Auto, if you want to try a new one. BTW the tests we have for the power transistor don't really work for this intermittent type failure (failure only when hot).

BTW the pics they show you in Rock Auto are usually the actual part, not "similar" or "representative".

Thank you. Im going to try a new/different PTU and see if it changes anything. I also got a tip from another forum to look at the 80A fuse?

 

[We're on Boost]

I also got a tip from another forum to look at the 80A fuse?

He's probably talking about the alternator fuse but I don't know why that would be it at all.

So weird that it keeps uploading the wrong log. This should be the one-

Ok, yeah I see some knock in that one.
3.2 degrees of knock retard at the place you mentioned, and 4.6 degrees retard at 723 seconds, and smaller amounts in a few other places. They are all during acceleration with 0.6 to 0.9 LoadFactor and around 2600 to 3200 rpm.

I don't think your timing numbers are high for E85. So I'm a little surprised about the knock.
But the easiest thing to do for a first shot at it is to reduce all the timing numbers in that area of the timing maps (both min and max octane maps) by 4 or 5 degrees. See if that pretty much gets rid of the knock. If it doesn't, then I think something else is the cause.
It would be this region that I greyed out in the pic below. And probably you will need to extend that region to the right as you begin to run it to higher RPM.
Also, since you are having some knock already at Load Factors below 1, you probably should not try extending your closed loop area. You might need it to go into open loop so it can run richer than stoic. In closed loop of course it will always try to give you stoic AFR.

You must be logged in to view this image or video.

 

[We're on Boost]

3.2 degrees of knock retard at the place you mentioned, and 4.6 degrees retard at 723 seconds, and smaller amounts in a few other places. They are all during acceleration with 0.6 to 0.9 LoadFactor and around 2600 to 3200 rpm.

At 1210 and 1234 you have about the same conditions but no knock. So it's a little strange. Maybe you are in the early stages of having bad HLA's (hydraulic lash adjusters). Still, try the reduced timing to see what happens.

 

[OldGregg]

He's probably talking about the alternator fuse but I don't know why that would be it at all.


Ok, yeah I see some knock in that one.
3.2 degrees of knock retard at the place you mentioned, and 4.6 degrees retard at 723 seconds, and smaller amounts in a few other places. They are all during acceleration with 0.6 to 0.9 LoadFactor and around 2600 to 3200 rpm.

I don't think your timing numbers are high for E85. So I'm a little surprised about the knock.
But the easiest thing to do for a first shot at it is to reduce all the timing numbers in that area of the timing maps (both min and max octane maps) by 4 or 5 degrees. See if that pretty much gets rid of the knock. If it doesn't, then I think something else is the cause.
It would be this region that I greyed out in the pic below. And probably you will need to extend that region to the right as you begin to run it to higher RPM.
Also, since you are having some knock already at Load Factors below 1, you probably should not try extending your closed loop area. You might need it to go into open loop so it can run richer than stoic. In closed loop of course it will always try to give you stoic AFR.

You must be logged in to view this image or video.

Thank you! I’ll start with these recommendations and see if it eliminates that knock. I have a new PTU in the mail and will definitely follow up once I get that part in.

 

[OldGregg]

Thank you! I’ll start with these recommendations and see if it eliminates that knock. I have a new PTU in the mail and will definitely follow up once I get that part in.

New PTU came in today, threw it in….. now it’s throwing a p0300 random misfire code. I’m at a loss, speedometer needle was all over during warm up- like bobbing up and down. I’m at a loss of words..

 

[We're on Boost]

New PTU came in today, threw it in….. now it’s throwing a p0300 random misfire code. I’m at a loss, speedometer needle was all over during warm up- like bobbing up and down. I’m at a loss of words..

Bummer.
Looking at what the FSM says about P0300, I think the most likely probable cause they have listed is the Crankshaft Position Sensor, which was already suggested by P0304 and is suggested again by P0300:

You must be logged in to view this image or video.

Here is a really crude diagram showing the general location of the crank sensor on the front end of the engine. I believe this is correct for the 1997, it's from the 1997 Electrical factory manual:

You must be logged in to view this image or video.

The speedometer jumping around sure seems weird, but there is no "speedometer cable". The speedometer is electrical all the way from a sensor on the transmission. There is a sensor there called the "Vehicle Speed Sensor" and I don't know anything about it, except that it is shown on a diagram as one of the sensors that feeds into control of the ignition system. So I wouldn't throw it out 100% as part of the problem. But I wouldn't mess with it either without getting some buy-in from a real 2g guy, which I am not!

BTW what brand and part number is the new PTU you got?

 

[We're on Boost]

now it’s throwing a p0300 random misfire code.

Found something you should try. Uncheck the checkbox called "Cylinder misfire".

You must be logged in to view this image or video.

Log it like that for a while and see if it changes anything.
This is something I ran across in a 2g thread. Had never heard of it (we don't have this on 1g) but it makes sense after everything I looked at in the FSMs yesterday.
The thread is here. Pay attention especially to post #2.
But I also thought it was interesting in post #1 where the OP says " The only way to fix it is to turn the ignition switch off and back on" and also mentions an ecu repair to replace a transistor used by cyl 1 and 4 ignition.
Anyway try turning that checkbox off, we might learn something about what your problem is from it.

 

[OldGregg]

Bummer.
Looking at what the FSM says about P0300, I think the most likely probable cause they have listed is the Crankshaft Position Sensor, which was already suggested by P0304 and is suggested again by P0300:

You must be logged in to view this image or video.

Here is a really crude diagram showing the general location of the crank sensor on the front end of the engine. I believe this is correct for the 1997, it's from the 1997 Electrical factory manual:

You must be logged in to view this image or video.

The speedometer jumping around sure seems weird, but there is no "speedometer cable". The speedometer is electrical all the way from a sensor on the transmission. There is a sensor there called the "Vehicle Speed Sensor" and I don't know anything about it, except that it is shown on a diagram as one of the sensors that feeds into control of the ignition system. So I wouldn't throw it out 100% as part of the problem. But I wouldn't mess with it either without getting some buy-in from a real 2g guy, which I am not!

BTW what brand and part number is the new PTU you got?

Thank you and apologies I meant to say the RPM needle was jumping not the speedometer. W're getting down to the last few options here and I know we're close (with the CPS being the last option since it's pretty deep in there). Since I didn't need the metal plate I ordered a HERKO HLX061 PTU replacement. Oddly enough when I spoke with my friend (Mechanic who sold me the car) he said that he switched out PTU's and it never changed anything. We're down to injector wiring and testing resistance from plug to ECU, or sending the ECU in for repair (which he also mentioned that it was sent in once before...)

 

[OldGregg]

Found something you should try. Uncheck the checkbox called "Cylinder misfire".

You must be logged in to view this image or video.

Log it like that for a while and see if it changes anything.
This is something I ran across in a 2g thread. Had never heard of it (we don't have this on 1g) but it makes sense after everything I looked at in the FSMs yesterday.
The thread is here. Pay attention especially to post #2.
But I also thought it was interesting in post #1 where the OP says " The only way to fix it is to turn the ignition switch off and back on" and also mentions an ecu repair to replace a transistor used by cyl 1 and 4 ignition.
Anyway try turning that checkbox off, we might learn something about what your problem is from it.

Thanks for this recommendation! I'm willing to try anything, so I'll give this a shot today and report back