1991 Eagle Talon Tsi AWD

[We're on Boost]

I'm starting to think this isn't a GM original MAF

Last night I took a bunch of pics of my GM MAF that I bought new from MAPerformance in 2009 and it is supposed to be a genuine GM MAF.
These pics might not do you any good, because when I look online, I see that some of the cheapo fake "compatible" MAFs look pretty darn much like mine. But here they are anyway

I would say this MAF worked right on "nominal" (accurate). My AFRs in the high power part of the range were 10.5 with zero tweaks to it in the ECU+. So close I didn't bother to tweak that part of it. I did tweak the idle range a little to make that part richer.
The "target" AFR that Jeff Oberholtzer had in mind when he burned the chip for the ecu was 11 at high power. Close. I had the ECU+ translating to 2g Hz, because Jeff said the 2g scale would be more suitable for high power. So that's what we did, that's the scaling that he put into his DSMchips ecu for me.
That's the setup you see in the Specs and Mods under that other profile of mine, the one that's called "up through 2015".

Switching back to a stock 1g MAF would have the advantage of, you kind of know where you're at with it, so it should be easier to get the rest of your system back to nominal operation. Assuming that the old 1g MAF is working right. It'll become your limiting factor once you hit about 2000 Hz, which is not hard to do once you have a 3 inch downpipe, 3 inch exhaust, and no cat.
Switching to SD would make sense because that's what you are going to do anyway, probably.

In these pics I've only taken off the air cleaner. The short pipe that goes to the 14b is still attached to the outlet side of the MAF, and we are looking through that pipe in the last pic.

 

[Spleen8urLSX]

Thanks for that. Based on this, I don't expect the FPS settings would fix any of our problems. I suspect we have a weak fuel pump and/or a restricted fuel filter. We have replaced the fuel filter and I have ordered a new fuel pump to eliminate those potential problems.

Since that GM MAF cable isn't available, and we can't connect the GM MAF directly to the ECU, I think we are just going to go direct to SD.

Not a bad move since you plan to go SD anyway and seem to be on track with your use of ecmlink thus far. There are things I like about using a maf as well that it seems a lot of the dsm crowd has gotten away with but you can’t go wrong with speed density, takes away some of the headache involved with using any maf in general.

 

[Spleen8urLSX]

Last night I took a bunch of pics of my GM MAF that I bought new from MAPerformance in 2009 and it is supposed to be a genuine GM MAF.
These pics might not do you any good, because when I look online, I see that some of the cheapo fake "compatible" MAFs look pretty darn much like mine. But here they are anyway

I would say this MAF worked right on "nominal" (accurate). My AFRs in the high power part of the range were 10.5 with zero tweaks to it in the ECU+. So close I didn't bother to tweak that part of it. I did tweak the idle range a little to make that part richer.
The "target" AFR that Jeff Oberholtzer had in mind when he burned the chip for the ecu was 11 at high power. Close. I had the ECU+ translating to 2g Hz, because Jeff said the 2g scale would be more suitable for high power. So that's what we did, that's the scaling that he put into his DSMchips ecu for me.
That's the setup you see in the Specs and Mods under that other profile of mine, the one that's called "up through 2015".

Switching back to a stock 1g MAF would have the advantage of, you kind of know where you're at with it, so it should be easier to get the rest of your system back to nominal operation. Assuming that the old 1g MAF is working right. It'll become your limiting factor once you hit about 2000 Hz, which is not hard to do once you have a 3 inch downpipe, 3 inch exhaust, and no cat.
Switching to SD would make sense because that's what you are going to do anyway, probably.

In these pics I've only taken off the air cleaner. The short pipe that goes to the 14b is still attached to the outlet side of the MAF, and we are looking through that pipe in the last pic.

View attachment 720401 View attachment 720402 View attachment 720403 View attachment 720404 View attachment 720405

Hey now, knocking out a few honeycombs and backing out the calibration screw goes a long way with the good ol 1g maf, I know you’ve been in the game long enough to remember that trick ! But all joking aside and while that does work there’s absolutely no reason to do that in the year 2024 when working 2g maf’s are 20 bucks and speed density is obtainable for under 100 bucks given you have the engine management for it. Not to thread jack but I like those trips down memory lane…and I shouldn’t talk as I was doing that in 2017 when 2g maf’s had hit that disposable price.

 

[justacanuck]

Last night I took a bunch of pics of my GM MAF that I bought new from MAPerformance in 2009 and it is supposed to be a genuine GM MAF.
These pics might not do you any good, because when I look online, I see that some of the cheapo fake "compatible" MAFs look pretty darn much like mine. But here they are anyway

I would say this MAF worked right on "nominal" (accurate). My AFRs in the high power part of the range were 10.5 with zero tweaks to it in the ECU+. So close I didn't bother to tweak that part of it. I did tweak the idle range a little to make that part richer.
The "target" AFR that Jeff Oberholtzer had in mind when he burned the chip for the ecu was 11 at high power. Close. I had the ECU+ translating to 2g Hz, because Jeff said the 2g scale would be more suitable for high power. So that's what we did, that's the scaling that he put into his DSMchips ecu for me.
That's the setup you see in the Specs and Mods under that other profile of mine, the one that's called "up through 2015".

Switching back to a stock 1g MAF would have the advantage of, you kind of know where you're at with it, so it should be easier to get the rest of your system back to nominal operation. Assuming that the old 1g MAF is working right. It'll become your limiting factor once you hit about 2000 Hz, which is not hard to do once you have a 3 inch downpipe, 3 inch exhaust, and no cat.
Switching to SD would make sense because that's what you are going to do anyway, probably.

In these pics I've only taken off the air cleaner. The short pipe that goes to the 14b is still attached to the outlet side of the MAF, and we are looking through that pipe in the last pic.

View attachment 720401 View attachment 720402 View attachment 720403 View attachment 720404 View attachment 720405

Yeah, that looks almost identical to mine, except ours has the honeycomb screen rather than the standard mesh screen shown in your pictures. From what I have read, that may be the only way to differentiate an OEM GM MAF from the aftermarket clone.

I just heard back from ECMLink, they confirm you cannot run the GM MAF directly to their ECU, at least not without the special cable they don't even sell any more. Very confusing...

We do have 3" straight-through exhaust, 3" intercooler piping, and an upgraded intercooler, apparently from a Supra. This is kind of why I didn't want to put the factory MAF back on it.

I hope SD is easy to setup and tune. I'll probably try wiring it up temporarily this weekend. I would rather not wait for the custom cable, althought now I'm not sure: Do we need that custom cable to wire in the MAP and GM IAT? or, will it work if I wire it manually? (i.e. does this custom cable have embedded electronics like the GM-MAF cable they used to sell?)

 

[We're on Boost]

Hey now, knocking out a few honeycombs and backing out the calibration screw goes a long way with the good ol 1g maf, I know you’ve been in the game long enough to remember that trick ! But all joking aside and while that does work there’s absolutely no reason to do that in the year 2024

Yeah those hacks were already old stuff when I got interested in these cars and started reading about them. I decided I didn't want to do any of those hacks. So my old 1g MAF is still original to this day, unhacked, and it was working "perfectly" when I last logged it. It's not even particularly dirty. It's been living comfortably in a nice cardboard box ever since!
I think that someday it should go to a person who wants to maintain a 1g that they want to keep in original form, not modified.

 

[Spleen8urLSX]

If you WERE to opt to toss a stock maf back on it for sake of working out the bugs it’s not going to care there’s a 3 inch exhaust or Supra sidemount on it, what the maf cares about is if more air is moving through it than what it can read (that 2000hz cap that wereonboost made mention of). With a 3” exhaust, ported 2g manifold, ching chong tubular o2 housing, frontmount intercooler, 9:1 compression non turbo pistons, and evo 3 16g at 14psi with a stock ecu and stock 450cc injectors my untouched 1g maf didn’t max out and I wasn’t seeing fuel cut. Mind you this set up was just to get the car running and work the bugs out and datalogs and the wideband showed me things were safe, I will say any more than that I would have definitely been at the limit. I had a spare 1g maf and “hacked” it (removed the lower honeycomb and backed the calibration screw out, not something that will work right without a way to compensate for it) when I installed my SAFC and 560cc injectors and left it at that for the rest of the year as the hacked maf wasn’t an issue at 22psi with all mentioned above and made 284 at the wheels on a low reading Mustang dyno. The following year I switched to ecmlink and speed density amongst other changes.

My advice, if you’re planning on speed density just do it now and it’s one more thing to cross off the list! You’ve done your reading on ecmlink and I will tell you there’s tons of support out there for SD for ecmlink users to where I don’t think you’ll have a problem at all, don’t let it be intimidating.

 

[We're on Boost]

I just heard back from ECMLink, they confirm you cannot run the GM MAF directly to their ECU, at least not without the special cable they don't even sell any more

Right, absolutely.
Apparently they couldn't get the translation thing done entirely within their chip. They had to put some electronic components outboard of the ecu. So they put them in the cable.
They also found that the ground wire coming out of the GM MAF should just go directly to any good chassis ground point in the engine bay. Which I did in my own wiring of the MAF just because it was simple. In their later version of the cable they had a chassis ground wire as part of it.

Have you seen this wiki Basic MAF Data ? This has some good info about the GM MAF and the various Mitsu MAFs, including frequency limits and mass limits, and they even give GM part numbers for the GM MAFs. I don't think GM put those part numbers right on the MAF anywhere unfortunately.

We do have 3" straight-through exhaust, 3" intercooler piping, and an upgraded intercooler, apparently from a Supra. This is kind of why I didn't want to put the factory MAF back on it.

Good. Do you have a 3 inch downpipe though? The downpipe is so important. If you still have the stock downpipe everything else is almost a complete waste. The first few inches of it could be 2.5 inches but the rest of it should be 3 inches. It would be good to put your exhaust info on your Specs and Mods page!

honeycomb screen rather than the standard mesh screen shown in your pictures. From what I have read, that may be the only way to differentiate an OEM GM MAF from the aftermarket clone.

I'm not even sure this is true anymore. I think the whole thing has become very confusing now to the point where if you had to start out with a new GM MAF and a new Translator, you'd have trouble doing the research to make sure that the MAF is for real and that the Translator was even correct for that particular MAF, the way it is scaled. You'd probably end up relying on info from some dealer who's been in the GM MAF Translator business for a long time like FullThrottleSpeed or R'Venge Performance. Maybe MAPerformance would have the right info.

 

[justacanuck]

My advice, if you’re planning on speed density just do it now and it’s one more thing to cross off the list! You’ve done your reading on ecmlink and I will tell you there’s tons of support out there for SD for ecmlink users to where I don’t think you’ll have a problem at all, don’t let it be intimidating.

Thank you for the encouragement.

Do we need a special cable for SD, or can I just wire the MAP and IAT directly into the ECU? We've got the stuff on the bench right now to solder it up.

 

[justacanuck]

Good. Do you have a 3 inch downpipe though? The downpipe is so important. If you still have the stock downpipe everything else is almost a complete waste. The first few inches of it could be 2.5 inches but the rest of it should be 3 inches. It would be good to put your exhaust info on your Specs and Mods page!

Yes, it is a 3" downpipe from the Turbo, all the way to the back, with no CAT, and no real muffler even.

 

[Spleen8urLSX]

Thank you for the encouragement.

Do we need a special cable for SD, or can I just wire the MAP and IAT directly into the ECU? We've got the stuff on the bench right now to solder it up.

You can indeed do your own wiring for speed density! There are write ups on here that explain it, there’s no built in resistors or anything like the cable made for the GM maf. Don’t get me wrong, ECMtuning’s plug and play cable is very nicely made and still in production, but not at all necessary. Speed density wiring is pretty much universal whether you’re on ECMlink or a standalone, little known secret Extreme PSI sells a speed density cable that is very nicely made by Fullthrottle Tech (same company that made the maf translator) that’s a couple dollars cheaper than ECMtuning’s, or you can make your own pretty easily using some connectors found in the junk yard and the ones off your maf translator if you don’t care to cut it up can be utilized as well but they’re common GM connectors, the maf translator is a good place to get the male side that plugs into the cars existing maf pigtail again if you don’t mind to cut it up. You even could hack the maf pigtail off the cars harness and hard wire it directly for the IAT sensor and MAP sensor if you wanted but I like not cutting it and having the easy option to change it if I want.

 

[justacanuck]

You can indeed do your own wiring for speed density! There are write ups on here that explain it, there’s no built in resistors or anything like the cable made for the GM maf. Don’t get me wrong, ECMtuning’s plug and play cable is very nicely made and still in production, but not at all necessary. Speed density wiring is pretty much universal whether you’re on ECMlink or a standalone, little known secret Extreme PSI sells a speed density cable that is very nicely made by Fullthrottle Tech (same company that made the maf translator) that’s a couple dollars cheaper than ECMtuning’s, or you can make your own pretty easily using some connectors found in the junk yard and the ones off your maf translator if you don’t care to cut it up can be utilized as well but they’re common GM connectors, the maf translator is a good place to get the male side that plugs into the cars existing maf pigtail again if you don’t mind to cut it up. You even could hack the maf pigtail off the cars harness and hard wire it directly for the IAT sensor and MAP sensor if you wanted but I like not cutting it and having the easy option to change it if I want.

That is what I understood from the instructions

So, we just wired in the IAT and MAP, as per the instructions, and enabled the settings in the ECU, as shown here: https://www.ecmtuning.com/wiki/sdsetup

It started but was still running like garbage. The IAT is showing -40F, so I'm starting to think, maybe we fried the 5V output on the ECU when I had incorrectly wired the GM MAF to 5V instead of the 12V circuit?

Or maybe I missed a setting somewhere.... Log attached.

 

[justacanuck]

Because the IAT was reading -40F, we went to Direct Access tab and zeroed out the SDTempWeighting table.

Seems to be running pretty decent now. But IAT not working must be a problem with wiring or the ECU.

 

[Spleen8urLSX]

Did you use a known good IAT sensor? I had a parts store special go bad and log -40 degrees, tossed in a real AC Delco one off a car at pull a part my buddy pocketed and had in his tool box and that one never let me down.

 

[justacanuck]

Did you use a known good IAT sensor? I had a parts store special go bad and log -40 degrees, tossed in a real AC Delco one off a car at pull a part my buddy pocketed and had in his tool box and that one never let me down.

It should be good... it was brand new, purchased from ECMLink. I suppose it could be a bad sensor. We have an extra, we will plug it in to confirm.

 

[Spleen8urLSX]

Being it’s new I doubt it’s the sensor but if you have a spare worth testing. It is possible like you said attempting to wire in the GM maf directly cooked something in the ECU if the sensor and the wiring check out.

 

[justacanuck]

Confirmed the other sensor behaves the same. If the ECU needs to go back for repair again, that would suck.

But IAT only uses ECU sensor ground and IAT input signal... we traced the wires back to the ECU to make sure they are connected right and found a broken wire near the MAF connector. FIxing now.

 

[justacanuck]

IAT sensor is working!!

Now for the speed signal to ECU:
- Tested reed switch in speedometer = Works.
- Tested wire from Reed switch to ground = works
- Tested wire from Reed switch to ECU = works

Must be another bad 5v driver in the ECU which is not functioning as expected.

Going to start some SD tuning.

 

[justacanuck]

Our diagnostics on the speed sensor circuits somehow fixed the problem. We now have a speed signal to ECU.

 

[We're on Boost]

The IAT is showing -40F, so I'm starting to think, maybe we fried the 5V output

In the log you showed there I noticed that the raw voltage RawGMIAT was 4.71 volts.
That is what the IAT input gives you when there is either nothing connected to it or a dead thing (a thing with ~infinite resistance) connected to it. About 4.7 volts.
My ecu logs 4.69 volts in that condition. This voltage is always somewhere around 4.7 volts on the 1g ecu.
So it makes sense that you found a broken wire!
Very glad it wasn't the ecu.
In the future if you have doubts about that input (and some of the other inputs are similar) you can sort of test the input with a small battery, like a 1.5 volt AAA, preferably one that is mostly shot so it can't put out a lot of current. See if the raw voltage you get from it in ECMlink is close to the voltage you get by testing the battery with a multimeter.
Just make sure your test battery is less than 5 volts and I think some of the inputs you wouldn't want to test this way. The thermistor inputs I think are ok to test this way.

 

[justacanuck]

Car is running good. Did some minor adjustments in the SD table.

Did some WOT pulls in 3rd gear. What's next? (log attached)

 

[Spleen8urLSX]

I will say, now that all of that’s solved I have no issue getting off track, I’m envious of that mkiv Supra smic! Hard to get a hold of one these days.

 

[justacanuck]

Speed sensor is intermittant. Going to take the dash cluster out, and clean the contacts this time. I think that will finally fix the issue completely.

Currently, we are running conservative 2G timing, and 1G fuel maps right now, that seem pretty rich IMO.

So, next I think we will try swapping OpenLoopMaxOct table and timing tables with 2008 EVO data.

 

[justacanuck]

The previous logs were collected while using the stock 2G timing tables, and the stock 1G AFR/fuel map.

Thinking about those logs, it seemed to me that it was very likely too rich, and the 2G timing map we were using is apparently fairly conservative. So I decided to turn the boost controller down again, and switch to EVO8 Fuel and Timing map. We had previously bumped into 100% injector duty cycle, at just over 5K, so I also turned the rev limiter down to 5500 just to make sure I didn't push it too hard, at least until I could see what happens to the injector duty cycle with these new settings.

Now, the car seems to be running even stronger. This time, I don't think we got much beyond 80% Injector duty cycle. I do see a few degrees of knock in a couple spots, but nothing that seemed too serious. I'll work on pulling a couple degrees of timing out, in a few of those trouble spots, only where the timing does seem a little too high, before I try it again. If the timing adjustments fix those few areas where I saw the knock retard kick in, then I'll probably try a few more turns on the manual boost controller.

Turning up the boost controller does seem to make a fairly significant difference in how hard and how fast the boost hits, although I am really doubting that this turbo could actually put out much more boost that what we saw in the logs today. It does look like the factory turbo but I do believe it has an aftermarket exhaust housing, or possibly it comes from a different car entirely. It is entirely possible that this turbo has other modifications. IDK, but I am really surprised how fast the boost hits, and how much boost it makes.

It is relatively cold today (-2C or +30F) but I'm really thinking we have an issue with the thermostat. It must be stuck open because while we are driving, it will not stay above 160F, even with some cardboard covering a good portion of the radiator. I should also mention, the IAT isn't actually installed into the intake piping yet, I just have it hanging in the engine bay over the intake piping, so at least it should be giving us a good approximation of IAT based on the temperature under the hood. I am waiting on a steel bung to install it properly. In the current weather conditions, I would not expect intake temperatures to change much at these boost levels.

I did use today's log to perform some adjustments to the SD table, but I don't think that will show up in the attached log, because I didn't make the change until the end of this logging session.

I'm really liking this speed density setup but I don't really have any experience tuning, so I would welcome any feedback on our log (attached).

 

[We're on Boost]

Looks like success! Man you get more done in a day than I get done in a week
I see 31 lbs/min airflow and 23 - 24 psi of boost with AFR in the mid to high 10's and knock is 0 or at most 1.4 degrees of retard. Seems good!

Your wideband agrees well with the narrow band when you are in closed loop (stoic). I was a little worried about that because it is using the EGR Temp input which doesn't always work out so well on the 1g ecu (didn't on my wideband) but yours seems good. Your wideband must have a good robust analog output to the ecu.
I'm glad you have a stock type narrow band on it still for running the closed loop stuff.

Your closed loop fuel trims look good to me, they aren't all plus or all minus, and the numbers aren't big. Quite a lot of your short term trim is slightly minus which, if I understood Lucas English correctly, is how he likes to have it. He said he likes to have fuel trims slightly "pulling" which I took to mean negative, reducing fuel a little from what it would be if you were forcing open loop. He said "because it drives better that way".

 

[justacanuck]

Yeah, I think I am close but I also think I missed a couple of steps.

I could not figure out why AFR seems to jump, the STFT/CombinedFT would not always stay within target. The logs seem to indicate I am getting close to where we need to be, but it seems I am still running a little leaner than it should be, particularly when it switches to open loop.

The Fuel Trim video here http://www.ecmtuning.com/demos/fueltrim.mp4 , in the 5-6 minute timeframe, shows how to set the Injector dead time properly. It also goes into some detail showing how to extend the closed loop operation to collect more data to help auto-tune everything.

I'm going to have to go backward a little bit, and make sure I have the injector deadtime dialed in properly.