ECMlink Tuning Direct maps CoolantTempFuelAdj CrankingFuelAdj

[justacanuck]

I just started tuning with ECMlink, and we did have some difficulty but we did end up getting the car started when the temp was only 12F. However, now that the temp has dropped closer to -9F we can't get it to start. Thus far we have tried changing the CoolantTempFuelAdj and CrankingFuelAdj so the -2F field is set closer to the 19F field (i.e. the coolant temp when we DID have a successful start).

I am wondering if anyone would share some wisdom regarding the settings we should be looking at to help with VERY cold startup scenarios. (e.g. -9F outdoor temperatures)
- CollantTempFuelAdj
- CrankingFuelAdj
- InjBatterAdj

Thanks in advanced for any advice.
/Josh

 

[We're on Boost]

I just started tuning with ECMlink, and we did have some difficulty but we did end up getting the car started when the temp was only 12F. However, now that the temp has dropped closer to -9F we can't get it to start. Thus far we have tried changing the CoolantTempFuelAdj and CrankingFuelAdj so the -2F field is set closer to the 19F field (i.e. the coolant temp when we DID have a successful start).

Cranking Fuel Adjust is a great thing but you should verify that it is actually working.
To do that, make sure that in Edit, Displayed Values, you are displaying "Key Start" and "CrankingFuelAdj".

When you are looking at your logs then, you should see Key Start as a 1 when you are actually turning the key to "Start" and a 0 the rest of the time.

During cranking, your log for CrankingFuelAdj should display as whatever number you set it to in the table. Then as soon as the engine fires, you should see the logged number trickle downwards and go to zero in about 20 seconds, in cold weather.

I would think that 200% should be a good number for it at -2 deg F and as far as I know that is the default for it there. But my coldest Talon startup since having ECMlink was only 38 deg F. So I'm maybe not going to be much help here . That startup shows an Adjust of 95% during cranking. You would need much more than that for your temps.

My only experience with below zero F starts was way back in carburetor days when I lived in northern Wisconsin. There I used an engine coolant heater inserted into a heater hose.

 

[justacanuck]

Cranking Fuel Adjust is a great thing but you should verify that it is actually working.
To do that, make sure that in Edit, Displayed Values, you are displaying "Key Start" and "CrankingFuelAdj".

When you are looking at your logs then, you should see Key Start as a 1 when you are actually turning the key to "Start" and a 0 the rest of the time.

During cranking, your log for CrankingFuelAdj should display as whatever number you set it to in the table. Then as soon as the engine fires, you should see the logged number trickle downwards and go to zero in about 20 seconds, in cold weather.

Thanks for that. I was working that direction already. I added CrankingFuelAdj to the logs and I am not seeing any values displayed in the log. I didn't think to check before, but I am also not getting "Key Start" signal either.

I don't understand how that is possible, it was definitely cranking.

I'll come back to this thread if my experience, as documented in this thread, produces anything of value.

 

[We're on Boost]

Thanks for that. I was working that direction already. I added CrankingFuelAdj to the logs and I am not seeing any values displayed in the log. I didn't think to check before, but I am also not getting "Key Start" signal either.

I don't understand how that is possible... it was definitely cranking.

Aha, then I think we can get you fixed up just fine.
To be clear about your post, you are logging Key Start and it shows 0 all the time?

If that's what you have, then it's the same problem or similar, to what I had in 2022 and should have a similar solution.

Here is the thread for that:
https://www.dsmtuners.com/threads/my-crankingfueladj-seems-to-not-be-working-at-all.540792/

I had completely bypassed and removed my starter relay, which is a little 4 leg relay in the cabin under the dash and near the left end.
I have a lot of logs and in the older ones Key Start worked. At some point it quit working. So all I had to do was think about what mod I did to the car at that time. It was when I totally bypassed the starter relay and its socket.

It turned out that the upper left position of the starter relay socket needs to have 12 volts on it just only when you are cranking. So all I had to do was run 1 wire to that upper left position, from my starter button circuit that has 12 volts in it whenever I'm pushing my starter button.

Another way to get into that circuit with 12 volts is by finding the wire that goes to pin 108 on the 1g ECU. That was pointed out by jim95redgsx in post #2 in that thread. I haven't tried that but I think that is the same circuit that goes to the upper left starter relay leg that I went to. On the 1990-1994 DSM ECU Pin Out diagram from ECMtuning, pin 108 is labeled "Ignition switch (ST) input".

Take a look at post #6 in that 2022 thread. Picture shows the starter relay socket with my wire going into the upper left hole. That is correct. But the blade contact I was using there was bad because the little wings that stick out to either side kept it from inserting all the way, so I had to put a different blade on there for full insertion. Otherwise, good.

BTW at one time earlier than all this, I had the starter relay removed and I had the upper 2 ports of the relay socket jumpered to each other with a short jumper wire. That worked too as long as my ignition switch was good. Well I didn't have ECMlink then (when I was using the jumper) but my cold starts were good. Later it seemed like the "Start" position in my ignition switch went bad, so I bypassed the ignition switch and starter relay with new wiring coming from a starter button instead of from the ignition switch, and my problem happened about midway through the evolution of that system.

 

[justacanuck]

That sounds like our problem exactly!

I have seen the Keystart signal show up a few occasions, so it could well be a loose connection on the start relay. The KeyStart signal was only ever active for a very short time however, not the entire cranking cycle. The few times we did get it to start cold, we did see the keystart signal momentarily.

Is that little connection on the passenger side near the firewall/hood hinge? I'll go outside tomorrow and fix it.

 

[We're on Boost]

Is that little connection on the passenger side near the firewall/hood hinge? I'll go outside tomorrow and fix it.

Well the location I was working with was the starter relay, and that is on the driver side. You have to take the lower panel off to see it, I think (4 screws). The cover that is to the left of the steering column. Between the steering column and the door.
Here is the "best" diagram I have of where it is located. This is from the 1990 factory manual. Looking sort of underneath the dash with that lower cover taken off.

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

But if your start circuits are stock, the starter relay has to be connected and working to even get the starter to go. So I'm thinking your wiring connection problem might be in a different place, because your starter is going.

There is a wire somewhere that takes current from the outbound leg of the starter relay, to pin 108 on the ECU. That might be the wire that is bummed on your car. I've never seen that wire and I don't know where it is in the car. The "outbound" leg of the starter relay is that upper left position in the relay socket that I connected to, like I showed in that other thread. If you can see the wires coming out from the back of the relay socket, the wire that comes from the "outbound" leg should at some point branch into 2 branches. One branch would go into the engine bay to connect on the starter solenoid, the other branch would go to 108 on the ECU. I think.

I was hoping that you had deleted your starter relay like I did. Then your problem would be more exactly like mine was. But if your starter relay is still there, then it's basically the same problem but originating in a different spot, I think.

There is a schematic wiring diagram in the Chilton manual (page 6-52) that I think would be correct for your 1991 turbo.
I can scan it and post it later on Saturday. It shows what I'm saying about the outbound leg and the wire going from it to the "MFI system" which they mean the ECU (MFI control unit). That might be the wire that has your bad connection.

 

[1990TSIAWDTALON]

If you don't see "Key Start", your car will definitely be hard to start at any time. Mine would be super hard (20 attempts) when it was not working, even in the summer. I can't remember how I remedied it, but I think Steve was the one that pointed it out to me.

 

[justacanuck]

Well the location I was working with was the starter relay, and that is on the driver side. You have to take the lower panel off to see it, I think (4 screws). The cover that is to the left of the steering column. Between the steering column and the door.
Here is the "best" diagram I have of where it is located. This is from the 1990 factory manual. Looking sort of underneath the dash with that lower cover taken off.

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

But if your start circuits are stock, the starter relay has to be connected and working to even get the starter to go. So I'm thinking your wiring connection problem might be in a different place, because your starter is going.

There is a wire somewhere that takes current from the outbound leg of the starter relay, to pin 108 on the ecu. That might be the wire that is bummed on your car. I've never seen that wire and I don't know where it is in the car. The "outbound" leg of the starter relay is that upper left position in the relay socket that I connected to, like I showed in that other thread. If you can see the wires coming out from the back of the relay socket, the wire that comes from the "outbound" leg should at some point branch into 2 branches. One branch would go into the engine bay to connect on the starter solenoid, the other branch would go to 108 on the ecu. I think.

I was hoping that you had deleted your starter relay like I did. Then your problem would be more exactly like mine was. But if your starter relay is still there, then it's basically the same problem but originating in a different spot, I think.

There is a schematic wiring diagram in the Chilton manual (page 6-52) that I think would be correct for your 1991 turbo.
I can scan it and post it later on Saturday. It shows what I'm saying about the outbound leg and the wire going from it to the "MFI system" which they mean the ecu (MFI control unit). That might be the wire that has your bad connection.

We definately have a small mess of wires under the dash in that location. A bunch of wires go into a mysterious black box labeled "Genesis by Delta". It appears to have some relays for doors, trunks etc. Still trying to find the starter relay...

 

[justacanuck]

We definately have a small mess of wires under the dash in that location. A bunch of wires go into a mysterious black box labeled "Genesis by Delta". It appears to have some relays for doors, trunks etc. Still trying to find the starter relay...

so, I found the starter relay. It seems we are getting power from the BY wire on pin 3 going to the ECU pin 108 but still can not see KeyStart signal logged in ECMLink.

 

[justacanuck]

Plugged in a different ECU and we can now see "keystart" signal as well as the corresponding CrankFuelAdj. Looking like we have to send back this ECU for further repair.

Thanks for the help.

 

[We're on Boost]

Plugged in a different ECU and we can now see "keystart" signal as well as the corresponding CrankFuelAdj. Looking like we have to send back this ECU for further repair.

Aha. Good actually, in that it's a perfectly reasonable outcome, and apparently the wiring in the car is ok. Good that you had another ECU to try.
Good luck with it!

I googled "Genesis by Delta car alarm system" and yup, that's apparently what that is. That location is where the aftermarket installers usually put them in these cars.

 

[justacanuck]

I am still struggling as to why we couldn't get the "good" ECMLink ECU to start the car during really cold weather. It seemed to me that it may have been sending too much fuel but I can't be sure. I would really like to see someone's "really cold" startup logs so I can compare a known/good startup log and corresponding ECMLink settings.

It would also be nice if someone could post the details of their "CrankFuelAdj" and "CoolantTempFuelAdj" settings for cold weather startups. I would also benefit from seeing the logged InjOn times (ms) that are seen during very cold startup. I was seeing ~40ms InjOn time during our (unsuccessful) initial crank/cold startup. Doesn't that seem _really_ high? Even when I turned "CrankFuelAdj" and "CoolantTempFuelAdj" off, I was still seeing what I thought could be excessively high injector times during cold start/cranking so I suspect something else in the configuration may be adding a LOT of extra fuel. InjOn time, after startup with cold idle, is closer to 4-5ms and it goes down to ~2ms InjectorOn time during warm idle. I was really expecting to see something closer to ~10ms InjOn time during cold startup cranking.

FWIW, I also logged AirFlowPerRev close to 1gm/rev during initial cranking. We are currently running 560cc (Evo6-8?) injectors, and a MAF translator which I suspect could be sending some funky data that ECMLink might not be happy with.

 

[1990TSIAWDTALON]

Try or look at stock Evo settings. They should be close to your set up I would think.

 

[We're on Boost]

I am still struggling as to why we couldn't get the "good" ECMLink ECU to start the car during really cold weather.

Do you still have the FIAV on the car - or have you blocked it off like some people do?
On my car there is no FIAV and I run the ISC at 120 until it warms up to 95 deg F to compensate.
I guess you can step on the gas a little instead, but then your IdleSw will say 0 instead of 1, and I'm not sure if that messes anything up for cold starting.

I was seeing ~40ms InjOn time during our (unsuccessful) initial crank/cold startup. Doesn't that seem _really_ high?

40ms sounds high but then when I get out my calculator it doesn' t seem much different than my car.
Like this: My car uses about 14ms while cranking at ~220rpm with 40 deg F coolant. My injectors are ~1500cc actual (1650cc nominal).
1500/560 = 2.68 so my injectors are 2.68 times bigger than yours.
2.68 x 14ms = 37.5ms
Making it a little more messy is that my fuel in that log was 39% ethanol. Also my base fuel pressure is 43psi.
Anyway, maybe 40ms is ok for around 0 deg F cold starts. I wouldn't actually know!

InjOn time, after startup with cold idle, is closer to 4-5ms and it goes down to ~2ms InjectorOn time during warm idle.

This sounds about right.

I gather you are using a GM MAF going to a MAF Translator, then I guess that sends the Hz signal to the normal Hz input on the ecu. You are not using the ECMtuning MAF translation cable, or the translation features in ECMlink. You are doing all the translation in the MAFT which is converting to 1g MAF hz which then goes to the ecu. Right?
I used a GM MAF for several years and I liked it a lot. But I never had it in weather below freezing, and I was doing my MAF translation with an ECU+ Silver Box which did an awesome job of it.

I'm going to go ahead and attach my log that I just looked at for the 40 deg start so you can look at some stuff, for whatever good it might do. Some of my ECMlink settings are a little unusual but it works. It was not setup by me, it was setup by English Racing who had just built the engine, and the results have been real nice. Sorry I am not logging AFR, I'm still using an AEM wideband of the previous generation that didn't log well at all into the EGR Temp input and it destroyed the analog output part of the wideband. But the gauge still works fine and seems very accurate. BTW this log is the one I was taking when I shot my video "Watching for closed loop" which is on my dash-four YouTube Channel. So you could actually watch that at the same time as looking at the log.

 

[justacanuck]

Try or look at stock Evo settings. They should be close to your set up I would think.

Good idea. I found this, which may be helpful.

It does appear to show some smaller adjustments to InjBatteryAdj settings than what I saw in our settings. But still, I would not expect these tiny tweaks to injector deadtime based on battery voltage to make any significant changes to expected injectorOn time during startup/cranking.

Do you still have the FIAV on the car - or have you blocked it off like some people do?
On my car there is no FIAV and I run the ISC at 120 until it warms up to 95 deg F to compensate.
I guess you can step on the gas a little instead, but then your IdleSw will say 0 instead of 1, and I'm not sure if that messes anything up for cold starting.


40ms sounds high but then when I get out my calculator it doesn' t seem much different than my car.
Like this: My car uses about 14ms while cranking at ~220rpm with 40 deg F coolant. My injectors are ~1500cc actual (1650cc nominal).
1500/560 = 2.68 so my injectors are 2.68 times bigger than yours.
2.68 x 14ms = 37.5ms
Making it a little more messy is that my fuel in that log was 39% ethanol. Also my base fuel pressure is 43psi.
Anyway, maybe 40ms is ok for around 0 deg F cold starts. I wouldn't actually know!


This sounds about right.

I gather you are using a GM MAF going to a MAF Translator, then I guess that sends the Hz signal to the normal Hz input on the ecu. You are not using the ECMtuning MAF translation cable, or the translation features in ECMlink. You are doing all the translation in the MAFT which is converting to 1g MAF hz which then goes to the ecu. Right?
I used a GM MAF for several years and I liked it a lot. But I never had it in weather below freezing, and I was doing my MAF translation with an ECU+ Silver Box which did an awesome job of it.

I'm going to go ahead and attach my log that I just looked at for the 40 deg start so you can look at some stuff, for whatever good it might do. Some of my ECMlink settings are a little unusual but it works. It was not setup by me, it was setup by English Racing who had just built the engine, and the results have been real nice. Sorry I am not logging AFR, I'm still using an AEM wideband of the previous generation that didn't log well at all into the EGR Temp input and it destroyed the analog output part of the wideband. But the gauge still works fine and seems very accurate. BTW this log is the one I was taking when I shot my video "Watching for closed loop" which is on my dash-four YouTube Channel. So you could actually watch that at the same time as looking at the log.

Yes, we blocked on the FIAV. I'm kind of regretting that now .

Thanks for the insight into your Injector/log data. That gives me more confidence that something else is probably wrong with our setup.

Yes, we are using GM MAF through MAF Translator. After I found the instructions for the MAFT, I was able to better tweak the MAFT settings for our injector, using the stock ECU, it fired right up today at -14C.

Thanks for posting your logs. It will certainly help me understand ours

 

[1990TSIAWDTALON]

I don't run the FIAV but I run SD and it has different sensors. For some reason, mine start and run and KNOCK ON WOOD!!!

 

[We're on Boost]

Good idea. I found this, which may be helpful.

It does appear to show some smaller adjustments to InjBatteryAdj settings than what I saw in our settings. But still, I would not expect these tiny tweaks to injector deadtime based on battery voltage to make any significant changes to expected injectorOn time during startup/cranking.

Yeah.
My injBatteryAdj at 9 volts is a little extreme because I was fiddling with it when my CrankingFuelAdj wasn't working, and I never set it back to what ER had it at. I might set it back at some point. Here's what it was from ER:

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

That's 1.4ms difference at 9 volts which sounds like a lot, but during cold cranking it's pretty much nothing compared to what CrankingFuelAdj puts on it. It would be more relevant to what works if I'm trying to drive home with a dead alternator sometime. And I've had that happen.

Yes, we are using GM MAF through MAF Translator. After I found the instructions for the MAFT, I was able to better tweak the MAFT settings for our injector, using the stock ECU, it fired right up today at -14C.

That's awesome!
I think we almost have a song for that!
Cuz everytime we start, at 14 minus ...