Question on adjusting the BISS & idle

[XC92]

I've been a bit confused about how to properly adjust the BISS on my '92 Talon TSi. I realize that to do it properly you have to ground 2 connectors to disable the ECM's auto idle adjust and use a tachometer to adjust the BISS. What I'm confused about is whether adjusting the BISS without grounding these 2 connectors will "confuse" the ECM and result in poor idle control even if it seems to work for a while, and if you literally MUST ground them to properly set idle and can't do it otherwise no matter how finely you set the BISS?

The background for asking this is that I had both rough idle and poor fuel consumption so I recently removed and overhauled the entire intake and exhaust systems, cleaning everything up really well, replacing all gaskets and o-rings, vacuum hoses and even the throttle shaft seals, putting in a new O2 sensor as the old one wasn't working, having the injectors professionally cleaned, etc., yet I'm still experiencing rough idle and poor fuel consumption, which I'm guessing is due to either an improperly adjusted BISS or faulty IAC and/or FIAV. But I want to start out by first properly adjusting the BISS, and see if that fixes it, and if not only then moving on to other possible causes.

Btw, not asking anyone to diagnose my issues, but what tend to be the most common causes of rough idle and poor fuel consumption in this particular model at this point in its service life, over 31 years now? When I tested all the sensors either the fuel or EGR solenoid (I forgot which but I have it written down somewhere) gave out of range readings. Could it be this? I mean the ones attached to the firewall on the driver's side that are connected to the fuel regulator and EGR value via vacuum hoses and to the ECM via cables.

Btw I overhauled the gas tank a bit over a year ago and the pump and sending unit were fine, and at the same time I replaced the fuel filter with a Wix, and there are no fuel leak smells or indications.

 

[steve]

Rough? Does that mean missing or unstable?

I'm sure there are people out there that adjust their BISS based on logging ISC steps without grounding the connectors. I believe that's a normal adjustment by DSMLink users who aim for around 30 steps at warm idle. I've done it both ways and noted small differences in the resulting ISC steps but no sigificant difference in idle quality. In my case it's always at a 750 RPM target idle. I would expect higher idle settings (via changing the programming) might differ from that # of steps.

I can't say I've noticed any change in fuel consumption from this adjustment.

Are all your cylinders contributing equally and the ECU able to maintain closed loop at idle and cruise?

 

[XC92]

Rough? Does that mean missing or unstable?

I'm sure there are people out there that adjust their BISS based on logging ISC steps without grounding the connectors. I believe that's a normal adjustment by DSMLink users who aim for around 30 steps at warm idle. I've done it both ways and noted small differences in the resulting ISC steps but no sigificant difference in idle quality. In my case it's always at a 750 RPM target idle. I would expect higher idle settings (via changing the programming) might differ from that # of steps.

I can't say I've noticed any change in fuel consumption from this adjustment.

Are all your cylinders contributing equally and the ECU able to maintain closed loop at idle and cruise?

So adjusting the BISS manually without grounding the connectors won't give the ECM bad readings or something like that, and what's "wrong" with this method is that it's imprecise, not that it's inherently wrong?

I did notice that I only had it out 1.5 turns and most people have it out 2 turns and maybe a bit more. 1.5 gave me around 750rpm, although sometimes it dipped below that. I turned it out another half turn and will see if that helps, with the rough idle at least.

And by rough I mean just that, it feels rough, not purring the way it used to, oscillating between 650 and 850 and sometimes dying when I'm either in 1st or reverse and feather the clutch and brake to move the car slowly, especially when parking or getting out of a spot and I have to be careful.

As for the fuel consumption, I'm stumped. I'm talking 10-12mpg in urban driving. I used to get around 17-19mpg. So many things could be causing this of course, from somewhat seized calipers to bad bearings to an improperly rebuilt trans to all sorts of engine, IAC/FIAV, sensor or vacuum issues. I've ruled out the fuel system as I've overhauled all of it.

I'm basically looking for advice on what tend to be the most common causes, so I start looking there.

 

[pauleyman]

If yiu can find any ability to data log you will find the problem
Coolant temp sensor comes to mind. Your bad fuel mileage is not because of the biss or isc.

 

[DSMPT]

what tend to be the most common causes of rough idle and poor fuel consumption in this particular model at this point in its service life, over 31 years now?

Assuming you have checked all sensors, then I would check for throttle body adjustment (Many people pay attention to only BISS, ISC and FIAV, and miss or skip the throttle plate position), leak, base ignition timing, and valve timing. Those would affect the idle stability and fuel consumption.

 

[steve]

So adjusting the BISS manually without grounding the connectors won't give the ECM bad readings or something like that, and what's "wrong" with this method is that it's imprecise, not that it's inherently wrong?

I did notice that I only had it out 1.5 turns and most people have it out 2 turns and maybe a bit more. 1.5 gave me around 750rpm, although sometimes it dipped below that. I turned it out another half turn and will see if that helps, with the rough idle at least.

It's wrong if your not monitoring the ISC steps while you are doing it.

Putting the ECU into adjust BISS mode stops the ECU from adjusting the ISC so that when you turn the BISS the ECU isn't working against that change with the ISC. When the ECU isn't in adjust BISS mode you note that you can make quite a bit of adjustments without really changing the idle speed. That's because you have to drive the ISC out of it's adjustment range before the idle speed changes.

Of course if you do that the ECU no longer has the ability to respond to changes in load or perform it's dashpot emulation with the ISC.

10-12 MPG is problem. Can you easily push the car around on flat ground when it's in neutral? It should roll very easily. Are you sure that it's going into closed loop?

 

[XC92]

It's wrong if your not monitoring the ISC steps while you are doing it.

Putting the ECU into adjust BISS mode stops the ECU from adjusting the ISC so that when you turn the BISS the ECU isn't working against that change with the ISC. When the ECU isn't in adjust BISS mode you note that you can make quite a bit of adjustments without really changing the idle speed. That's because you have to drive the ISC out of it's adjustment range before the idle speed changes.

Of course if you do that the ECU no longer has the ability to respond to changes in load or perform it's dashpot emulation with the ISC.

10-12 MPG is problem. Can you easily push the car around on flat ground when it's in neutral? It should roll very easily. Are you sure that it's going into closed loop?

What would the neutral push test reveal? Mechanical drag (or slip) in the clutch, trans, t-case, prop shaft, rear diff, bearings or brakes?

No idea how to check for closed loop or even know what that means. I have no scan tools or any of those "Link" things. Total newbie regarding anything to do with the ECM, diagnostic SW or tools, etc., on this car. I only have experience with OBD2 scan tools on newer cars.

So far most of the work I've done on the car has been mechanical, not electronic or digital. Not asking for a refresher here as I'm sure there's plenty of useful info out there already. But would I need a custom ECM or EPROM and such to do any of this, and if so where/how do I get started, like what do I need to buy or do or read up on?

Also, on a whim I checked the ECM. I had a bad one replaced nearly 12 years ago with a Cardone reman. The original was MD166262, which was the correct # for my car. But the reman says MD159966 on the outer metal case, which apparently was for the 3000GT. The engine's run fine since then, but I'm wondering if this could be causing fuel issues.

Till now everything I've done has been the "easy" part, by which I mean it can be difficult but not especially complicated. Even rebuilding my trans was fairly straightforward with no concepts too hard to grasp. I just had to put in the work. But engines and auto electronics are a whole other thing, far more complicated from my perspective and very much specialized, and from what I understand the heart of what this site is really all about, i.e. tuning, and I've yet to get into all that.

 

[steve]

3000GT ECUs share the same physical top case dimensions so Cardone isn't very careful about keeping parts together making me wonder about the care inside. Clearly you don't have a MD159966 ECU since the 6G72 ECU wouldn't work very well on a 4G63. You'll have to open it up and look for some of the other ID's to figure out what it really is. If you're lucky the black number on the white connector is still present, that's the last 4 digits of the E2T number and traceable.

Without any insight into the status of variables in the ECU software you have to do the BISS adjustment by the book. Ground the connectors and adjust for stock idle speed. Buying or building a 1G datalogger cable would be a first step along with getting one of the old programs to talk to the ECU (TMO, MMCd, TunerPro)

WRT "neutral push test", it just gives you a simple indication as to how much rolling resistance you have. If it's easy to roll then you don't need to keep looking. If it's not start jacking up corners and check the brakes for drag. From there is gets more difficult to isolate causes.

 

[XC92]

3000GT ECUs share the same physical top case dimensions so Cardone isn't very careful about keeping parts together making me wonder about the care inside. Clearly you don't have a MD159966 ECU since the 6G72 ECU wouldn't work very well on a 4G63. You'll have to open it up and look for some of the other ID's to figure out what it really is. If you're lucky the black number on the white connector is still present, that's the last 4 digits of the E2T number and traceable.

Without any insight into the status of variables in the ECU software you have to do the BISS adjustment by the book. Ground the connectors and adjust for stock idle speed. Buying or building a 1G datalogger cable would be a first step along with getting one of the old programs to talk to the ECU (TMO, MMCd, TunerPro)

WRT "neutral push test", it just gives you a simple indication as to how much rolling resistance you have. If it's easy to roll then you don't need to keep looking. If it's not start jacking up corners and check the brakes for drag. From there is gets more difficult to isolate causes.

Better yet, I took photos of the old and "new" reman ECU's before I had the latter installed, and just verified that the latter is what's installed right now by taking a peek under the dash.

Old/original:

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

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

"New" reman:

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

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

Looks like the old one had an EPROM but the new one does not. Was I had and need to find a proper OEM ECU? Car's been running "ok" since then, which was nearly 12 years ago when I basically knew nothing about cars, but then I never started paying too close attention until a few years ago. I returned the old one as a core exchange so no longer have it.

What if anything should I be worried about with the reman ECU, and could it have anything to do with the excessive fuel consumption issue?

Finally, so you're saying that 1G ECUs put out way more data than the dozen or so error codes you can get by counting the number of long and short pulses from the OBD1 connector, and all you need is a proper cable and SW to get at it? I've never gotten into that whole aspect of DSMs.

I'm sure that I've asked about this before but haven't had a good reason to get into it until now, but now that most of the mechanical fixes and restorations have been completed all that's left is the "tuning" tasks.

Btw, just did a roll test. Car moves easily back and forth. I could even pull it from the rear bumper--gingerly. So that's not it.

 

[steve]

6578M = E2T36578 = MD166262 The M means Mask Programmed. The E means EPROM. So two 91 Federal MT AWD ECUs, one EPROM and on non-EPROM. The EPROM one looks in much better shape overall but needs it's capacitors replaced and whatever was wrong to start with fixed. If I understand your post it's gone, which is a shame.

The Mitsubishi ECUs of this general vintage have a diagnostic protocol above just pulsing error codes.

It's a 8bit (generally) command and response dialog that abstracts the physical location in processor address space for common variables the ECU keeps track of, like RPM, Injector Pulse Width, Timing Advance, TPS voltage, ISC position, O2 voltage and others. There are also some functional commands like disable an Injector, run the Fuel Pump, clear Fault Codes.

On top of that there are a couple of digital inputs that are used to give commands to the ECU, for example, Set Basic Timing Adjustment , Set BISS Adjustment, Enter Diagnostic mode.

Open/closed loop refer to the mode where the ECU fueling decisions are based off the feedback from the O2 sensor (closed loop feedback for 14.7:1 AFR) or from the fuel maps for current load. Cruise and Idle are meant to be closed loop.

WRT the Cardone it's pretty much in the same shape I've seen over and over. They used through hole capacitors to replace damaged SMD parts and it's not clear how well/closely they inspected and cleaned up the corrosion damage that caused them to change those parts in the first place.

 

[XC92]

6578M = E2T36578 = MD166262 The M means Mask Programmed. The E means EPROM. So two 91 Federal MT AWD ECUs, one EPROM and on non-EPROM. The EPROM one looks in much better shape overall but needs it's capacitors replaced and whatever was wrong to start with fixed. If I understand your post it's gone, which is a shame.

The Mitsubishi ECUs of this general vintage have a diagnostic protocol above just pulsing error codes.

It's a 8bit (generally) command and response dialog that abstracts the physical location in processor address space for common variables the ECU keeps track of, like RPM, Injector Pulse Width, Timing Advance, TPS voltage, ISC position, O2 voltage and others. There are also some functional commands like disable an Injector, run the Fuel Pump, clear Fault Codes.

On top of that there are a couple of digital inputs that are used to give commands to the ECU, for example, Set Basic Timing Adjustment , Set BISS Adjustment, Enter Diagnostic mode.

Open/closed loop refer to the mode where the ECU fueling decisions are based off the feedback from the O2 sensor (closed loop feedback for 14.7:1 AFR) or from the fuel maps for current load. Cruise and Idle are meant to be closed loop.

WRT the Cardone it's pretty much in the same shape I've seen over and over. They used through hole capacitors to replace damaged SMD parts and it's not clear how well/closely they inspected and cleaned up the corrosion damage that caused them to change those parts in the first place.

Thanks, that clears up a lot. And yeah, I shipped them the original to get the core charge back. That's $50 I wish I hadn't bothered to get back, in retrospect, as it certainly does look to be in better shape than the one they sent me. They actually sent me 2, the first being too defective to work. But that's ancient history now. If only I'd known then what I know now, I'd have replaced the caps as I did know how to solder back then. I've been modding and fixing circuit boards since I was a teen.

So would I need another ECU with an EPROM to get the diagnostic codes, or would I be able to get those even with my current ECU? If so, then what would I need an ECU with an EPROM for, to swap it out for a custom EPROM so I could do more advanced tuning on, e.g. ECMLink (if I have the name right)? And if I'm not looking to increase the power on my Talon, is that something I wouldn't need to get into?

Btw, with today's vastly more advanced and integrated digital electronics wouldn't it be possible to make a much more compact, reliable and sophisticated version of the original ECUs that would be much smaller and have more features like say Bluetooth, WiFi, cellular, USB, massively more onboard storage, removable SD cards, etc., and much more robust real time diagnostics and adjustments? And if so, has anyone done it, or is it just not worth the bother?

 

[steve]

So would I need another ECU with an EPROM to get the diagnostic codes, or would I be able to get those even with my current ECU? If so, then what would I need an ECU with an EPROM for, to swap it out for a custom EPROM so I could do more advanced tuning on, e.g. ECMLink (if I have the name right)? And if I'm not looking to increase the power on my Talon, is that something I wouldn't need to get into?

No you don't need another ECU to do that but what DSM/ECMLink can do is an order more useful just on the diagnostic side. It builds on the stock diagnostic features and improves all of it. You don't have to use it to "tune" anything but it's an expensive solution if all you do is datalog.

BTW, with today's vastly more advanced and integrated digital electronics wouldn't it be possible to make a much more compact, reliable and sophisticated version of the original ECUs that would be much smaller and have more features like say Bluetooth, WiFi, cellular, USB, massively more onboard storage, removable SD cards, etc., and much more robust real time diagnostics and adjustments?

And if so, has anyone done it, or is it just not worth the bother?

The hardware side isn't the difficult part, as you note there are huge improvements available compared to the second generation microprocessor used. Getting the software right is another story.

 

[XC92]

No you don't need another ECU to do that but what DSM/ECMLink can do is an order more useful just on the diagnostic side. It builds on the stock diagnostic features and improves all of it. You don't have to use it to "tune" anything but it's an expensive solution if all you do is datalog.



The hardware side isn't the difficult part, as you note there are huge improvements available compared to the second generation microprocessor used. Getting the software right is another story.

Is the source code for either the original PROM or any of these 3rd party logger/tuner SW available, or at least some sort of guide to what kind of data is logged and available and the control codes to send the ECU to get that data (and, way more advanced and of course dangerous, actually control the engine) if one wants to write one's own PROM?

I doubt that I'd ever tackle such a thing as it requires a really advanced understanding of auto engines which I don't even begin to have, but it does make me curious as I do have a SW background and have written code at the machine level, but ages ago.

But I'm guessing that the folks who make and sell these tuning SW aren't going to open up their code but you never know.

It might though be fun to develop a simple device to obtain, display and maybe log all the basic sensor data, since as you say the ECU does make it available. That I think I can do, with an Arduino and some tinkering.

 

[steve]

Go to github and look up mmcd. It's for the Palm Pilot but you can get an idea how the diagnostic protocol works. I've written up background info before and worked on the OEM software disassembly.

 

[XC92]

Go to github and look up mmcd. It's for the Palm Pilot but you can get an idea how the diagnostic protocol works. I've written up background info before and worked on the OEM software disassembly.

Thanks, just wanted to be pointed in the right direction, this topic has been covered endlessly in the past with lots of great content I'm sure and it's just now that I'm joining the discussion, due to circumstances.