Question about Idle Speed Maps

[Mscromer]

The F9 DMEs that have a built-in MAP sensor connector have a quirk that can force them onto the wrong idle-speed map if you do not have a MAP sensor installed. I’m not sure if that applies to your setup, but if it does, it may be contributing to your current idle issue.

The issue is that it can reduce the idle target slightly, from the normal 840 rpm target down to 800 rpm, and it also defeats the A/C idle-speed increase. With any performance cam, a little more idle speed usually helps smooth things out, while lower idle speed makes the engine less stable. So if your idle is already lower than normal, it will not idle as nicely as stock --and adding a performance cam makes that worse. I’d be inclined to raise the idle in 40 rpm increments --880, 920, 960, etc. -- until you are happy with the idle quality.

You are right that Pin 28 needs to be grounded for the idle-speed circuit to work correctly. However — and this is the big quirk — even if the harness wire at Pin 28 is connected to ground, that wire is not connected inside the F9 DME as needed for the code to select the right idle map.

F9 followed the configuration originally created by Vitesse Racing, and later used by Rogue, where the Pin 28 signal was repurposed as a MAP input. Vitesse and Rogue had you cut the harness wire at Pin 28 and connect the MAP sensor to it. F9 wanted to avoid requiring people to cut wires, so they made that change on the circuit board instead.

The only problem is that if you do not install a MAP sensor, that input floats and is not grounded. So even if you ground the harness wire at Pin 28, that ground signal never reaches the ADC circuit inside the DME that needs to see ground for the idle-speed maps to work correctly.

Until we discovered the issue here on Carpokes, no one (including F9) was aware that cutting that signal without installing a MAP sensor would alter which idle speed cells were used by the DME. Prior to the F9 DME, the issue never came up because the only people cutting the pin 28 wire were doing it in order to install a MAP sensor. The F9 DME introduced the scenario where pin 28 is 'cut' (on the circuit board) even if no MAP sensor is installed.

Fortunately, it is easy to tell if you have this issue, and easy to correct if you do.

To test it, connect the Ostrich and open TunerPro RT using the Carpokes XDF. You will see three idle-speed maps you can adjust. The main idle-speed map has only three cells, which should be set to 840 on a stock map. Once the car is warmed up, try changing all three cells to 1000 and see if the idle speed increases.

If the idle speed goes up, your DME does not have the Pin 28 quirk, and all is good.

If nothing happens, try the “alternate” idle-speed map. That one has four cells, and on a stock chip they should all be set to 800. Change all four of those cells to 1000 and see if the idle speed goes up. If so, that means you need to deal with the Pin 28 ground quirk. To do that, plug the pigtail that comes with the DME into the MAP sensor connector and short the signal wire to the ground wire (and cap off the power wire so it can't accidentally touch anything). That has the same effect as grounding Pin 28 on a standard DME, and should allow the DME to use the normal three-cell idle-speed map.

You want to use the normal three-cell map if possible, not the alternate map, because the alternate map does not let you change the A/C idle speed.

See this post, and the surrounding discussion on idle speed, for more background:

https://carpokes.com/viewtopic.php?t=3574&start=130#p49326

Hope that helps. It was a real pilgrimage to uncover all that info, with thanks to people like johnb, Dave W, and Whalenlg, and you won't find it anywhere else on the Internet unless/until our little Carpokes discovery starts making the rounds elsewhere...

I can’t thank you, Johnb, Dave W and Whalenlg enough for helping everyone with these cars. Carpokes is the best Porsche forum out there. Thank you!

Thank you for saying that. Sincerely appreciate it. I like to say we have the highest signal-to-noise ratio in the business. You'll get 20 answers within an hour on Facebook, but you won't get this.

p.s., you do have an F9 DME with MAP connector and no MAF sensor?

Correct. The DME does have the map connector and I still have the pig tail for it. I do not have a MAF yet. That’s will be in the near future. Now that you mention the MAP sensor where would that go on the 944 turbo car. Didn’t know u could put that on one.

[Tom]

I can’t thank you, Johnb, Dave W and Whalenlg enough for helping everyone with these cars. Carpokes is the best Porsche forum out there. Thank you!

Thank you for saying that. Sincerely appreciate it. I like to say we have the highest signal-to-noise ratio in the business. You'll get 20 answers within an hour on Facebook, but you won't get this.

p.s., you do have an F9 DME with MAP connector and no MAF sensor?

Correct. The DME does have the map connector and I still have the pig tail for it. I do not have a MAF yet. That’s will be in the near future. Now that you mention the MAP sensor where would that go on the 944 turbo car. Didn’t know u could put that on one.

Sorry, that was a typo. I was trying to ask if you have the F9 DME that's set up for the MAP, but without a MAP (not MAF) sensor installed. Sounds like that's your exact situation, so unless Joe fixed it on yours, that's probably part of why you are unhappy with your idle. My idle was 'fine' but seemed weaker than I wanted, which is why I started down the path that finally uncovered the issue. Surprisingly, just bumping it back to the stock 840 made a noticeable difference, although I think I settled on 880 for both the base idle and the A/C-on idle. Beware the a/c-on idle only triggers when the compressor is on, so if it's more than 40rpm higher than the base idle speed, the idle will seem to surge up and down when the a/c is on, as the a/c clutch engages and disengages. Also know the DME only increments in multiples of 40, so if you put 850 in a cell, the DME will be round to the closest multiple of 40 -- i.e., 840 -- so think in terms of 840, 880, 920, 960, etc.

Edit: re your question about the MAP sensor, it is not used by the stock DME. The F9 DME (obd+ versions) made using the Rogue Tuning M-Tune and A-Tune easier to install, back when those were in production and popular. The F9 onboard connectors are still nice for logging though.

[Mscromer]

The F9 DMEs that have a built-in MAP sensor connector have a quirk that can force them onto the wrong idle-speed map if you do not have a MAP sensor installed. I’m not sure if that applies to your setup, but if it does, it may be contributing to your current idle issue.

The issue is that it can reduce the idle target slightly, from the normal 840 rpm target down to 800 rpm, and it also defeats the A/C idle-speed increase. With any performance cam, a little more idle speed usually helps smooth things out, while lower idle speed makes the engine less stable. So if your idle is already lower than normal, it will not idle as nicely as stock --and adding a performance cam makes that worse. I’d be inclined to raise the idle in 40 rpm increments --880, 920, 960, etc. -- until you are happy with the idle quality.

You are right that Pin 28 needs to be grounded for the idle-speed circuit to work correctly. However — and this is the big quirk — even if the harness wire at Pin 28 is connected to ground, that wire is not connected inside the F9 DME as needed for the code to select the right idle map.

F9 followed the configuration originally created by Vitesse Racing, and later used by Rogue, where the Pin 28 signal was repurposed as a MAP input. Vitesse and Rogue had you cut the harness wire at Pin 28 and connect the MAP sensor to it. F9 wanted to avoid requiring people to cut wires, so they made that change on the circuit board instead.

The only problem is that if you do not install a MAP sensor, that input floats and is not grounded. So even if you ground the harness wire at Pin 28, that ground signal never reaches the ADC circuit inside the DME that needs to see ground for the idle-speed maps to work correctly.

Until we discovered the issue here on Carpokes, no one (including F9) was aware that cutting that signal without installing a MAP sensor would alter which idle speed cells were used by the DME. Prior to the F9 DME, the issue never came up because the only people cutting the pin 28 wire were doing it in order to install a MAP sensor. The F9 DME introduced the scenario where pin 28 is 'cut' (on the circuit board) even if no MAP sensor is installed.

Fortunately, it is easy to tell if you have this issue, and easy to correct if you do.

To test it, connect the Ostrich and open TunerPro RT using the Carpokes XDF. You will see three idle-speed maps you can adjust. The main idle-speed map has only three cells, which should be set to 840 on a stock map. Once the car is warmed up, try changing all three cells to 1000 and see if the idle speed increases.

If the idle speed goes up, your DME does not have the Pin 28 quirk, and all is good.

If nothing happens, try the “alternate” idle-speed map. That one has four cells, and on a stock chip they should all be set to 800. Change all four of those cells to 1000 and see if the idle speed goes up. If so, that means you need to deal with the Pin 28 ground quirk. To do that, plug the pigtail that comes with the DME into the MAP sensor connector and short the signal wire to the ground wire (and cap off the power wire so it can't accidentally touch anything). That has the same effect as grounding Pin 28 on a standard DME, and should allow the DME to use the normal three-cell idle-speed map.

You want to use the normal three-cell map if possible, not the alternate map, because the alternate map does not let you change the A/C idle speed.

See this post, and the surrounding discussion on idle speed, for more background:

https://carpokes.com/viewtopic.php?t=3574&start=130#p49326

Hope that helps. It was a real pilgrimage to uncover all that info, with thanks to people like johnb, Dave W, and Whalenlg, and you won't find it anywhere else on the Internet unless/until our little Carpokes discovery starts making the rounds elsewhere...

Well I just checked my idle speed maps and only the 4 cell alternate map works so I’ll plug the pig tail in and ground pin 28. So once I get a MAF I can unplug that pig tail and not have the 28 pin quirk on the F9 DME?

Edit. My pig tail has red black and orange. I’m assuming black is ground and orange is signal? So connect the orange to black to short it out and cap off the red?

[Tom]

Well I just checked my idle speed maps and only the 4 cell alternate map works so I’ll plug the pig tail in and ground pin 28. So once I get a MAF I can unplug that pig tail and not have the 28 pin quirk on the F9 DME?

That means your F9 DME is like mine and you need to ground the map sensor input signal in order for the DME to use the correct idle map.

Just to be clear, you don't need to do anything with the DME harness wire on pin 28 -- grounding that wire will do no good because it is not connected to anything inside that DME.

To fix the issue, plug the MAP sensor pigtail into the board. On that pigtail, there should be a power wire (probably red), a ground wire (probably black) and a MAP signal wire (probably green, maybe yellow). What you want to do is connect the ground wire in that pigtail to the MAP signal input wire in that same pigtail (I always solder and cover with adhesive lined shrink wrap). Then be sure to protect the remaining wire(s) in the pigtail so they cannot make electrical contact with anything.

I can't find color-coded pin-outs anywhere for the MAP pigtail, so if it's not clear to you which signal is which, I can always open mine up and tell you as a last resort -- but hopefully Joe can tell you.

When this first came up, Joe @FTECH9 said there was a spot on the board for a pull-down resistor that would fix the idle issue with or without a MAP sensor installed. He also said he may start installing that resistor on future DMEs. That would be the more elegant solution, and one I'd rather use, but I'm not sure where that resistor resides on the board and/or what value would be best (if not just a random 10k resistor or the like) -- so hoping maybe Joe sees this and offers an update.

[Mscromer]

Well I just checked my idle speed maps and only the 4 cell alternate map works so I’ll plug the pig tail in and ground pin 28. So once I get a MAF I can unplug that pig tail and not have the 28 pin quirk on the F9 DME?

That means your F9 DME is like mine and you need to ground the map sensor input signal in order for the DME to use the correct idle map.

Just to be clear, you don't need to do anything with the DME harness wire on pin 28 -- grounding that wire will do no good because it is not connected to anything inside that DME.

To fix the issue, plug the MAP sensor pigtail into the board. On that pigtail, there should be a power wire (probably red), a ground wire (probably black) and a MAP signal wire (probably green, maybe yellow). What you want to do is connect the ground wire in that pigtail to the MAP signal input wire in that same pigtail (I always solder and cover with adhesive lined shrink wrap). Then be sure to protect the remaining wire(s) in the pigtail so they cannot make electrical contact with anything.

I can't find color-coded pin-outs anywhere for the MAP pigtail, so if it's not clear to you which signal is which, I can always open mine up and tell you as a last resort -- but hopefully Joe can tell you.

When this first came up, Joe @FTECH9 said there was a spot on the board for a pull-down resistor that would fix the idle issue with or without a MAP sensor installed. He also said he may start installing that resistor on future DMEs. That would be the more elegant solution, and one I'd rather use, but I'm not sure where that resistor resides on the board and/or what value would be best (if not just a random 10k resistor or the like) -- so hoping maybe Joe sees this and offers an update.

No need to open up your DME. The pig tail has red black and orange. I’ll solder the black and orange together and cap off the red. Thanks for all your help!

[chris white]

chips are best served with guacamole……
there, I said it…
BTW – I have always said the best improvement to engine management is to replace all the wiring. Its old and cranky.

[Tom]

chips are best served with guacamole……
there, I said it…
BTW – I have always said the best improvement to engine management is to replace all the wiring. Its old and cranky.

My dietary kryptonite....

IMG_0648.jpeg (130.96 KiB) Viewed 121 times

[Mscromer]

IDLE ISSUES FIXED
Just want to give an update on how I finally fixed my idle issues on my newly rebuilt turbo engine. Hopefully this will educate others so they don’t have to go through what I did. A little back story, had my engine, turbo and wastegate rebuilt by Lindsey Racing. Everything on the engine is new including a Kroon harness. The only thing still original to the car is the MAF. Will be getting a Storch MAF soon. Started it up for the first time back in February. I have been fighting a rough idle ever since. Went down every rabbit hole I could. Ohmed out the MAF, TPS replaced the new DME temp sensor incase I got a bad one, messed with the idle adjustment. Never found anything wrong and idle never improved. Finally decided to go down the tuning rabbit hole. Ordered a Ostrich 2.0 and downloaded TunerPro RT. Started messing with different parameters and still nothing helped. Finally with @Tom help he figured out I was having issues with my F9 DME. Evidently there is a quirk in the F9 DME where if you’re not running an aftermarket MAF the DME will use an alternate idle map and not allow the ICV to operate properly. I emailed F9 and they did verify this quirk. To get around this you have to short the signal wire to ground using the MAF pigtail sent with the DME and plug it into the MAF plug on the F9 DME. Ever since then the car has been idling great. Even when I turn the AC on the engine doesn’t idle down or stumble. After all these months of idle issues @Tom finally figured it out and F9 verified they know about this quirk with their DME. If they know about it why don’t they send a little note with their DME telling you how to fix the issue if you’re still using the original MAF. This has kind of pissed me off a little with F9. Well anyway, after months of going down every rabbit hole it’s fixed. Hopefully this will help some else in the future. I can’t thank @Tom enough. Without him I would still be having issues and pulling my hair out. THANKS @Tom !!

[Tom]

IDLE ISSUES FIXED
Just want to give an update on how I finally fixed my idle issues on my newly rebuilt turbo engine. Hopefully this will educate others so they don’t have to go through what I did. A little back story, had my engine, turbo and wastegate rebuilt by Lindsey Racing. Everything on the engine is new including a Kroon harness. The only thing still original to the car is the MAF. Will be getting a Storch MAF soon. Started it up for the first time back in February. I have been fighting a rough idle ever since. Went down every rabbit hole I could. Ohmed out the MAF, TPS replaced the new DME temp sensor incase I got a bad one, messed with the idle adjustment. Never found anything wrong and idle never improved. Finally decided to go down the tuning rabbit hole. Ordered a Ostrich 2.0 and downloaded TunerPro RT. Started messing with different parameters and still nothing helped. Finally with @Tom help he figured out I was having issues with my F9 DME. Evidently there is a quirk in the F9 DME where if you’re not running an aftermarket MAF the DME will use an alternate idle map and not allow the ICV to operate properly. I emailed F9 and they did verify this quirk. To get around this you have to short the signal wire to ground using the MAF pigtail sent with the DME and plug it into the MAF plug on the F9 DME. Ever since then the car has been idling great. Even when I turn the AC on the engine doesn’t idle down or stumble. After all these months of idle issues @Tom finally figured it out and F9 verified they know about this quirk with their DME. If they know about it why don’t they send a little note with their DME telling you how to fix the issue if you’re still using the original MAF. This has kind of pissed me off a little with F9. Well anyway, after months of going down every rabbit hole it’s fixed. Hopefully this will help some else in the future. I can’t thank @Tom enough. Without him I would still be having issues and pulling my hair out. THANKS @Tom !!

Awesome — so glad to be able to help. Not sure why Joe hasn’t addressed that, since it seems like low-hanging fruit. Or maybe he has and we both just have early versions. I dunno.

As much as I’d like to take credit for figuring this all out, @johnb traced the DME and discovered the key CPU-logic routing issue that led to unearthing the F9 quirk. He’s the Gordon Ramsay of the group. I just put it on a pretty plate, add parsley, and serve it.

For posterity, just a few clarifications. The factory used an Air Flow Meter, aka AFM, which measures the volume of air with a big flapper door, aka barn door. Aftermarket MAFs like the Storch are Mass Air Flow sensors, which measure the mass of air with a hot wire. The F9 quirk doesn’t relate to whether you use an AFM or MAF as your main air-measuring device — either is fine.

The F9 quirk relates to a third thing known as a Manifold Absolute Pressure (MAP) sensor. The F9 DME is designed to accommodate a MAP sensor, which is used for logging vacuum/boost and/or as a supplement to some Rogue Tuning and Vitesse Racing kits. This is where the quirk is.

If you have the F9 without a MAP sensor installed, you need to short the MAP sensor signal to ground, whether you have an AFM or MAF. If you don’t, the DME will use the wrong internal idle maps — johnb’s discovery — which produce a lower, weaker idle. The ISV still works; it’s just being told to maintain a slow, weak idle. You also lose the ability to adjust RPM when the A/C is on, potentially causing more idle instability.

By installing a MAP sensor, or grounding the input for the MAP sensor, the DME then knows to use the correct idle map and all is well again. See my post above (here) for more on that.

[Mscromer]

IDLE ISSUES FIXED
Just want to give an update on how I finally fixed my idle issues on my newly rebuilt turbo engine. Hopefully this will educate others so they don’t have to go through what I did. A little back story, had my engine, turbo and wastegate rebuilt by Lindsey Racing. Everything on the engine is new including a Kroon harness. The only thing still original to the car is the MAF. Will be getting a Storch MAF soon. Started it up for the first time back in February. I have been fighting a rough idle ever since. Went down every rabbit hole I could. Ohmed out the MAF, TPS replaced the new DME temp sensor incase I got a bad one, messed with the idle adjustment. Never found anything wrong and idle never improved. Finally decided to go down the tuning rabbit hole. Ordered a Ostrich 2.0 and downloaded TunerPro RT. Started messing with different parameters and still nothing helped. Finally with @Tom help he figured out I was having issues with my F9 DME. Evidently there is a quirk in the F9 DME where if you’re not running an aftermarket MAF the DME will use an alternate idle map and not allow the ICV to operate properly. I emailed F9 and they did verify this quirk. To get around this you have to short the signal wire to ground using the MAF pigtail sent with the DME and plug it into the MAF plug on the F9 DME. Ever since then the car has been idling great. Even when I turn the AC on the engine doesn’t idle down or stumble. After all these months of idle issues @Tom finally figured it out and F9 verified they know about this quirk with their DME. If they know about it why don’t they send a little note with their DME telling you how to fix the issue if you’re still using the original MAF. This has kind of pissed me off a little with F9. Well anyway, after months of going down every rabbit hole it’s fixed. Hopefully this will help some else in the future. I can’t thank @Tom enough. Without him I would still be having issues and pulling my hair out. THANKS @Tom !!

Awesome — so glad to be able to help. Not sure why Joe hasn’t addressed that, since it seems like low-hanging fruit. Or maybe he has and we both just have early versions. I dunno.

As much as I’d like to take credit for figuring this all out, @johnb traced the DME and discovered the key CPU-logic routing issue that led to unearthing the F9 quirk. He’s the Gordon Ramsay of the group. I just put it on a pretty plate, add parsley, and serve it.

For posterity, just a few clarifications. The factory used an Air Flow Meter, aka AFM, which measures the volume of air with a big flapper door, aka barn door. Aftermarket MAFs like the Storch are Mass Air Flow sensors, which measure the mass of air with a hot wire. The F9 quirk doesn’t relate to whether you use an AFM or MAF as your main air-measuring device — either is fine.

The F9 quirk relates to a third thing known as a Manifold Absolute Pressure (MAP) sensor. The F9 DME is designed to accommodate a MAP sensor, which is used for logging vacuum/boost and/or as a supplement to some Rogue Tuning and Vitesse Racing kits. This is where the quirk is.

If you have the F9 without a MAP sensor installed, you need to short the MAP sensor signal to ground, whether you have an AFM or MAF. If you don’t, the DME will use the wrong internal idle maps — johnb’s discovery — which produce a lower, weaker idle. The ISV still works; it’s just being told to maintain a slow, weak idle. You also lose the ability to adjust RPM when the A/C is on, potentially causing more idle instability.

By installing a MAP sensor, or grounding the input for the MAP sensor, the DME then knows to use the correct idle map and all is well again. See my post above (here) for more on that.

Well thank you @johnb also. Thanks to all you guys that are deep in the trenches on these cars. Thanks for clarifying the F9 quirk. Tried to do my best understanding what was actually going on. That makes more sense on the ISV now cause on all my rabbit holes, one was to make sure you could fell it vibrating, which I could. But now with just the key on I can hear that thing buzzing. Lol