*HELP*DSM afc related

[tallrd]

Thanks for the responses everyone.

I do not have an a/f guage, BUT I did confirm with many people that the vac source which I used off the intake mani is indeed correct (cuz it was wrong the first time-- (hand) I got some bad information initially). I'm 99.9% sure it's the DSM because it never did this until now--just about the same time/mileage the ECU take to settle in after being reset. And that's the only mod I did at that time. The whole reason I purchased this unit was plug-and-play. I was hoping to avoid cracking it open and having to load a new map into the unit.

Look at my sig for my mods, and yes, I am running the FMIC map intended for 8psi. I'm not currently running 8psi, but my problem is under idle and not boost, so that shouldn't matter, right?

I will get in touch with John (or at least try) to see if he has a solution for the idle problems. The DSM unit still makes a great difference at wot, so that's my saving grace so far.

 

[peepsalot]

i was watching the car's inputs on a laptop and recorded it all at vac it showed it was pulling values of 10. But the value of 10 does not = 100% untouched. I just never figured out what it was doing.


And believe me i checked all the wires, all the vac hoses. Everything

From everything I've been hearing, it seems like this is the case. I mean I have seen the maps and I understand that the stock maps are set to 10 for anything in vacuum, but apperently 10 doesn't exactly leave the fuel values untouched.

I was wondering if maybe the way it alters the signals causes the sequential fire of the injectors to delay and therefore not synch up with the valve opening. I don't know exactly how the inner working of the signal processing go though. It wouldn't matter is higher rpm because at that point the stock system basically fires in batch mode anyways.

This page has some explanation of batch vs sequential, in case anyone isn't familiar.
http://www.sdsefi.com/techseq.htm

But I'm pretty sure that only the MAF signal is altered by the AFC, which should not affect the timing of the injection. So yeah the simpler answer that 10 is not equal to 100% for whatever reason, is more likely.

I've had idling problems for a while now, which I didn't attribute to the AFC since they existed before the install. I have since found and fixed a couple vacuum leaks but am still having problems, so I think the AFC could have something to do with it with so many others having similar problems.

I've had problems getting access to a laptop to do any tuning of the map on my own, but I've recently worked that out.

 

[peepsalot]

Ok, I checked the install instructions again because I couldn't remember which signals were being altered. Looks like MAF and O2 are the only ones the AFC is wired to alter.

 

[Mallard]

Ok, I checked the install instructions again because I couldn't remember which signals were being altered. Looks like MAF and O2 are the only ones the AFC is wired to alter.

that's correct.

 

[DSMConvert]

wow not sure how i didnt see this thread earlier but none the less heres the run down on the idle problem. As peeps indicated the afc only modifies the o2 and maf signals, so it doesnt have any effect on injector timing or anything like that. Ok now we know the problem only occurs at "cold" engine start up temps. THe 3 sensors that have the biggest effect on cold start up conditions are the IAT, the IAC, and the o2 sensor, all 3 need to heat up before they work efficiently. Hence we've narrowed the problem down to these areas. Now the afc is indeed only set to function at above 1psi, please make sure that this is the case under your options menu. Below this value the afc is set to 10 and is essentially "off" as far as signal modification goes. Now with any other car made by man this is fine b/c the ecu is able to deal with it, but if what we're finding is correct the ecu is not dealing with this...its like it needs the afc to regulate idle too...Now as indicated above the ecu is NOT able to learn around the afc, no way no how, BUT what it is doing, we THINK, is getting used to the leaner values its seeing in the higher rpm range and extrapolating it to idle conditions, and since our afc doesnt do anyhthing to the idle values the ecu gets "confused". THe more mods the car has the more the idle seems to be affected...ie more cold air from a fmic just means more confusion for the ecu at idle. Now we've been working on updated maps with idle corrections but the problem is we're getting conflicting info some of you are telling us you're lean at idle while others are telling us you are extremely rich. Now as indicated before 10 is indeed pass through no alter, but the problem lies on the fact that the ecu has become completely reliant on the afc for its signals...kinda weird b/c most ecus do the exact opposite and just ignore the altered signals coming in. SO there are 2 ways on going about fixing the idle issue. Either A. I can build a map for everyone to use, BUT this may give idle issues to those that currently do not have them, or B. you can load up the software and adjust the values and decided which produced the optimum idle for your application. As we've seen we have people running fine, people runnign rich, and people running lean, so it seems to be a random ecu problem. IN all the applications we've had hands on experience it seems to be rich conditions, but a lot of you are saying lean, so at the moment it seems to be random which way the ecu gets confused. On the fmic map note, that map was built for a very specific set of conditions and as you know requires a dyno to fine to it to your setup, I've had a lot of people that get better results with the stock map until after they've had a chance to go to the dyno. So if you guys could please send me all the info you can on idle problems, everything from weather condtions, to mods, to a/f readings...the more info I can get the better.

 

[peepsalot]

Hmm, I think I might hook up a DPDT switch in a way to be able to switch between the altered and unaltered signals to the ECU. That would be pretty pimp. I'm pretty sure it would just take that switch and a couple wires. Might even be able to build it into the box if there is room.

 

[noclue119]

Hmm, I think I might hook up a DPDT switch in a way to be able to switch between the altered and unaltered signals to the ECU. That would be pretty pimp. I'm pretty sure it would just take that switch and a couple wires. Might even be able to build it into the box if there is room.

ya whatever u do don't cut power to the AFC while the car is running. That is no good talk about a CEL coming on quickly.

 

[tallrd]

well, DSM's info gives me a direction to try. At this point, it would seem like an awful coincidence for it to NOT be the afc unit. btw, what do those 3 abreviated terms which you said are the other possibilities stand for?

I will load the software into my PC/laptop and double check the settings. Hopefully, I'll have my answer there. A lot of the directions for the more sophisticated settings were beyond my realm of knowlege so I didn't want to start tweeking values.
I don't have an a/f guage; do I NEED one to this?

 

[mofoaka]

well, DSM's info gives me a direction to try. At this point, it would seem like an awful coincidence for it to NOT be the afc unit. btw, what do those 3 abreviated terms which you said are the other possibilities stand for?

I will load the software into my PC/laptop and double check the settings. Hopefully, I'll have my answer there. A lot of the directions for the more sophisticated settings were beyond my realm of knowlege so I didn't want to start tweeking values.
I don't have an a/f guage; do I NEED one to this?

it woudl be best if you had a wideband A/F gauge with datalogging. then you'd know exactly how to adjust it.


BTW john, do you think a datalogged a/f graph from a wideband would be suitable for you to remake my map?

 

[speed1016]

Thanks John. Well me for one Im lean at idol. But Im a little confused when it comes to downloading maps. Do I just plug in a laptop with the software and go from there? And is it going to hurt anything if I go to the dyno with the afc/ecu acting like this with 10 psi on forged internals? Thats the only way Ill be able to provide proper afrs.

 

[ZMN BY U]

i have same problem car wants to cut off on cold starts on very hot days rough idle. car runs lean until its idles for about 30 secs then its ok. I run supermap with my mods listed below. I work for mazda and i have no codes in ecu at all they think its 02 sensor in exhaust going bad or the signal from my afc to the ecu that has it going haywire. also the ecu will learn the map settings after awhile this also is a factor. john sorry your wrong the ecu does learn its there and the ecu trys to compensate for the added fuel and maf settings to correct the problem thats why we get idle problem.

 

[tallrd]

i have same problem car wants to cut off on cold starts on very hot days rough idle. car runs lean until its idles for about 30 secs then its ok. I run supermap with my mods listed below. I work for mazda and i have no codes in ecu at all they think its 02 sensor in exhaust going bad or the signal from my afc to the ecu that has it going haywire. also the ecu will learn the map settings after awhile this also is a factor. john sorry your wrong the ecu does learn its there and the ecu trys to compensate for the added fuel and maf settings to correct the problem thats why we get idle problem.

uh oh. I'd love to see how this pans out.

 

[DSMConvert]

zmn perhaps you misread my post or perhaps we're unclear on terms...when an ecu learns around a device it renders the device useless at any rpm or boost level, if this was the case the afc would not work period to lean out the car ever....over a year now on our test cars....still runnin perfect a/f ratios through dyno runs....only thing affected is the idle, which is when the afc IS NOT activated....so the ecu hasnt learned anything, it simply doesnt know what to do when it doesnt get its signal from the afc....in order for an ecu to learn around any piggy back setup it would require it to either have a true maf signal or a true o2 signal....our afc interrupts both, thus the ecu has NO WAY of realizing that its running leaner than it was programmed to...I'm not yelling or anything like that I just want to make sure we understand the terms eachother are using...

 

[ZMN BY U]

John i have tried two things 1 running a cat less dp with a cel elimnator in it. 2nd way running a new 02 sensor with factory dp. Tech said try that see if it clears it up it didnt clear it up. Im an service advisor at a vw mazda store in virginia. The techs have said there is no detection of the afc but we can see at idle where the ecu is trying to comensate for the overly rich tune thats in the afc. Also im unflashed. And to add one more comment ive had your afc for roughly 6 months and ive pmd you about concerns and you have been more then helpfull to me, but the ecu is trying to compensate for the signal interference it sees after you boost and the afc does its thing the ecu learns what the afc was doing so it does this on idle. Ecu says ok car run lean at start up well it has just been running so rich and so much fuel being dumped its overloaded so yes it will learn its there we see the ecu trying to fix the problem it finding but cant because its learned from the afc to do what we have told it to. I have hooked scan tools up to it watch it run at idle at wot and you should see the difference in the ecu and what it does. It must be a mazda scan tool to get accurate readings. I love your afc not doging your product but maybe I made a mistake in buying it. It has caused more issues with idle then i want. Thanks just trying to help out guys not doging the afc or john hes been nothing but awesome in the forum but maybe it cant be fixed guys. I tihnk the damage has been done and its not fixable at idle.

 

[speed1016]

hmmm?

 

[jeffmsp]

make a rpm based switch so the afc is not altering signals below 1500RPM-just an idea.

 

[ZMN BY U]

I dont think the afc is altering the idle signal i think the ecu is trying to fix the afc. Because it sees it as a problem. So it still thinks it is being changed when in boost. my car only acts like this when i drive it and hit boost. So when i start it up in the morning to go to work it idles great. But after getting into boost and say i go get gas cut the car off go inside get coffee and come back out i start it up and it wont idle because the ecu still thinks the afc is in control so it runs lean because the afc is not being used at that time. There is noway to fix it but remove the afc is what im told. But even that might cause some damage to the ecu because its been tricked into thinking what to do with the afc. What if we disconnct the 02 sensor wire from the afc to the ecu and wire it up the factory way what will happen?

 

[tallrd]

strange because I have the exact opposite symtoms. Idle sucks when I first start it up, but when under boost the car is night and day better with the DSM. Typically, this only happens at first start up and seems to do so with random/unpredictable levels of severity. Either way, I have yet to hook it up to a laptop and check it --too lazy to go get an RS232 DB9 cable.

On a related note of help request from you all, I put a mbc in yesterday and it created a new problem which I think is DSM related (or ignorance related on my part). Here's the flow so we're on the same page: turbo wategate line --> mbc-->"t" for boost guage vac -->"t" for DSM vac source--> intake mani. I also "t" into the bb line for my vac source on the greddy. So now that you now what's getting vac where, here's my problem:

The car stutters like mad and certain levels of throttle (but, again, seemingly random), and I'm mostly certain that it's because I have the vac source for the DSM AFTER a mbc and it's confusing it with readings which are too variable or something.

$1M question: Does the DSM need a dedictate vac line, or vac/boost line, and can it share a line? If so, where in that chain should it go and on what vac line?

Thanks to all you guys for the help/support.
Shawn

 

[Yay-Boost]

$1M question: Does the DSM need a dedictate vac line, or vac/boost line, and can it share a line? If so, where in that chain should it go and on what vac line?
Shawn

I would put the wategate/MBC line off the Brake booster and put the boost gauge and DSM unit off the manifold.

 

[tallrd]

I would put the wategate/MBC line off the Brake booster and put the boost gauge and DSM unit off the manifold.

Thanks for the reply, but is doesn't clearly answer my question.

How does a mbc control boost if it's not hooked up to the wastegate, but into the brake booster line?? If you know something which I don't, please explain.

Also, the wastegate DOES go into the mani so I'm not sure what you mean when you say the DSM unit taps off the mani. Which line off the mani?