A Call for local Michigan Members - 95 OBD 1.5 Help

[vipmiller803]

Does anyone near me have a scan tool capable of getting 95 Buick ECM codes?

 

I have been struggling with a check engine light for a while now, and now that spring is upon us, I would like it fixed for summer driving. The car is a 95 Buick Regal with a 3800. A few hundred miles ago, I started getting a SES light that comes and goes. At the time, it seemed random. At one point, I replaced a cracked EGR connector and thought that fixed it, only to get it again a couple weeks later. I have no idea what difference that made, but it did make it stop for around 10 days - maybe just a coincidence. It has started happening more and more often since then. Now it lights up every time the engine reaches operating temperatures without fail.

 

Here is how it goes: At start-up, the light is on if it was on the last time the car ran. The light turns off during warm-up. The light comes back up a few miles away, right around the time the engine warms up. I imagine it is around the time it tries to go into closed loop. I am wondering if I am running in open loop full time because of this. I have replaced the oxygen sensor not too long ago (with a new bosch one). I have also replaced the EGR valve and connector (not new, but made no difference). I just replaced the TPS this morning and its connector because a couple of times I have gotten a sudden high idle in the past. Had no effect on the SES light. I would prefer not just switching out sensors till it is fixed, although with the fun OBD 1.5, that is a real possibility for me.

 

I did try to use tunerpro to get some data from the computer, but this is the first time I have ever used it and so far I get some data, but it is erratic enough to call random. For the connection, I used RobertISaar's suggestion from another form for the USB cable. Like I said, not really an expert in tunerpro and have never tuned an EFI system before so I am not looking to do that, I just thought it would be an inexpensive alternative to getting raped by the local shops just to pull codes. I used the A and M pins for the connection. Only one other pin is used and I do not believe it has anything to do with the ECM.

 

Now that your eyes hurt from reading my long post, does anyone around here have a tool capable of pulling codes from these cars?

 

Comments and suggestions are also welcome.

 

Thank you,

 

Miller

[Psych0matt]

I'm not too far from you, but unfortunately I don't have one

[vipmiller803]

Alright. Thank you anyways. I know there are a few others around here. Any other OBD 1.5 experts?

[Nas Escobar]

I've managed to pull codes with TunerPro before. I've been meaning to do a how to on here.

 

If you can't read the data because it's too erratic, your computer may not be able to run the program. IIRC, TunerPro needs a good processor otherwise it lags and acts crazy.

 

What computer did you try to do this with if I may ask?

 

I know I had the erratic problem with a Dell tablet, bought a knock off MIcroCenter tablet, had the same issue until I decided to plug it in. Processor ran better and I now use that as a datalogging app.

 

We can talk tunerpro if you're sure your computer can run it. Maybe a screen capture of what you're talking about can help us fix your issue so you can use Tunerpro.

[vipmiller803]

That is actually a great point. I am using a crappy HP elitebook tablet looking thing that is quiet old. I had it as a spare and figured I could dedicate it as my tunerpro computer. It was not plugged in any of the times I tried it. I do have a car plug for it so I can try that. If that does not work I will try my regular laptop and see if that does it. That one has an intel i5 in it so it should handle it.

[Nas Escobar]

Haha I should have known.

 

The laptop will work for sure. I think tunerpro needs at least XP style specs.

 

Sent from my SAMSUNG-SGH-T989 using Tapatalk

[vipmiller803]

Got the real specs for the computer. Core2Duo 1.4GHz. Not exactly a stout processor, but it did come with XP (probably would run better with it too). I did get much better readings with it plugged in, but it still hops around a few times about every 3 seconds. Working on freeing up some more processing power and trying it before permanently writing off this computer. I would rather be able to use it than lug the big laptop around.

 

One weird thing, just had a drive in the Buick, no SES. Came back home, hooked it up (with laptop plugged in), and it showed loop status open and codes all over the place, some blinking. SES never went on once. Bulb did not burn out.

 

Am looking into data logging a short drive now. More to follow.

[RobertISaar]

can you post your ADX? I may have a fix or two to get the datastream corrected.

[vipmiller803]

I just tried to attach it but neither adx nor zip files are allowed here.

 

Running the crappier computer with selective startup services and while plugged in seemed to reduce the glitches to a tolerable level, so I think the adx should be fine. Thank you for the offer though. Also, RobertISaar, thank you for your post from the other form about the USB cable! Extremely helpful.

 

Back to the Buick:

 

I took a nice long drive and data logged. The SES seems to be gone. Maybe I just did not give the TPS enough time? I figured around 1.5 hours of driving with it should be enough to be definitive, especially with a non-OBDII system. Maybe I was wrong.

 

I am still working on getting graphs of the data I logged. I feel like it should be trivial but I am a big newb with tunerpro, so does anyone have a link or instructions for that?

[vipmiller803]

I will start with the Buick:

 

No SES light so far. I will update that if it changes anytime soon.

 

For tunerpro, I did want to put up what I did log, so if anybody feels like looking at a short drive, feel free to look below.

 

I tried to display it as nicely as possible, but it is still pretty crowded. I think I like what I am seeing, but if I were to see something at the end of its life, I doubt I would be able to recognize it due to lack of experience.

[vipmiller803]

Well the SES light is back on. This time I had the tunerpro computer with me. I am not sure if the code being thrown is the same, although I imagine it would be, given how soon it came back on.

 

I data logged just a few seconds in order to save all those "error"s I see, but I am not sure, just from seeing them, which ones are actual codes.

 

Here is a link to a google drive folder that includes the data collected today (carrecord2) and the adx file.

 

https://drive.google.com/file/d/0B6AKqXKswy6VNHotNTF2TEN3WW8/view?usp=sharing

 

It also has the log from before (carrecord1) and the data backing it, considering the poor image quality from before.

[RobertISaar]

that has CSVs in it, but is the XDL still kicking around? that would be helpful for troubleshooting.

[RobertISaar]

I also see a few issues in the ADX... may explain the goofy looking data. i'll see if I have one of mine for this application kicking around. I know Andrei has used it without issues back a few years ago, I just need to find it.

[vipmiller803]

I used CSVs to look at graphed data. I could not figure out plotting in tunerpro and good old excel is easier for me to manipulate.

 

I will get original files up there in a few. The data does still go erratic at times, although it is a ton smoother now.

 

From what is there now, though, are any of these "error"s actual codes?

[RobertISaar]

the way the ADX was setup, I don't think that data should be used.... most, if not all of the codes are held in the mode 1 message 1 response. the only request the ADX was sending to the PCM was mode 1 message 0 but the codes were setup to respond to both.

[vipmiller803]

I see. I did download that adx right from the tunerpro definition download list.

 

Might be my time to learn how to write/edit those files, I guess.

 

I will not bother upload the original xdl's if that data is not to be used.

[RobertISaar]

https://www.dropbox.com/s/ywfpagorua8qpwb/A221.adx?dl=0

 

there is my version I just created, it should produce significantly more useful results. I had to split the item list into 2 pieces due to running into the item number limit of 256. one is nothing but codes.

[vipmiller803]

Thank you for that. It is really awesome to have such great support!

 

I will use that adx to log my trip home.

 

Thank you again!

[vipmiller803]

https://drive.google.com/file/d/0B6AKqXKswy6VZ09CNHYzeUlaclk/view?usp=sharing

 

Here is the link to the xdl using that last adx. I do see the codes tab there and there were two history codes on. The data was stable but it still hit a connection error every 10 seconds or so, with the frequency oscillating somewhere around 10 Hz.

 

Assuming this data is to be trusted, it is quiet exciting to be able to read those codes after all the trouble with communicating with this ECU. So much for 85 dollars just to hook up the machine.

 

It looks like there is one O2 sensor code and one ECU code. The O2 sensor was replaced recently but I could warranty it if needed. The ECU code might be my fault, considering I have rigged fan 1 with a manual switch. That was done a long time ago, though, and the SES has only been recent.

 

To me, the O2 is reading fine, although the issue might be intermittent, considering the now intermittent SES light.

[Nas Escobar]

Try grabbing the BLM and fuel trim readings. A screen shot of the gauges on TP would suffice to see more data of what the engine/computer is doing.

[RobertISaar]

10Hz seems about right, and the data looks trustable. a connection error every 10 seconds or so is somewhat odd, but nothing alarming. you could likely get rid of it by slowing down the request rate, but I would just live with 1 dropped sample out of 100 unless you can slow the stream down by less than 1%, otherwise you're at a net loss of data acquired.

 

anyways, your manual fan switch is likely the cause of code 670, the 4 items that quad driver controls are the SES lamp, fan 1, fan 2 and the a/c relay.

 

code 134 can be set by a slow-reacting O2 sensor. could be a faulty sensor, or if the sensor's heating element is faulty, that can cause it as well. depending on how much the car has been driven since replacing it, it could be from the old sensor. can clear code data via the Data Acquisition Send Command function. otherwise, you may have to wait ~50 drive cycles or so for it to be dropped off automatically. or pull power to the PCM for a minute, but that can cause drivability issues until certain default operating values are learned out.

 

your log starts 5 minutes after the engine was fired, so I can't comment on how the O2 reacts cold, but when warmed, looks like 5 transitions/second under a normal load. fuel trims seem about how I expect them to be.

[vipmiller803]

Nas,

I am using an adx with no gauges set up right now. It is within the links so far here and so is the xdl. I will take a couple screenshots of numerics, but a few data points seem like a small sample to judge from.

 

Robert,

I will not touch the request rate (not that I already know how, hah) seeing how your logic is pristine on that.

 

As soon as I saw that quad4 code, I thought of that switch. Looking up what it controls, as you said, confirms it. I am wondering if the ECU damage is permanent with my switch. I will have to check the wiring on that switch, maybe when it gets warm out. I can add a relay to give back the ECU's ability to run the fan. I think I just used a diode so I do not short the ECU (which happened back then at least once).

 

For the 134 code, if I am doing the math right, I may or may not have had 50 cycles since the sensor replacement. I replaced it in December and drove it a while, but I drove the jeep for a month and half after that . It is a nice cool 30 degrees outside now so I will log the cold start and trip to work to see how that sensor is doing.

 

More food for thought:

quad4 = SES light control, does that mean erratic light behavior? Fan 2 control still works well. As of last year, AC clutch control also works well. Seems to me like the quad4 is just a bunch of (four) MOSFETs ordered by ECU logic, which would make erratic light behavior highly unlikely.

 

Lastly,

Thank you, to both of you, for all this support. I would have not been very far without the help.

[vipmiller803]

https://drive.google.com/file/d/0B6AKqXKswy6VN2V6bFM2NnJEZUU/view?usp=sharing

 

Full cold start xdl, as promised. Forgot screenshots.

[RobertISaar]

Seems to me like the quad4 is just a bunch of (four) MOSFETs ordered by ECU logic, which would make erratic light behavior highly unlikely.

 

they're a little more complicated than that, but largely, yes. they also have overcurrent protection and fault sensing, all packed into a 15 pin IC.

 

they're tough to damage, so imagine if you return wiring to stock, it wouldn't occur again. it has been a while since I've looked at it, but I do believe a second relay is necessary to kick on a fan manually without setting the code.

[RobertISaar]

looking at the newest log: I don't know how long you had the key rolled forward before cranking the engine, but it looks like the primary O2 sensor was up to temp almost immediately after firing, it started producing believable results within 10 seconds. considering the engine was around 30*F when fired, I'd say the O2 looks like it is doing its job.