Autonerdz forums

Ok, so here's another set of microamps captures.

These are Gm MAF sensors.  One is a 00 Silverado 6.0 and for sure is a hot wire sensor.

The other is a 96 Olds 3.1.  Information is hard to get, but we believe that is is a hot film sensor.  We were unable to verify from any good sources, but according to "Advance Auto Parts" is it a hot film.

We are not trying to say we understand exactly what we are looking at, we do not yet.  But it is a way of looking at stuff that is different, and perhaps as a group we can make some sense of this!

All captures are the same as far as channel layout

A:  MAF volts (normal scope reading)
B:  MAF amps
C:  MAF frequency, taken from Channel A
D:  TPS volts for reference

As before, we are going to break this up so it fits.

These are the silverado ones.

First is an overview of the entire screen.

Second is a zoom that shows how closely the frequency of the sensor follows the amp draw of the sensor.

Third is a real tight zoom that shows that the amps and frequency do indeed peak at the same time.

(more to follow)

John and Mike

Now here's the part they we don't get.

The capture of the truck shows that the maf frequency and amps peak at exactly the same time.  It is for sure a hot wire sensor.

The capture from the Olds shows something different.  The MAF amps peak about 6 milliseconds before the frequency peaks.  This stays consistent across the capture, 6 to 7 milliseconds every time.  We think it is a hot film sensor.

Why?  this is NOT a test!  We do NOT have the answer yet, hopefully some one out there knows!

Does this show the sensor take 6 msec to respond to changes?  Is that a function of the hot film instead of the hot wire?  That's our only theory.

Obviously the hook up and microamp tester are working well, the truck pattern matches.

The next question is amp draw.  There is no spot where the 2 captures have exactly the same frequency, but they are close.  The truck at 6.8 khz draws 346 milliamps.  The car at 6.7 khz draw 152 milliamps.

So in round numbers the truck takes twice the current.  Is this a function of one being hot wire and one hot film?  Or is this a function of displacement?  Twice the displacement takes twice the current?

We don't know! 

These 2 vehicles were chosen simply because they are in my garage.  My wife's car has had so many holes poked in the wiring harness from our experiments it is a wonder it even runs anymore!

Thanks

John and Mike

John and Mike

That was a lot of work on a weekend.

A big thank you for your efforts.

The Pico software is very accurate and I would have to guess that it’s a displacement issue about the Amps draw, but I am still thinking about the 6ms of difference.

Mick

Pedro

Thanks

I know the truck is hot wire, it's in the data.

But no data source we have describes the olds maf.  Now we have one parts store saying one thing, and another saying the opposite.  I could not find what style it was in the napa listing, but i logged on through the regular site.

Good guess on the car.  3.1 vin M  except with 5 kids it's a station wagon!

This is the style that mounts directly on the throttle body, with three terminals marked A, B, and C.  + and - are over B and C (forgot which)

We know both styles exist, but until now never had to tell which one was which.

If both are hot wire, then we are even more lost!  That was our only theory!

Got hold of a hot wire frequency MAF here today.  91 Ciera 3.3.

A: MAF frequency graph
B: MAF voltage
C: MAF current
D: TPS

The MAF frequency graph is so clean because I am using a slightly hacked version of PicoScope.  This improvement is in the works for the next release.

Noted that the frequency was dead on correct with the voltage frequency measurement.  Also noted when adding 2x vertical zoom to C and overlaying the current and frequency graphs that they followed each other almost exactly.

We will be exploring this on other MAF designs as the opportunity presents itself.

Yes, interesting.  Like this?

Quote:
i dont mean to change the subject but did that car have a slightly rough idle, like a random lean misfire every so many  seconds when above 200 degrees? i have one of these and every one i've seen has the same idle issue, usually on the #5 and #3 cylinder?

Yes, they were rough. I think that was due to valve overlap and no EGR. The 3.8 had this problem too before it was totally redesigned with a balance shaft in about 88 or 89. I remember try to fix both of these engines when they were brand new. Couldn't do it. We were installing different balancers and that helped a little. (I just know Jim recalls this) Then GM came out with different engine mounts. Sorta funny really because this didn't fix the imbalance problem, it just isolated it from the driver. That engine would still shake, you just couldn't feel it as much.

Problem was solved with the 3800. That has been a pretty good engine although we did have to correct a few that had the balance shaft walking a little.

Great topic this guys!

I am so pleased that this feature is finally being worked on. I have been emailing Peter (French) @ pico for a while now & it is good to see it in action. The "cleaned up" signal that you posted Tom looks really good & I hope it isn't too long before we see the improvements in a new software release.

Are there any plans to introduce the same graphing feature, but for PWM (Duty cycle) signals? The benefit of being able to study these values along side other output & input signals is going to be very useful.

If there was one thing that Snap-on had in their arsenal that I really wished Pico would pick up on, it is this graphing function.

Well Done Pico (& Autonerdz for your undoubted contribution   )

Cheers all

PS. If anyone is clever enough..... I would love an "APP" for my iPad or iPhone, so that I could open Pico files, when I am away from my PC/Laptop. There seems to be an "App" for everything else!  

Dave

John,

I think you found a bug in the software. You're right, the frequency does follow the analog MAF voltage. That isn't right.

The only thing I can think of is... although your scaling does indicate Hz, I believe you are still in a voltage range. Try choosing a frequency scale instead of "Auto" and see what that does. Your frequency really isn't changing that much. Maybe it is due to all the noise. I'm not sure.

I'd send that to Pico and have them take a look at it. Or maybe Mark will read this and respond. I'd still send the .psdata file though. I'm sure they'll take care of it.

Regards,

At this time, files saved using frequency do not open with the correct probe settings for the frequency channel.  The behavior is different with hardware or in demo mode but neither way correct.  The channel scaling on the screen opens correctly though.

Some of the images I posted show the correct probe settings only because I cheated and changed them before snapping the image.