Newly rebuilt M20 AFM - $70 (Standard MF20040) NOT a forsale thread

[Gregs///M]

You're the electrical engineer that wants to charge kids $120 to bend a wiper track and hook up an AFM to an o-scope, you should already know the answer.

The AFM testing procedure I linked works just fine using a DVM.

I'm afraid you are the one who incorrect. But hey, here is your opportunity to prove me wrong in front of everyone in this forum.

A DVM does not nearly acquire voltage signal as fast as a oscilloscope. Let me explain this the best that I can.

Let's say we want to test for bad AFM. Customer reports they have severe "bogging" at idle speeds. So let's test with a DVM vs an oscilloscope.

DVM:
Measure voltage output while the wiper arm is moved from fully closed to half open. The DVM does not have the resolution of an oscilloscope so its averages the voltage over the time acquired. BTW, in general, DVM acquire voltage using three methods...max volts, minimum volts, and average volts. The DVM is going to provide you an average of the voltage acquired.

Therefore, with a DVM, the voltage feedback would appear to be within spec and you pass the AFM and return to the customer. Customer spends hundreds of dollars replacing other items and troubleshooting, etc.

Oscilloscope:
Measuring voltage output while the wiper arm is moved from fully closed to half open and displays this voltage difference over a time scale of say...10 seconds.

This graphs shows myself opening and closing the AFM from 0% to 10% nine times in a period of 10 seconds.

Notice the "dips" to zero volts. This is known as voltage drop out and is the reason why there is a sudden loss of power near a specific RPM range, which correlates to a certain speed range or air input to the AFM.

Also notice on the second curve that voltage drop out occurs near the beginning of travel and near the end of travel. It happens several times more over the period of 10 seconds.

But note: these drops to zero occur and only last a few milliseconds and even microseconds! A DVM CANNOT ACQUIRE NOR DOES IT HAVE THE RESOLUTION TO CAPTURE THESE MILLISECOND DROPS TO ZERO VOLTS.

I would fail this AFM and the customer would rest assured that he has found the problem to the solution.

So please, anyone reading this. Please understand a DVM is not sufficient to test a AFM. At bare minimum, a oscilloscope or data acquisition software must be used which has a resolution of microseconds or better. My oscilloscope can acquire data at a rate of Pico-seconds so it is more than sufficient to test accurately.


ForcedFirebird, you're welcome to dispute my research and findings. Please enlighten us.

[ForcedFirebird]

I see the same thing when emulating an eeprom in the ECU as well, during data tracing.

Perhaps I should have said DMM not DVM, Fluke, whatever you want to call it. You are posting a pic of a worn wiper track, ok, but the wiper arm can just be bent.

I see your finding, and don't disagree that you can show a graph, but to open an AFM, bend the arm and verify an AFM is working. Do you take them completely apart and soda blast the cases, check door spring tension, anything other than moving the wiper and testing with your service?

[Gregs///M]

I see the same thing when emulating an eeprom in the ECU as well, during data tracing.

Perhaps I should have said DMM not DVM, Fluke, whatever you want to call it. You are posting a pic of a worn wiper track, ok, but the wiper arm can just be bent.

I see your finding, and don't disagree that you can show a graph, but to open an AFM, bend the arm and verify an AFM is working. Do you take them completely apart and soda blast the cases, check door spring tension, anything other than moving the wiper and testing with your service?

DMM=digital multi meter
DVM=digital volt meter

Same thing and these two terms are used interchangeably in the industry.

Have you even read my ad on AFM rebuilds? Have you even read this thread?

Again my process is this:
1. Test intake air temperature resistance over the heat range via Bentley repair manual.
2. Test spring tension vs. a known recently calibrated AFM using Newton scale meter. (Fuel Injection Corp. unit)
3. Disassemble wiper arm and connector assembly. Clean incoming, and outgoing contacts and bend them to contact fresh material. Clean wiper arm contacts and primary grounding post. Adjust wiper arm to contact fresh material.
4. Reassemble wiper arm and connector.
5. Measure voltage output with oscilloscope and adjust if necessary. I run a variety of tests including quality repeatability, linearity testing (or non linear for eta AFM's,) quality repeatability in the idle genre and the fully open genre.
6. I provide my customers with physical photos of the entire process and my oscilloscope print outs of their individual AFM being tested.

I do not perform any cleaning or sand blasting of the AFM casing. However, I am working with a forum vendor to begin these services soon. Unfortunately, my AFM's might arrive a little dirty but electrically they are 100% working again. I've had zero negative feedback and only 100% positive feedback in the 20+ units that I've sold this year. I have yet to hear any negative feedback on my AFM refurbs.

Again let me state...Do not trust ForcedFirebird or anyone else's advice of using a DVM or DMM to test AFM's. I think this thread clearly identifies the difference and reason for my statement.

Additionally ForcedFirebird, you have failed to prove my previous post incorrect or inaccurate. So either I am right or you are right, if so, we are waiting for your proof.

[Gregs///M]

Here is another bad AFM that I tested for a customer.


Notice now the voltage spans from 0 volts to almost 5 volts. This is 14 cycles from fully closed to fully open. Noticed the short duration drops to zero. This occurs in the microseconds and cannot be captured by a DVM/DMM. Only an oscillocope or data acquisition software will have the resolution to capture these drops to zero.

A DVM/DMM would only show the average as you move the wiper arm from fully closed to fully open. It would appear the AFM is good, when indeed it is bad.

[E30-325iS]

At what rate does the ECU sample the AFM signal?

[Gregs///M]

They interchange. The Motronic 1.0 uses 5v, 1.3 uses 12v and they are a direct swap for each other (153 ECU is plug and play for a 173).

I have been hacking the OBD1 ECU's for years.

This is another incorrect statement regarding Air Fuel Meters. Feeding a 12V reference into a Motronic AFM would damage the AFM, ECM, or both.

Motronic 1.0 systems utilize a reference voltage of 12V DC.
Motronic 1.1 - 1.3 utilize a reference voltage of a highly regulated 5V DC.

This is a result of my findings through AFM research from the many that I've tested over time.

[ForcedFirebird]

If you reread my post, you can clearly see that there is a typo. 153 ECU is M1.1 and 173 ECU is M1.3. The 173 has 12v reference and 153 has 5v.

[Gregs///M]

If you reread my post, you can clearly see that there is a typo. 153 ECU is M1.1 and 173 ECU is M1.3. The 173 has 12v reference and 153 has 5v.

Again, the 173, M1.3 is 5V referenced. Not 12V. You cannot feed 12 volts into the DME for AFM signal or it will be damaged. Same with M1.1

[delatlanta1281]

Why would you buy a rebuilt afm, or even a new afm considering how cheap and reliable a maf swap has become?????

[Gregs///M]

Why would you buy a rebuilt afm, or even a new afm considering how cheap and reliable a maf swap has become?????

Many people, including myself, prefer to keep the car in original condition. I don't need/want extra air or performance. Plus an AFM is a direct swap and does not require modification to install.

Also can a MAF even work on the OEM air box? Or will a cone filter be required. If that is the case, owners will not pass the visual smog test, esp here in CA.

[ForcedFirebird]

Again, the 173, M1.3 is 5V referenced. Not 12V. You cannot feed 12 volts into the DME for AFM signal or it will be damaged. Same with M1.1

Hmm, it's only the TPS that changed reference voltage. You are correct on the AFM voltage.

Just so happens we have had 11 e30's at the shop today, just checked it out - down to 6 now actually after a few pick-ups, the rest will be here a little longer...

[Gregs///M]

"Before testing the switch, check that the switch is receiving
voltage. With the ignition on, there should be at least 5 volts
(VOG) between the center terminal and either of the outer
terminals of the switch harness • connector. It voltage is. not
present, test the electronic control unit as described under 5.4
Electrical Tests."

As per the Bentley manual. 5V reference voltage for the TPS. Same with the AFM. This applies to most, if not all, E30's.

ForcedFirebird, you have too many errors and typos. It is my suggestion that you no longer post anymore into this thread. There will be thousands of people who will read this thread in the future, and those people will need to rely on reliable and accurate information.

You and Noid have accused me of being incorrect and questioned my knowledge on AFM operation without having any professional background in electronics technology. I have thoroughly proved by research and findings. In addition I have provided the public advanced knowledge required to perform/test their own AFM's and posted some data that they can reference.

I don't care how many articles and videos are out there detailing AFM testing, I have clearly explained that most of those are inaccurate and incorrect. The only reference information that should be trusted is mine.

http://www.r3vlimited.com/board/show...83#post3152483

[ForcedFirebird]

I don't care what Bently says. I just put a DVM on a TPS switch on a 153 and 173 ECU's. Bently also says that the CPS is 1200ohms and it should be 580.

Not trying accuse you of being wrong, I was expressing that people can test their own AFM with the methods that have been used for decades. I just hook up an ECU with an eeprom adapter and an emulator and can read the feedback from the sensors as a diagnostic - but not everyone has that equipment and can use the FRWilk procedure and save $120.

[Gregs///M]

I don't care what Bently says. I just put a DVM on a TPS switch on a 153 and 173 ECU's. Bently also says that the CPS is 1200ohms and it should be 580.

Not trying accuse you of being wrong, I was expressing that people can test their own AFM with the methods that have been used for decades. I just hook up an ECU with an eeprom adapter and an emulator and can read the feedback from the sensors as a diagnostic - but not everyone has that equipment and can use the FRWilk procedure and save $120.

The first line in your post pretty much sums it up. Bentley wrote that manual in cooperation with BMW engineering, you should trust it. Perhaps you should unplug the TPS before testing its resistance or you will be reading an impedance figure which includes the ECM.

Just because there is an article or video on the "internet" does not mean it is 100% truth. Do not trust everything you read on the "internet."

Im sorry to tell you but those methods that people have been using for decades, from that Porsche site, are wrong. No DVM, No DMM, oscilloscope only.

[ForcedFirebird]

Ok, go put a CPS in an e30 with 1200ohms and see what happens. I don't trust the internet or "how to" books, I trust experience. Bently also says to torque the valve cover bolts to a torque that will snap the studs off. If you really were an e30 enthusiast and not here trying to make money on them, you would know about the Bently mistakes.