AFM testing with video

This 2nd time around is a good one. I know this one has helped loads of people definitively diagnose the problem with their Bosch Airflow meter (AFM) , and many of them have been able to repair theirs instead of shelling out a few Benjys for a reconditioned one. Like a lot of things similar to this I’ve seen, reconned units, especially electrical components, are often no better than the core you’re turning in, sometimes worse, so diagnosing and fixing the one you have is usually the better course of action. All you need to know is how to test it and what remedies you can apply to return it to normal function.

.

Vane-type AFM’s just like the Vanagon one are found in many many vehicles. They were the central component of all the iterations of Bosch L-jet, which first appeared in the late 70’s and was later named Digijet in the first 1983.5 waterboxer Vanagons when new digital fuel injection computers emerged. In the ’86 model year, Digijet the fuel injection system became Digifant the engine management system, where the engine computer took on spark control as well. It also became Digifant I and II in some other VW cars (so I sometimes refer to the waterboxer Digifant as “Digifant Zero” since they somehow forgot the normal convention of naming something’s second version “II”), and it appeared as other iterations of L-jet in lots of other German and Swedish cars. But the essence of these systems was the same, they all used a measurement of intake air mass as a proxy for engine load, and the AFM is the instrument to provide that signal. So this diagnostic is applicable to an awful lot of late-70’s thru 80’s European cars.

.

Here’s an iddy biddy viddy I took while testing an AFM that was giving some trouble. Symptoms were hesitation off-idle, and intermittent missing while under load in low-mid-range rpms (under 3200rpm).

.

The Bentley test confirms circuit integrity, but doesn’t tell you what you really want to know, which is that the ECU is receiving a 0-5V variable voltage signal that is smooth, uninterrupted, and increases steadily in relation to the air vane position. This test shows the actual signal as the ECU receives. By the way, this customer had already done the Bentley tests himself with an ohmmeter and found nothing that was outside specs. The reason being those tests will not show up the single most common problem these instruments develop, by far, which is wear on the carbon track exactly like what you’ll see in this one.

.

You want to use an analog meter; a digital one has a periodic refresh rate so it doesn’t do a good job of showing a steadily changing reading, when constancy and steadiness is what you need to confirm. Basically, it’s an analog signal, so it is best represented with an analog instrument. You could do the test with a digital meter, but you would want to move the vane very very slowly so you could see the parts where voltage drops out, if it does.

.

I removed the AFM from the filter box to have access to the inlet so I could deflect the air vane itself. This also obviates the need to remove the cover to do this test (although you will need to remove it to repair the meter if it does test as bad). On this particular AFM with the cover off, moving the wiper hub at top did not show the problem, while moving the air vane did, so even slight pressure on the wiper hub can alter the pressure of the wiper on the track enough to mask a problem, so you really want to take the two minutes to remove it from the air filter box so you can manipulate the air vane directly. Think about the $300 you’re saving.

.

After unbolting the AFM from the filter box, I plugged it back into the vehicle wiring, and switched on the ignition so the ECU would be feeding it’s regulated 5V output into the potentiometer. That voltage is reduced variably by the resistance of the wiper board, and what comes out is the 0-5V signal the ECU needs to calculate air mass. If you have a regulated 5V supply, you could bench test the AFM by inputting the 5V at pin 3, and I know some people have done this successfully with a 9V battery, the voltages will just be proportionally higher.

.

The voltage signal is on pin 2 (pins are 1-4 left to right when viewed from the rear of the van as installed; they are also numbered on the male connector flange of the AFM). Pin 4 is grounded inside the ECU, so you want to test V between 2 and 4. Use a low range on the meter; I have this meter on a 2.5V range. Even though the actual reading across the entire scale will be from just above zero to about 4.5V, I’m most interested in seeing the lower half of the range because that’s where the wiper spends most of its time, so that’s where the wear will occur. At idle the signal will normally be around 1V, so at every opening of the throttle it has to sweep thru the area just above one volt. Consequently that is where the most wear will be found, and that is the range that will cause the hesitation on tip-in and load hesitation that this customer complained of.

.

.

An AFM that shows dropouts can usually be repaired. First remove the cover, and then I first try spraying the wiper track with electrical contact cleaner while moving the wiper across it a bit to also clean the wiper tips. Then test again.

.

If cleaning doesn’t restore smooth signal, there are two ways to reposition the wiper to run on a different radius of the track. Simplest would to loosen the screws that hold the wiper board down and slide the board toward or away from the wiper hub. In this AFM I couldn’t get the screws to break free (heat may help soften whatever adhesive was used on them) , and most folks also report that the screws are very hard to get loose. So I used the second choice solution. I used a bent scribe and a small screwdriver to make a slight double bend in both sides of the copper wiper arm itself, just enough to move the wiper tips a tiny bit closer to the hub, so it will run on a fresh line on the track. This restored smooth signal output in a subsequent test, and the van now runs smooth and the customer is very happy. If you bend the wiper, take care that the pressure of the wiper points on the track is still very light so it doesn’t cause undue wear, your testing will confirm you still have good contact thru the full range, so just enough pressure is enough!

.

The other test to be done is the Intake Air Temp sensor, which is just an exposed NTC thermistor wired between pins 1 and 4. An ohmmeter and knowing the ambient temperature are all that’s needed to check that, according to the temp/resistance table in Bentley. The IAT is very rarely bad, whereas the AFMs physically wear so after 20 years of use it’s not at all uncommon to find them in this kind of condition.

.

The test will be exactly the same with both 2.1 and 1.9 WBX AFM’s in Vanagons. It can also be done on other L-jet injections, but the pin positions will vary. One person who experimented said the signal output on late 70’s VW bus L-jets was pin 7, ground was pin 6. I don’t have one of those or the manual to verify. If you test one of those with the AFM connected as I’ve done here, you’ll find that those were built before the industry convention to run sensors at 5V, so what you’ll see is a 0-12V range instead, so your idle range will be around 2-2.5V, and so on proportionally.

4 Comments

  1. Tom Oconnor

    Chris,

    How does any lover of old vanagons spend his saturday bfore leaving on a trip? Well of course one way is cleaning and re adjusting his Afm. I watched your video several times tested the afm and carefully adjusted the contacts within. Problem solved. Thanks for the info, it was a good Saturday.

    Tom

    Reply
    • Chris

      Good deal, glad it helped!

      Reply
  2. Jonathan Harnum

    Thanks so much for the precision explanation. Would love to see an “after” shot of the crimp you mentioned performing to help me from destroying the swing arm. Might try anyway, lol. Super-helpful, thanks again!

    Reply
  3. Samy

    Excellent info
    Thanks for your time and helpfully explanation
    Great jop
    Thanks again

    Reply

Submit a Comment

Your email address will not be published. Required fields are marked *