View original topic: Vanagon (Digifant) Air Flow Meter (AFM) Adjustment Procedure Page: 1, 2, 3, 4, 5  Next
Quixotix Fri May 08, 2009 11:08 am

Below is a procedure I cobbled together for adjusting the Air Flow Meter (aka – AFM, Intake Air Sensor, Air Flow Sensor, etc.) on my 1987 Vanagon.

First, let me make it very clear that I DON’T THINK YOU SHOULD USE THIS PROCEDURE. Whatever the problem is that you are having with your Vanagon, it is probably something else causing it, not an out-of-adjustment AFM (my problem was an intermittent connection to the TEMP II sensor). Check everything else first! At least do all the continuity checks in the chart on pages 24.61 and 24.62 of the Bentley manual (See Foot Note 1).

But, since like me, you are probably going to ignore others good advice and mess with the AFM anyway, hopefully the following procedure will help to keep your car running even after you work on it. :)

Why I’m I qualified to write this procedure? I’m not! Well, I did learn to drive in a VW van. And, as a teenager, I did successfully take the box of disassembled engine parts (including leftover parts from previous engine rebuilds) that my dad had in the back of the garage and put together a working engine for one of our VW buses. BUT, these were Type I buses. There’s not exactly a lot of computer control in those things.

So, why am I posting this procedure? Well, I couldn’t find any procedures at all for adjusting the AFM (in books or on the web), and this procedure seemed to work for me. Mainly though, I’m a mechanical engineer, so I think I know everything – Just ask my wife! :wink:

Just in case I’m wrong about knowing everything (it seems logical that some day I might actually be wrong about something :oops: ) I would appreciate any comments, corrections, or feedback on this procedure. Thanks.


The AFM has 3 things you can adjust on it: 1, the idle mixture screw. 2, the position of the wiper that runs on the carbon track. And 3, the return spring tension for the flapper vane. This procedure will also have you adjust the idle speed screw on the throttle valve.

The only special tools you will need for this procedure are a multimeter to measure voltages between 0 and 1 volt and a tachometer to check idle speed (if your car doesn’t have a tach on the dash). You will use the multimeter to measure the voltage produced by the oxygen sensor (also called a lambda sensor). SO, OBVIOUSLY, YOU MUST HAVE A GOOD O2 SENSOR TO USE THIS PROCEDURE. The voltage of the O2 sensor tells you if the engine is running rich or lean. Note in the graph (from Bosch Fuel Injection & Engine Management by Probst) how quickly the voltage goes from about 0.8 volts to under 0.2 volts. Because this change is so abrupt, the O2 sensor does NOT do a good job of telling you HOW MUCH to rich or HOW MUCH to lean the engine is running. It really only tells you that the mixture is somewhere on the rich side or somewhere on the lean side of ideal. An exhaust gas analyzer would be nice to have as it can give you more detail about how far off the mixture is, but using the O2 sensor is adequate for this procedure.

Adjusting the idle mixture screw is described in the Bentley manual on Page 24.51. I’ll give you an alternate procedure here that doesn’t need a CO meter. To get to this adjustment screw, you must remove a round plug from the AFM. The plug is about ½ inch in diameter and clearly visible on the top of the AFM body. It is right next to where the big rubber boot connects to the outlet of the AFM. As per the Bently manual, you may need to drill a small hole in the plug and then screw a sheet metal screw into this hole to pull the plug out with.

To get to the other 2 adjustments, you need to remove the square plastic cover from the top of the AFM. It is about 3 inches square as seen from above, and wedge shaped when viewed from the side. To remove this cover, carefully pry it off. It is held on with silicone sealant. Work your way all the way around the cover, trying to cut the silicone as you go.

CHECK THAT THE VANE IN THE AFM MOVES FREELY: With the plastic cover off the AFM, grab the little metal clamp opposite the wipers and swing the wiper back and forth. If it doesn’t move smoothly through it’s entire range, then the AFM must be replaced.

BEFORE MAKING ANY ADJUSTMENTS, I really, really recommend that you record the starting position of all 3 adjustments. This way, when your car runs worse after following my stupid procedure, you can put everything back where it was:

COUNT THE NUMBER OF TURNS THAT THE IDLE MIXTURE SCREW IS OPEN: Put an Allen wrench in the idle mixture screw, and turn it clockwise (CW) until it is closed. Write the number of turns down somewhere! Now open it back up counterclockwise (CCW) to its original position.

MARK THE POSITION OF THE WIPER; Use a felt tip pen to mark the position of the wiper on the circuit board. I recommend you mark it somewhere other than on the black carbon tracks.

MARK THE POSITION OF THE GEAR TOOTHED WHEEL. Use a pen or scratch-all to mark one tooth of the gear and the location of this tooth on the aluminum body.

IF THE CARBON TRACES ON THE CIRCUIT BOARD ARE WORN: You can loosen the 3 screws that hold the circuit board and then move the circuit board a little so that the wipers run on a new part of the carbon traces. See Foot Note 3.


1) The engine MUST be at full operating temperature before you make any adjustments. See Foot Note 2.

2) Disconnect the electrical connector from the idle control valve (this is the hotdog sized thing located on the top of the engine right in the middle of the engine compartment). Keep the idle stabilizer control unit (idle control module), located forward of the right taillight, connected.

3) Disconnect the single green wire from the oxygen sensor. Connect the positive lead of your multimeter to this wire (connect it to the side that goes to the sensor, not the side that goes to the computer). Leave the other 2 wires for the O2 sensor (the O2 sensor heater wires) connected. Connect the negative lead of your multimeter to a good ground point on the car. Set the multimeter to a setting that will read well in the 0 to1 volt DC range.


Start the engine. You should have already warmed it up, but if it not make sure you warm it all the way up before you adjust anything.

NOTE: All of the adjustments below (idle RPM, Idle mixture, wiper position, and spring tension) tend to interact with each other. You will need to go around and around through the various adjustments until you get them all to fall in the right range.

Adjust the idle speed to 880+/-50 rpm (Bentley’s page 24.51). Use the big slotted screw (plug) on the throttle body (not on the AFM) to adjust the idle speed. DON’T use the small stop screw on the throttle arm to adjust the idle speed. This small stop screw should only be adjusted per the procedure in Bentley’s (page 24.37). If you cannot get the idle speed down low enough, then your throttle body is probably worn. Repair or replace the throttle body before adjusting the AFM.


Rev the engine a little and then let it drop back to idle. Now watch the multimeter reading for a few seconds. Adjust the Idle Mixture Screw until the voltage stays between 0.5 and 0.8 volts. If the voltage is too low, then turn the screw CW. If the voltage is too high, turn the screw CCW. Note, this screw may need to be adjusted a number of turns. If the idle RPM changes, then alternately adjust the idle speed and the idle mixture until both of them are in the correct range.

If you cannot get the voltage in the correct range, then you will need to adjust the wiper position or the spring tension as described next.


Rev the engine up to about 1100 (high enough to open the idle position switch on the throttle body) and hold it there for a couple of seconds. When the voltage stabilizes on the multimeter, note what it is. Now rev the engine to around 2000 RPM; hold it there and read the voltage again. Do the same at around 3000 RPM. You can do 4000 RPM too if you want, but if you don’t like to rev it that high then you don’t need to. Bring it back to idle and let the voltage stabilize. If at all RPM’s, the voltage always stabilizes between 0.1 and 0.8 volts, then no more adjustments are needed.

If the voltage always stabilizes ABOVE 0.8 volts - OR BELOW 0.1 volts - then adjust the wiper position. To do this, loosen the single Philips screw in the slot in the black plastic part that holds the wipers. If the voltage was always too high, rotate the wipers CW (move the wipers closer to the closed position on the carbon traces). If the voltage was always too low, rotate the wipers CCW (move the wipers closer to the full open position on the carbon traces). MOVE THE WIPERS JUST A TINY BIT AT A TIME. This adjustment is very sensitive.

If the voltage was high (over 0.8 volts) at higher RPM’s and low (under 0.1 volts) at low RPM then rotate the toothed plastic wheel CW. If the voltage was low at higher RPM’s and high at low RPM then rotate the toothed plastic wheel CCW. To rotate the wheel, push the end of the wire clip out of the tooth and turn the wheel with your fingers or gently with a tool. I recommend that you move the wheel only 1 or 2 teeth at a time.

Adjusting either the wiper position or the spring tension will definitely change the idle mixture. And if you adjust the spring tension wheel CW you will most likely need to adjust the wiper position CCW (or visa versa) to compensate (See Foot Note 4). So, Start back at the idle mixture adjustment, and keep going through all the adjustments until all the tests give the correct results.

NOTE: If you can get the correct voltages at most of the RPM’s but not all (for example, you might be OK at idle, 1100, and 3000, but 2000 RPM might read 0 volts) then you MIGHT have a bad spot in the carbon traces. OR, my procedure might be no good because I’m full of $#!+. However, I would try hard to get a good voltage reading at idle, and at at least one RPM above idle. If you can do this, the oxygen sensor (when reconnected) should allow the computer to adjust the fuel/air mixture correctly.


1 – Here are some things that might help you do the tests in the chart on pages 24.61 and 24.62 of the Bentley manual. When the chart says “bridge”, it means connect a jumper wire between the listed terminals. In step 6, the reference should be page 24.45 (in my book, no page number is listed). Although the graph on page 24.45 looks like two graphs, it is really just one with the ohm scale expanded for higher temperatures. The same graph is used for both the Temp II sensor (page 24.62, step 6) and the intake air sensor (page 24.62, step 9). In step 8, if you have one in your car, I think you must REMOVE the pigtail wire installed at the AFM that is there to fix the “Vanagon Syndrom”. After you remove the pigtail, reconnect the standard electrical connector to the AFM. With this pigtail installed, I couldn’t read the AFM wiper resistance.

2 – For the adjustment procedure above to work at all, the engine must be at operating temperature, AND the engine control unit (computer) MUST know that the engine is at operating temperature. The computer knows this by reading the Temp II sensor (this is the temperature sender on the side of the thermostat housing). I highly recommend you warm the engine up, then turn it off and disconnect the big electrical connector from the computer. Check the resistance at this connector (refer to pages 24.62 and 24.45 in Bentley’s). Make sure the resistance between pins 6 and 10 is between 200 and 400 ohms. If the resistance is higher than 400 ohms, then either the engine is not fully warm, or worse, the computer thinks the engine is cold when it is not. If the resistance is under 200 ohms, then the computer is getting the signal that the engine is hot. Once you finish making the checks at the computer’s connector, reconnect the computer.

3 - If you measure the resistance from pin 2 to pin 3 of the AFM as described in Bentley’s, page 24.57, the resistance does NOT change smoothly up or down as you swing the air flow vane. Don’t let weird resistance readings make you think the carbon traces are worn. BUT, if you connect a volt meter from pin 2 to pin 4, turn on the ignition (with the AFM connectors all connected) and then swing the air vane, you should see a steady increase in voltage as you open the air vane. It will change from about 1 volt when closed to 4 to 5 volts when all the way open.

The 3 screws that hold the circuit board in the AFM may have lock-tight on them and be hard to turn. Make sure you have a screwdriver that fits well and push down hard on the screwdriver as you turn to keep from stripping the screw head.

4 – At first glance, adjusting the wiper position and adjusting the spring tension may seem to do the same thing. That is, moving either one CCW will make the engine run richer (increase the O2 sensor voltage) across the entire airflow range. While this is true, there is a subtle difference between the two adjustments. Adjusting the wiper position makes the same change in wiper angle across the entire range of air flows. But, adjusting the spring tension will have a bigger effect on wiper angle during low airflow conditions than it does during high airflow conditions. Thus, only the spring tension adjustment can be used to correct differences in the mixture ratio (voltage) at different airflow rates. Note: Airflow rate changes with RPM and with engine load. But, since engine load is harder to change than RPM when the car is parked, I ask you to change only the RPM to get different airflow rates.

stevey88 Fri May 08, 2009 4:20 pm

Thanks for taking the time to write the procedure. I have a question concerning the O2 sensor. As you disconnect the connection to the ECU, the ECU will see 0V at its input and think the mixture is too lean. It will then try to make it richer, more injector open time I guess, until it max out at its software preset ?
The second question is concerning the idle stabilizer. Since you disconnect it for the adjustment for the idle speed, wouldn't it become impossible to lower the idle speed by the idle stabilizer control module after all is done and the connection re-established ?

How about applying a fixed 0.4V to the O2 input of the ECU so it think it is in range and not trying to adjust the mixture ? What if we jam the idle valve open a bit so the idle module has enough range to lower the rpm after AFM adjustment ?

Quixotix Fri May 08, 2009 5:13 pm

Those are good questions stevey88. I can't tell you what the logic is in the ECU (boy would it be nice to have the engineering logic diagrams for that thing).

But, having the oxygen sensor and idle stabilizer valve disconnected comes from Bentley's. These are required conditions for adjusting the idle mixture and basic idle speed per page 24.51. And, the 880 +/-50 RPM comes straight from the same page.

I think you are correct that adjusting the basic idle RPM to 880 will mean that the idle will never drop below that (assuming no increase in engine load). I assume this is what VW wanted.

Regarding the zero volt reading with the O2 sensor disconnected: Since I got this from Bentley's, I have to assume that the ECU has logic to recognize that the sensor is disconnected. It must either detect an open circuit or see a true zero reading (vs. say a .01 V reading) and know that this is not a just a lean situation. That's my guess anyway. Or, you could try applying a .4 volt input and see if it works better. :) Just don't blame me if it fries your ECU. :shock:

stevey88 Sun May 10, 2009 7:48 pm

Here is my speculation :

As the idle/full throttle switch is on and rpm is low, the ecu knows the engine is at idle and will not look at the O2 sensor. It will just use the AFM output voltage to look up the map setting for the injectors. This will not be true if the throttle is off idle.

It is not practical to measure low voltages at this noisy environment so I do not think the ECU will know the O2 is disconnected. Having said that, Some member did report their engine run better with the O2 wire disconnected.

Dogpilot Sun May 10, 2009 8:51 pm

The ECU does look at the O2 at idle. It does sense the absence of the signal and go to open loop mode. It uses a map and does not adjust the fuel injector pulses/duration to adjust mixture. if you don't disconnect the O2 the ECU will constantly adjust the mixture as you adjust the AFM wiper, effectively nulling your actions.

The engine will run better if the O2 is disconnected if it is shorted or at the end of its life. It runs on its map, so it will be better than trying to cope with the Garbage in mode.

For fun, hook an Oscilloscope to the O2 output, it is not a linear value, but rather a sawtooth waveform.

Check it out:

Great job on the write-up!

Quixotix Mon May 11, 2009 11:49 am

Thanks Dogpilot.

Just one thing though to avoid someone from getting the wrong impression.

The "sawtooth" output from the O2 sensor ONLY occurs when it is connected to the ECU. That is, even the ECU can't control the mixture accurately enough to stay right at the desired 0.45 Volts. So the ECU makes it richer until it sees a higher voltage, then it makes it leaner until the voltage drops. It keeps doing this giving the sawtooth pattern.

With the O2 sensor disconnected, you won't get the sawtooth voltage output.

PS - I almost included a graph similar to the one you referenced in my original post. But I cut it out hoping to avoid the "They blinded me with science" problem. Too bad I wasn't successful. :wink:

Dogpilot Mon May 11, 2009 12:06 pm

The truth will set you free. Include as much science as you can. Yes I suppose I should have pointed out that the Sawtooth is when hooked up to the ECU. It took me a while to realize what was happening since there is, as you aptly pointed out, a lack of a logic diagram. So explaining anything to do with the ECU becomes kind of a black box issue, we see effect and cause, but not the decision matrix. It is kind of why I made the pass through version of the Digitool. I wanted to be able to observe the signal, many of which are not best interpreted as straight voltages, but rather waveforms. I also wanted to play with a few en-route to the ECU, modifying them for my own diabolical purposes. I did learn some, not nearly enough, to satisfy my curiosity.

I am glad to see others share the bug to delve into secrets and clarify them.

iceracer Mon May 11, 2009 12:42 pm

This is awesome! ANd it works. I did it last night on my 86 westy refreshed motor that has been giving me hesitation issues. It worked and I didn't have to get into the AFM adjustment. Just needed some tweeking of the idle screw and the AFM screw. And now it runs great and no more sluggishness. Way to go.

dbeierl Fri Oct 16, 2009 3:48 pm

stevey88 wrote: I have a question concerning the O2 sensor. As you disconnect the connection to the ECU, the ECU will see 0V at its input and think the mixture is too lean. It will then try to make it richer, more injector open time I guess, until it max out at its software preset ?
No, the ECU biases the line with just under half a volt, at a high enough impedance (that is to say, a very tiny current limit) that the sensor will overpower it when it heats up and comes on line. So disconnecting the sensor will leave the ECU looking at its own bias voltage, which will tell it to stay in open-loop (no O2 sensor) operation.

Quote: The second question is concerning the idle stabilizer. Since you disconnect it for the adjustment for the idle speed, wouldn't it become impossible to lower the idle speed by the idle stabilizer control module after all is done and the connection re-established ?
Yes, it is impossible. That's ok, there's no need for the idle stab to lower the rpm, only to supply additional air to keep idle rpm constant with varying loads from steering, alternator etc.

peterkissam Wed Oct 21, 2009 10:33 pm

I think there is a solution to your problem if you are willing to buy a Wideband O2 sensor.

Look at this post down-page by "tencentlife" about mixture control and his setup for 2 modes of operation to solve the "sawtooth issue" of OEM O2 sensor. Wide Open and Cruise. Conditional input.
His sounds like the answer to "tighten up" the EFI operations of Digifant.

Concurrently, I've grafted a GM TPS sensor in place of the wiper/resistor board of the flapper in AFM. I've got it smoothly outputting 0-4.8 volts and am hoping to try it in place of stock potentiometer/voltage divider circuit. I hope to hang the GM TPS sensor on all the same wires (Ground#6, +5volt#17, var.AFM 0-5v #21)
It's the impedance of this circuit with AirTemp and voltage divider of AFM pot that has me perplex (basically unsure if...)
I don't want to "brick my ECU. I'm right there...

Anyway, go here for the answer...Page one, 16th posting by "tencentlife"

sbclayton Thu Nov 19, 2009 5:52 pm

Reviving this thread to point out something.

I used Quixotix's procedure to adjust my AFM, and it worked pretty well; the hesitation issue disappeared and I got better fuel mileage. THEN, along the lines of "I am so stupid", I tried tweaking the AFM for even better performance. BAD MOVE!

When I couldn't get the correct O2 Sensor voltage with the Idle Mixture Screw, I returned the IMS to it's original position and moved the AFM wiper arm - that's when everything went really wrong. At one point, I couldn't even get the engine to stay running long enough to make adjustments.

I finally realized something - the wiper arm position tells the ECU what load the engine is under, and that affects the ignition timing! I had fallen off the ECU's map! (Dogpilot's reply, 13th post down in this topic: I left the wiper arm at its best-running position and re-timed the ignition with tencentlife's procedure, found in this post - The engine stayed running well enough that, using Quixotix's tuning procedure, tencentlife's timing procedure, and working between the throttle body idle speed plug, the wiper arm and the timing, in that order, yes - I got a smooth-running engine and also got that extra bit of response and mileage I was originally looking for. Would I do it again? Yep - now that I've already been through the meatgrinder. :oops:

Bottom line: if you move the wiper arm, or change the spring tension, be sure to redo your timing, and recheck all your engine adjustments - they're all interconnected!

Also, a great tip: when moving the wiper arm over the carbon track, voltages were all over the place. I moved the board - same thing. Light came on :idea: - I *gently* slid a torn-off piece of printer paper under the contacts and moved the wipe arm - black tracks and black dust everywhere! When I removed the paper and read the voltages again, everything was nice and smooth.

Mzwo Fri Nov 20, 2009 2:30 am


forgive me if this isn't entirely relevant.

I noticed something missing in the very first picture, namely the inductor plugged in between the airflow meter and the cable running to the ECU. This has not been fitted as standard but was offered as an upgrade. In the US I believe it was free of charge and done during routine inspections. Hence my surprise at its absence.

The part is still available from VW (025 906 302A) at the rather hefty cost of around 200€ and is nothing much more than a bunch of cables and two plugs. What it does is cure the impedance problem most WBX will develop at some point. Symptoms are sudden jerks and temporary loss of power when travelling at motorway speeds. (something to do with uncalled-for oscillations the meters are prone to develop over time. sorry, can't be more technical, am not much of expert ... )

This happened to me (and turned the 13-hour drive from Vienna to Rome into a three-day odyssey :twisted: ). After replacing pretty much every vacuum hose, sensor and electrical connection, I stumbled upon the above and, hey, problem solved.

For those of you who read German, you'll find more here: (I'd be happy to translate :wink: ). I did try the DIY solution described in this link, but it didn't really help to solve the problem, although the symptoms were less serve (occasional refusal to accelerate or maintain speed rather than kangaroo-style hopping about ... )

sorry if this already common knowledge. I did search and couldn't find anything on the topic.


sbclayton Fri Nov 20, 2009 7:09 am

Matt, herzlich Willkommen auf TheSamba!

I think what you are seeking is the "Vanagon Syndrome" - search for that and I think you will find your information.

Thanks for the T3-Info site - great information there.

Mzwo Fri Nov 20, 2009 7:31 am


thanks for the pointer. You're no strangers to hopping vans either, it would seem ... :roll: :wink:

In any case: It has been strongly recommended to me by people in the know to fit said part before fiddling with the airflow meter and related gadgetry. Apparently, if you don't, any subsequent adjustments are likely to be off the mark.

The rest of is quite useful as well. Especially diesel drivers will find plenty of info. Diesels are incredibly popular over here, in Austria even more so than in Germany. Not much of a surprise, really - we pay €1,10 per litre for unleaded, in Germany €1,40 .... Diesel's around 40 cents cheaper.


tencentlife Fri Nov 20, 2009 9:29 am

Mzwo wrote: hi,

In any case: It has been strongly recommended to me by people in the know to fit said part before fiddling with the airflow meter and related gadgetry. Apparently, if you don't, any subsequent adjustments are likely to be off the mark.

No, they are in error asserting that. The cable mod, or capacitor mod, don't alter the overall voltage signal the ECU sees to determine engine load, they only prevent that voltage oscillating.

Mzwo Fri Nov 20, 2009 12:26 pm

OK - allow me to rephrase. when investigating the causes of the "vanagon syndrom" the cable mod should be fitted first, if not already in place, and only subsequently should other bits be toyed with.


peterkissam Fri Feb 19, 2010 7:02 pm

Just to throw something out there for some of you....My solution....

What I thought was "vanagon syndrome" turned out to be "bad fuel pump plug".

While checking every possible source of the traditional "syndrome" causes and not finding a solution, I took another whack at it by testing all areas but adding to this: I hooked a Voltmeter to the fuel pump and watched it's input. My situation involved the engine running great but under various and non-patterning scenarios, it died. Seemed like no fuel. ARRRGGGHHH!!!!!

I started by REALLY inspecting my wire -loom. Seems I could create the problem when I grabbed ahold of the bundle and moved it while engine was running. Random bumps underway jostled it just enough to cause engine to "cut out", and, maybe not to re-fire.

Solution: I started cutting off the factory heat shrink on the original harness to inspect for breaks or shorts. I re-did the couple of ground "crimps" hidden inside but it was still not 100%. Checking further, over in the right-hand forward corner of engine compartment and ahead of and below the AFM/Air cleaner is a 2 pin M/F connector for the fuel pump electrical supply from the relay box. It passes into engine area thru small grommetted hole on firewall.
Bound tightly from factory, the 2 pin plug is clustered up and into the bundle of wires in that corner. Tight. Too tight.

My theory: The factory heat shrink tubing KEEPS SHRINKING over time and continues to "tighten" (mine was hard as a rock; yours is too!) This cause the loom to tighten in all directions and caused the connector to pull out of it's locked position and, while it looked "mated", it wasn't. It was "intermittent". Bound tightly, it gave and disconnected just enough to look mated.

Upon removing the air cleaner and AFM, I saw the EXACT CAUSE. The plug had indeed pulled out due to "tightening" of the factory harness.

The engine harness is strung thru tie points that, once tightened, don't allow alot of additional movement. I tie-wrapped the loom back into a tidy, service loop way. Much better.

I cut out and butt-spliced the connector and now have 4 months and 5,000 miles on a motor that NEVER sputters, coughs or dies. Runs GREAT!!

NOW I know why I got such a deal on this van. They couldn't fix it! The motor was fresh but unreliable. Yippeee!!!! V A L U E for my dollar.

So, look for the 2 pin connector. The same thing may be causing your "Vanagon Syndrome". Visually, it looked fine. Re-mating confirmed it; Butt Splicing ELIMINATED IT.


djs94124 Sat May 26, 2012 8:06 pm

My '87 camper recently failed its smog test. HC and CO levels were significantly over the limit at 15 mph and 25 mph on the dyno tester. So, I replaced the O2 sensor, plugs/cap/rotor and changed the oil. This time it passed with no problem, so I assume the 02 sensor was bad.

However, when I went in for my re-test, the technician informed me that although I passed smog, my HC's at idle were still way too high. This was a "Test Only" station so the tech couldn't make any adjustments. Although emissions at Idle are not part of the California Smog Test for 2WD Vanagons, and I passed smog, I still wanted to address the problem since I will be selling the rig soon and want to ensure it's running well.

Following Quixotix' very well-written procedures above, I connected a volt meter to the O2 sensor to set my idle mixture. I was unable to get the voltage below 0.8V by adjusting the mixture screw on the AFM, so I opened the AFM, cleaned the wiper track and then adjusted the wiper arm position ever so slightly toward the "Leaner" end of the scale (I did not change the position of the spring tension wheel, just the arm position) I was then able to adjust the mixture screw on the AFM to get the voltage to stabilized between 0.1 and 0.8 volts. The voltage seems to bounce up and down a little but stays within that range.

It will be interesting to see if my gas mileage improves a bit since I've been running too rich for awhile and didn't realize it until I failed smog.

Thanks for a great write-up, Quixotix, and all who contributed to the discussion.


denwood Mon Jul 30, 2012 8:41 am

This post was a great help tuning up my i4 conversion a few years after the conversion. (Revisited last week) Three things to add:

1. I added a dash gauge that monitors the o2 sensor signal from the Digifant II (jetta in my conversion) control unit. This shows you the trends as you drive with respect to fuel cut off, full throttle operation and when the system is open or close loop. A simple gauge therefore diagnoses much more than o2 sensor output. If the needle is pinned at .5 volts for example with the engine know to start looking at the coolant temp sensor etc. Conversely if the needle is spending more time under .5 than over, you know the mixture is biased rich as the computer tries to lean it out. Under full power I can see the needle pinned at .8 volts as the full throttle enrichment (high rpm) takes the system closed loop. Pics here:

2. Any vacuum leak and you're wasting time attempting this tuning. I found injector seals leaking using a propane torch with a plastic tube attached. This is much safer than spraying flammable carb cleaner everywhere. I was able to pinpoint leaks quickly as idle surged when propane was sucked in. Idle now is refreshingly stable.

3. I found approximate range adjusting the toothed wheel only. I then adjusted the toothed wheel "richer" one tooth at a time until I could no longer control idle mixture at .5 V with the CO adjust at idle. This improved full throttle performance and tested well in my standard Gtech tests. I'm seeing 0-60 in 14.2s (90 Westy auto with AC, 2 liter i4 conversion) vs 20 seconds stock 2.1 boxer. The Gtech reports 1/4 mile elapsed in [email protected] So from dog-slow stock to less dog-slow converted :-)

Thanks for the great post :-)


timfry Tue Jan 29, 2013 1:36 am

What a wealth of information! Without your excellent, well-reasoned post I would still be wondering why my Vanagon was running so rich.

I am the proud third owner of a 1985 Vanagon. I've had this wonderful thing for about six months, and in that time I have applied my rigorous perfectionism and mechanical skill to bring it back to nominal factory standards. After fixing many, many vacuum leaks and replacing every part that was out of tolerance (or even could be out of tolerance) I was at a loss to why it was running so rich.

I came across your post and it saved me. I got to the idle mixture screw and found that *someone* (likely in an attempt to counteract the myriad vacuum leaks) had turned the idle mixture screw all the way down (clockwise). While taking your very wise advice of "turning the screw clockwise and counting the turns", I thought the thing was seized. I thought, "What the heck," and turned it CCW instead. To my amazement it turned freely. Thinking I may have found a galled spot (or something) in the screw itself (wishful thinking), I turned it 360 degrees CCW. Smooth as a baby's bottom the whole way around.

If you hadn't put me on to this line of thinking with your wonderful and detailed post, I would have never known this system had an idle mixture screw, let alone found out that mine was turned to the richest position possible.

This is a long and very round-about way of saying "Thank you very, very much" and "I owe you one". Thank you for helping and for making the world a better place, if even if it's just helping me; I truly appreciate it.

Powered by phpBB © 2001, 2005 phpBB Group