AFM Adjustment...

[Rustam]

For those who are interested in AFM adjustment procedure I added another webpage to my site. It talks about the theory, significance of the flap spring and bypass screw, and the adjustment procedure itself...

http://mysite.verizon.net/vzerdcib/e34a/id6.html

[SchnellE34]

Very nice website, keep up the good work.

[Rustam]

Very nice website, keep up the good work.

Thank you, I am glad you've enjoyed it...

[bahnstormer]

thanks! this is very cool, only problem i see
(gave it a quick skim maybe its addressed)
is that u are setting idle afr...what happens
at 6000rpm?

[Rustam]

thanks! this is very cool, only problem i see
(gave it a quick skim maybe its addressed)
is that u are setting idle afr...what happens
at 6000rpm?

what is "afr"? air flow ratio?
you should read the theory - the ratio is not set for the idle only but is set for the whole operational band. It just happens to be set at the idle...

[nariusb]

This is a great help. Thanks Rustam.
One question, please. Is what you call the "Link" a switch and does it disconnect the wire #2 to O2 sensor? Do all the readings/udjustments to be taken with this wire #2 disconnected from O2 sensor?
Thanks

[genphreak]

thanks! this is very cool, only problem i see
(gave it a quick skim maybe its addressed) is that u are setting idle afr...what happens at 6000rpm?Yea I like that write-up. It will be hard for some to follow but the good thing is its suitably technical for others...

I have been loking for a definate procedure to 'baseline' my AFM but had resolved to wait until I get a modern MAP sensor and ECU to replace the whole thing. Mine is good, but I suspect that many as they age suffer from spring deterioration which will affect the whole curve as well as shift the baseline air flow measurement. The point being; is the curve affected uniformly or is different depending on the Air Flap position... I mean by the time the AFM adjustment is out of whack the whole spring might be so out that the rest of the flap range may be way out of whack, thus it may be worth replacing the spring.

Does anyone have a way to buy some of those springs and the PCB for Bosch for BMW AFMs? The AFM rebuilders have to get them from someone... given that this definately affects our gas mileage a lot, we really need to find a good fix for this friends...

Absolutely great write-up Rustam!

[Rustam]

This is a great help. Thanks Rustam.
One question, please. Is what you call the "Link" a switch and does it disconnect the wire #2 to O2 sensor? Do all the readings/udjustments to be taken with this wire #2 disconnected from O2 sensor?
Thanks

"Link" is a switch.
Wire #2 must be disconnected during the procedure.

The switch is really superficial, it just lets you connect the system to observe operation of oxygen sensor as described in Bentley manual.

However, there are multimeters that enalbe the person to record "average value". When one uses such multimeter the "Link" switch allows the person to see whether the average value of the fluctiating voltage in open loop ("Link" disengaged) and closed loop ("Link" engaged) are same. They must be same - this way the map margins are adjusted with margins of operation of oxygen sensor.

Without such multimeter one should simply note on the position of the screw at the point of getting the mixture lean and at the point of getting the mixture rich after the adjustment. The middle position of the screw between those points is assumed optimal.

As soon as I am able to invest in one of such multimeters I am planning to provide an update on "fine tuning". Alas the results are satisfactory already. If you feel that the engine still consumes too much gas or too hot try to turn the screw within the limits of optimal operation (while the voltage fluctuates the screw is still given some freedom).

[Rustam]

Yea I like that write-up. It will be hard for some to follow but the good thing is its suitably technical for others...

I have been loking for a definate procedure to 'baseline' my AFM but had resolved to wait until I get a modern MAP sensor and ECU to replace the whole thing. Mine is good, but I suspect that many as they age suffer from spring deterioration which will affect the whole curve as well as shift the baseline air flow measurement. The point being; is the curve affected uniformly or is different depending on the Air Flap position... I mean by the time the AFM adjustment is out of whack the whole spring might be so out that the rest of the flap range may be way out of whack, thus it may be worth replacing the spring.

Does anyone have a way to buy some of those springs and the PCB for Bosch for BMW AFMs? The AFM rebuilders have to get them from someone... given that this definately affects our gas mileage a lot, we really need to find a good fix for this friends...

Absolutely great write-up Rustam!

Thank you Gen I am glad that you have enjoyed my website. I recall you also were able to use some info on shifter switch.

regarding the question about affecting the curve by the spring - from practical standpoint I don't think that the spring deteriorates to the point of being unable to provide matching response to the curve.

It is fairly long - the longer is the spring the more uniform is the force it provides to countract compression within given limit. If the spring was short then we would expect it to provide progressively increasing resistance. But resistance in this case increases at uniform rate since the spring is long.

Bottom line - it can become less resilient, but the laxing effect would still be uniform through turning space making us able to simply tighten the spring and experience conforming operation again...

[genphreak]

This is a great help. Thanks Rustam.
One question, please. Is what you call the "Link" a switch and does it disconnect the wire #2 to O2 sensor? Do all the readings/udjustments to be taken with this wire #2 disconnected from O2 sensor?
ThanksYep, you need this to be able to control the ECU behaviour whilst testing... the Bently explains it, perhaps Rustam will embellish more around it at some point.

[myles]

Hi Rustam
Nice write up. I'm keen to try it on my car. I have a question though.
You state - "Count 3 turns as you turn now counter-clockwise. This is the base setting of the screw."

I was just wondering where you got this info? The AFM on my car has the screw turned out about 3/4 (three quarters) of 1 turn. And a second brand new AFM that I have has the screw turned out about 1.5 turns.

If you are right about the 3 turns being the base setting, then my car has deviated from that a long way.

[Rustam]

Hi Rustam
Nice write up. I'm keen to try it on my car. I have a question though.
You state - "Count 3 turns as you turn now counter-clockwise. This is the base setting of the screw."

I was just wondering where you got this info? The AFM on my car has the screw turned out about 3/4 (three quarters) of 1 turn. And a second brand new AFM that I have has the screw turned out about 1.5 turns.

If you are right about the 3 turns being the base setting, then my car has deviated from that a long way.

Thank you, Myles...

I got this information from some e24 forum on the internet. The copyright was not claimed.

It is relatively simple to check whether the "three turns out" happens to be the "base setting". Base setting must block half of cross sectional area in the bypass channel. I will have to experiment with the flow meter more in order to provide this information.

For now you can use a technique that C.Probst uses in his book "Bosch FI & Engine management" once you find that you've achieved the fluctuation between .8V and .2V -

- turn the screw in just till you see the voltage turn relatively stabile above .5V and note the position of the screw

- turn the screw out just till you see the voltage turn relatively stabile below .2V and note the position of the screw

- the base setting is between these two extremes
__________

Since I had discovered that my evaporative purge valve did not operate properly having cleaned it I realized that the setting of my AFM was off. I'm going to have to remove it and quite possibly reset the spring - I will try to experiment with the position of the screw, and update my website.

I don't know when I will experiment with checking on whether the spring does in fact provide conformity through the band of operation, so I m not making a promise at this point.

As far as the tightness goes you can try to experiment making it tighter or looser and observing how it kicks when you just step on accelerator, and how easy it is to accelerate up to speed. The compromise setting should give you unhindered climb in speed while preserving good kick at pedal nudge. It is possible to tighten the spring so much that the car will have noticable trouble accelerating past 80mph - my AFM took 3 cogs clockwise to get to that state. So adjusting the spring between these two extremes (good kick & acceleration up to speed) should not be too challenging, as it seems that it does not require turning too far to experience a drastic change.

I just updated a reply just above in the thread regarding spring conformity check:

http://www.bimmer.info/forum/showpost.php?p=164573&postcount=8

I don't know when I will have actual procedure on my website. But this should be enough to get you started.
__________

P.S. I would not take the position of the screw in brand new AFM as a datum for perfectly adjusted meter. It's very concievable that one is expected to adjust the screw after installation - every engine needs its own specific mixture adjustment.

P.P.S. Please keep me posted on your experimentation...

[Rustam]

thanks! this is very cool, only problem i see
(gave it a quick skim maybe its addressed)
is that u are setting idle afr...what happens
at 6000rpm?

By the way, C.Probst claims that Motronic relies on injection tables between 4000-6000rpm...

I need to experiment more to provide actual procedure on verification of conformity.

Generally, one can connect the voltmeter that is able to show average of fluctuating voltage and go for a ride, at 3500-4000rpm but not above...

Based on the average setting observed at this speed one knows whether the spring is too tight (lean mixture, reading below .5V) or the spring is too loose (rich mixture, reading above .5V). Then the adjustment of the spring is performed. Unfortunately I don't have such expensive multimeter, so I am planning to experiment with analog type in close future.

[genphreak]

yo tx rustam, i didn't look at it that way. That makes me happier about the AFM. Now I will definately baseline that sucker ASAP and seal it up w silica. Might do it tommorrow. How will a cheap Digital Multi Meter not work for me? Will it not measure the fluctuations well enough as they occur?

I could borrow some mother Fluke or perhaps a portable cro even from work... might even save me a few bucks in gas this week! Thanks a heap, your write up is excellent, do keep up the good work. Good people lurk, better ones contribute, Saints document and update for the whole community to share :) ** Many thanks **

Yes I sure did pull my tranny selector due to your site's prompting- I was going to put a relay in to stop the reverse light contacts melting again, but am proly going to convert to LED tail globes soon enough anyway :) Nick

[Bill R.]

wrong information. The setting for the afm used on bmw is stamped right on the unit. You check it by measuring the depth of the screw with a caliper or a depth micrometer. I've posted the procedure on here before.



Hi Rustam
Nice write up. I'm keen to try it on my car. I have a question though.
You state - "Count 3 turns as you turn now counter-clockwise. This is the base setting of the screw."

I was just wondering where you got this info? The AFM on my car has the screw turned out about 3/4 (three quarters) of 1 turn. And a second brand new AFM that I have has the screw turned out about 1.5 turns.

If you are right about the 3 turns being the base setting, then my car has deviated from that a long way.

[Rustam]

yo tx rustam, i didn't look at it that way. That makes me happier about the AFM. Now I will definately baseline that sucker ASAP and seal it up w silica. Might do it tommorrow. How will a cheap Digital Multi Meter not work for me? Will it not measure the fluctuations well enough as they occur?

I could borrow some mother Fluke or perhaps a portable cro even from work... might even save me a few bucks in gas this week! Thanks a heap, your write up is excellent, do keep up the good work. Good people lurk, better ones contribute, Saints document and update for the whole community to share :) ** Many thanks **

Yes I sure did pull my tranny selector due to your site's prompting- I was going to put a relay in to stop the reverse light contacts melting again, but am proly going to convert to LED tail globes soon enough anyway :) Nick

Thank you, Nick.

Cheap multimeter can show occurence of fluctuations but cannot calculate the average value about which the fluctiations occur. It is only good to show at which position of the screw the fluctuations stabilize above or below .5V. That is as far as it can be useful. But to "fine tune" to .5V you need use the likes of a "Fluke" to extract the average value of the fluctiations so that you can set your system to average at .5V ...

It may be possible to use analog multimeter, by simply noting the extreme values between which the needle jumps and deducing the average that way. It may even be possible to set it within a range such that the needle virtually does not jump and is stationary. By these are guesses I have not verified them yet with actual experiments.

Use the Fluke if you can get your hands on it.

[Rustam]

wrong information. The setting for the afm used on bmw is stamped right on the unit. You check it by measuring the depth of the screw with a caliper or a depth micrometer. I've posted the procedure on here before.

Thank you Bill, I really apreciate your comment.
I will use it for my AFM and update my website...

[Rustam]

yo tx rustam, i didn't look at it that way. That makes me happier about the AFM. Now I will definately baseline that sucker ASAP and seal it up w silica. Might do it tommorrow. How will a cheap Digital Multi Meter not work for me? Will it not measure the fluctuations well enough as they occur?

I could borrow some mother Fluke or perhaps a portable cro even from work... might even save me a few bucks in gas this week! Thanks a heap, your write up is excellent, do keep up the good work. Good people lurk, better ones contribute, Saints document and update for the whole community to share :) ** Many thanks **

Yes I sure did pull my tranny selector due to your site's prompting- I was going to put a relay in to stop the reverse light contacts melting again, but am proly going to convert to LED tail globes soon enough anyway :) Nick

Nick take a look at Bill's comment before you proceed with my information:

http://www.bimmer.info/forum/showpost.php?p=164610&postcount=16

[Rustam]

One important detail that has not been mentioned (work in progress) - the hose that connects to the charcoal canister past the evaporative purge valve, must be plugged during the adjustment. Otherwise gas fumes entering the engine would create a problem by enrichment.

[myles]

Thanks for that correction Bill.

This is all good info for getting the old AFM setup working a bit better. I have some testing to do this weekend.

[Rustam]

wrong information. The setting for the afm used on bmw is stamped right on the unit. You check it by measuring the depth of the screw with a caliper or a depth micrometer. I've posted the procedure on here before.

I took out my AFM and adjusted in accord to this information. I've found the setting to be 4 and 5/8 of turns out from stop. As opposed 3 turns out from the stop.

I cannot say that I necessarily got "wrong information", as the setting itself assumes ideal circumstances and in that mandates consequent adjustment. But I really appreciate the fact that you've provided the comment Bill. This is certainly a very informative information that will enable the reader to decide for themselves...

I am planning to add this information to the website. Do you care to see yourself being credited for the input Bill?
__________
P.S. The fact that you've provided this information earlier is news to me. Along with searches at various places on the internet I used this very place to look up and never found anything. Even now as I try to search for "AFM" in the archives I find no information... Where is the original message?

[myles]

What number do you have engraved into your AFM?

My new afm has 185 in it, which im assuming means 18.5mm, which is about 11 turns out. From memory the one on the car has about 245.

[Rustam]

What number do you have engraved into your AFM?

My new afm has 185 in it, which im assuming means 18.5mm, which is about 11 turns out. From memory the one on the car has about 245.

Mine says "25,2".
I assumed same - 25.2 millimeters