You know all this stuff too. You did a lot of work on your 964 engine! It shouldn't take this much effort, right?

I saw that, and it looked a bit worn. But I did test it and it tested perfectly fine using the oscilloscope:

http://forums.pelicanparts.com/uploa...1589530663.jpg

Your suggestion regarding the Singer take-offs is a good one. I'm assuming that the guy with all of the Singer parts is this guy?

https://www.ebay.com/itm/Porsche-911-964-C2-C4-OEM-Bosch-Air-Flow-Meter-Part-0280203023-96460605000/252081028870?hash=item3ab1333f06:g:aWkAAOSwgQ9V6g~ I

http://forums.pelicanparts.com/uploa...1589530792.JPG

I mean, who really has 76 of these units, unless you've been buying them off of Singer? He's also located right here in SoCal -> two pretty big clues there.

Tomorrow I will readjust the wiper on there and button the unit back up. Pretty disappointing, as I'm quite sure that this unit is not the one that I gave with the engine (that only had 65,000 or so miles on it). In the meantime, I will try to source a new AFM that hasn't been messed with.

-Wayne

Wayne i`m a little too far but i can give you a unit for free -i have 4 - but it will take a little time with the post....let me know if you like it later ,problem ,you would just pay for the postage....i guess it would take a week with DHL?

Ivan

Wow, that would be great, that will help with comparisons! Just sent you a PM.

-Wayne

Proporsche,

I have the same issue : AFM tracks worn to substrate. I have tried retracking them and it 'looks' Ok on the scope but with the engine running there are dips and peaks (even on idle) resulting in a small hesitation.
Due to COVID, I am unable to do a test drive (it is still forbidden to drive classic cars in BE 'for leisure'). If the retracking doesn't work out, I will be looking for a replacement unit soon.
If you decide to sell any of the units, plz contact me.

@SteveW : I see that the output of the AFM (pin 7 ECU) goes straight into the multiplexer and then the A/D converter. So somewhere in the EEPROM there must be a mapping table for the input ? Can this be used/remapped to make the ECU work with the 'aliexpress' replacement boards for the AFM ? Or is it impossible to recalibrate perfectly ?

Kind regards,
FrankM from Belgium

I might have one also. Did mention that in post #8.

Actually, what's typically used is an EPROM (UV) and not a EEPROM (flashable).
Yes, it can be re-coded if you disassemble the original Bosch EPROM code and re-write it in 8051 assembly language.
Also not that difficult if you have the Intel 8051 compiler and can write C code. If you have access to an old PC with DOS,
there's an assembler in it you can use to write the code in assembly language. Good luck!

Steve, I was not aware that the 964 unit is a drop-in replacement? Is this really the case? And yes, Autobahn in LA is where all the Singer take-offs go, I bought some 964 calipers from them. Nice guys and they are willing to deal on the prices somewhat
Bill

Just noticed something different with the 964 AFM, although I don't know if it's
really that important. Had the cover off my 964 AFM and noticed it has a set of three contacts on the copper wiper unit instead just two on your 3.2 AFM. I know the two units are interchangeable, but there is that subtle difference.
Three contacts surely offer better redundancy than two.

Can you post a photo of that? I don't think I've seen a photo of one before that has three contacts?

Thx,

Wayne

I tested the AFM tracking on the oscilloscope three or four times, and I didn't see any issues with drops. However, I did see some "noise" on the output while the engine was running. Frankly, my gut is that this is not caused by the resistance strip, but is caused by some type of noise coming from somewhere else. If it were from the resistance strip, then I would think we would see consistent drops or spikes in the voltages (not both):

<iframe width="560" height="315" src="https://www.youtube.com/embed/Tx-ElHkZVGo" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>

I'm not sure how susceptible the ECU would be to noise in the AFM? I've never seen another scope trace of an AFM while the engine is running, so I don't know if this is typical / normal, or if this is an issue?

Thoughts?

-Wayne

Wayne, this thread will be a gold mine to future trouble shooters. And somehow it makes me feel a little better as a guy scratching his head when my 3.2 runs a little or a lot funny. After 25 years of driving my 911 I have come across most issues and helped fund your hobby with a lot of parts buying.

You are way more experienced than most of the other guys posting on Pelican. And some of the top experts have chimed in. I am following along to see just what is the culprit. My coil caused me some real headaches, and until I replaced it with a old black Bosch coil did that issue go away. You have replaced the coil and tried all the logical things.

Good luck finding that gremlin, and I look forward to seeing the solution and your cool 914 back on the road.

Mine has spikes too, but also slight hesitations in idle...
https://youtu.be/C2qnNGzoXrs


Sent from my iPhone using Tapatalk

this my 964 AFM
http://forums.pelicanparts.com/uploa...1589567673.jpg

It's not just the lack of noise or spikes in V out, but the correct linear rise of voltage as the door opens, that was originally calibrated by Bosch when the meters were made. Your voltage curve is likely very much out of spec, both from the curve change from the track wear, and possibly a someone messing with the spring tension. From experience, what you have is responsible for a lot of problems such as you're having. You can try renew it, it only brings it back or closer to original spec, and certainly can't be any worse than it is now. The width of the wear spot on yours is alarming though, 3x thicker than what is normally seen, as the tips of the wiper are just a needle's width in contact. Either the wiper was moved around before, or the tips have flatspotted. If that's the case, you practically need to recalibrate the whole thing by monitoring AFRs while driving the motor under different loads and rpms to bring the whole thing to where it should be. In any case, if you decide to renew it, best is to reposition is as illustrated below:

http://forums.pelicanparts.com/uploa...1589570027.GIF


Move the tip inward maybe about 1mm, and angle the wiper tip to maintain the same angle it had originally to the carbon track. Also adjust the wiper pressure so it has just enough pressure to maintain contact, so as when it just makes contact give just enough pressure for the wiper to deflect another 1 to 1.5mm of pressure. Too much pressure accelerates wear of the carbon, which you don't want. The arm should be positioned a couple of mm higher up on the pivot hinge than before in order to allow the tip to come inward. If you have a contact enhancer, lube and treat the track with it. I always use a DeoxIT pen on those. But your meter could be at the point of no return. It's also a bit questionable that half your wiper tip is off track completely, as if someone loosened that philips screw and repositioned the angle more clockwise to address a rich issue. They usually should not be that far off track, and most of the time are usually all on the carbon track when parked.

Porsche superceded the AFMs on the 3.2s to a 964 part number. A few years ago, Porsche offered replacement AFMs on an exchange basis for something like $700. No idea who redoes them for Porsche. The 964 units do all have a triple tip wiper tip, along with a reduntant wire from the wiper arm to the copper wiper bridge contact above it to reduce noise and resistance. The triple tip AFM can be found on some of the later 89 3.2s. These units do seem for some reason to exhibit less wear over time then seen on the 3.2 units, most of the time they look really good. Whether the calibration is the same, I'm not absolutely sure. The last 964 unit I tested seemed to have noticeably tighter spring tension than the one on my 86, and the throttle response also seems to be not as sharp. I ended up recalibrating it, loosening the tension to be the same as on my 3.2, I forget but I think by about 4 or 5 teeth, which brought the response to where it was equal.

Autobahn is where a lot of the Singer parts get sold to. You could contact them directly and speak to Mike who's in charge of those parts.

I saw a fairly well-done video last night that showed how someone "updated" their AFM with an attachment wire very similar to the one shown on your 964 AFM. I thought it looked a little "ghetto", but now I see the same exact thing on your 964 AFM here! I wonder if this was factory, or if these were updated by someone at some time.

Actually, here is the video:

<iframe width="560" height="315" src="https://www.youtube.com/embed/Z-QlPtsjh7g" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>

-Wayne

I agree - the strip wear seems to be thicker than every other photo I've seen of them. Seems likely that someone already messed with this, and moved it to a new spot (which is now possibly wearing).

I will try to move it to a new section and see if that makes a difference, while moving to get one of those 964 Singer take-off units. It is interesting that there was an original 930 part number that then got superseded to a 964 part number, but the internal BOSCH number remained the same on both of those units. Something is a bit odd about that - I'm quite familiar with the way Porsche performs supersessions, but not so much on BOSCH. The 3-contact 964 unit *should* be a separate part number if done "the Porsche way", as it's an updated (and presumably more reliable design). Also, in the back of my mind, I would think that the displacement increase of the 3.6 versus the 3.2 (more than 10% difference) would mean that the airflow would be greater through the 3.6 engine than the 3.2. That would mean that the effective range of the unit would be less for the 3.6 if the two AFM units were exactly the same (the flapper box would max out opening / pushing against the barn door screen sooner than on the 3.2, resulting in a more limited range of the box). If BOSCH increased the spring tension on the door to compensate, that would make sense, although technically that *should* not be the same part number. Who knows, it could be either one...

-Wayne

As Steve W noted that wire seems to be part of a 964 AFM update.
My AFM came from a 10k mi unmolested 964 3.6 and was sealed as new.

BTW if you ever fix this thing, the next thing you should address is the clusterfuch of elbows om your intake. From dyno testing every 90 degree bend kills power by 4 hp to the wheels, and the one from your air filter to the AFM is horrible, at least a 10hp killer!

Ahh, the good ole days. Way back in high school I used to program in 6502 Assembly Language on the Commodore 64. I got so good that I used to write programs for RUN magazine at the age of 16. I even found a copy on the Internet!

http://forums.pelicanparts.com/uploa...1589572135.JPG

Gosh, things are so much easier these days with libraries, and modules, and apps, and graphical user interfaces, etc. And compilers!

As expected, *none* of the high school girls were ever impressed with *any* of this stuff...

-Wayne