Hello! Log in or Register   |  Help  |  Donate  |  Buy Shirts New!  See all banner ads | Advertise on TheSamba.com  
TheSamba.com
 
AFM Advanced Test
Page: 1, 2, 3  Next
Forum Index -> Bay Window Bus Share: Facebook Twitter
Reply to topic
Print View
Quick sort: Show newest posts on top | Show oldest posts on top View previous topic :: View next topic  
Author Message
telford dorr
Samba Member


Joined: March 11, 2009
Posts: 3036
Location: San Diego (Encinitas)
telford dorr is offline 

PostPosted: Wed Jun 06, 2018 9:50 pm    Post subject: AFM Advanced Test Reply with quote

The Issue

I've been trying to debug a FI problem on my '71 bus which has '75 bug FI, which up until now has been a stellar performer. Haven't fixed it yet, but I did come up with a pretty good method to test the AFM.

The L-Jetronic manual gives resistance measurement tests for the AFM, and they're good, as far as they go. But they don't check an important area of AFM operation: the unit output as you move the air flap from full closed to full open. These days, the resistive sliders on many of these units are starting to wear out. Here's a method of evaluating their performance.

Background

One would think that you could simply connect your ohmmeter between the slider output (pin 7) and one of the slider inputs: pin 6 (the "closed" end) or pin 8 (the "open" end), and measure the resistance, expecting to see some sort of linear resistance change. You can't. What you measure will seemingly increase and decrease randomly, which is contrary to expectations. But once you understand what's happening inside the unit, it makes sense.

What Bosch did was use Piecewise Linear Approximation: take a resistance element and divide it up into seven zones along the travel path of the wiper. These zones are wired up to a series string of low value laser-trimmed resistors. The values of the resistors in this string have much lower values than does the main resistance element that the wiper travels over. This allows the voltage at the connection points of these seven zones to be precisely set, and the voltage between these points varies linearly. So when you try to measure the resistance from the wiper (pin 7), you're measuring through the high resistance midpoint sections of the resistance element, which drops radically as you approach one of the seven zone connections. The measured resistance will go up, then down then up, etc. Brain hurt!

Solution

So: if we can't measure the resistance of this thing and have the readings make sense, how do we test it? Simple: apply a small voltage (I used 2.00 volts, because it's a full scale range on my meter) across the whole thing ("+" to pin 8, "-" to pin 6), and connect a voltmeter between pin 7 and pin 6. You push the flap from closed to full open and watch the voltage reading vary nicely from about 0.435 volts (full closed) to 1.91 volts (full open). The readings should vary smoothly and in one direction only. Different models of AFMs may have different endpoint voltages, but the voltage should vary smoothly. If the readings drop intermittently, or do anything else weird, then the resistance element is wearing out, or the resistance elements are damaged, and you will need a replacement AFM.

If you don't have a nice variable power supply to supply the 2 volts, a "D" sized flashlight battery (1.55 volts, brand new) should work nicely. The voltage readings will drop proportionally. Solder a wire on either end of the battery to act as terminals. You'll need a method of connecting to the terminals on the AFM. Extra small alligator clip leads from Fry's will work, or maybe some 0.110" or 0.187" female spade terminals from your local FLAPS would work.

I ran this test on my current AFM and an old broken one that lost its anti-backfire valve in the flap door (DANGER! Really hard on pistons!). Both measured EXACTLY the same (within 0.01 volts, anyhow), so these units are quite precision from the factory.

While I did this on a bug FI system, it's installed in a bus, and lots of type 4 buses have FI, so, being a bus guy, I posted it in this forum.

For what it's worth...
_________________
'71 panel, now with FI
'Experience' is the ability to recognize a mistake when you're making it again - Franklin P. Jones
In theory, theory works in practice; in practice, it doesn't - William T. Harbaugh
When you're dead, you don't know you're dead. The pain is only felt by others.
Same thing happens when you're stupid. - Philippe Geluck
More VW electrical at http://telforddorr.com/ (available 9am to 9pm PST)
Back to top
View user's profile Send private message Gallery Classifieds Feedback
airschooled Premium Member
Air-Schooled


Joined: April 04, 2012
Posts: 11360
Location: San Diego
airschooled is offline 

PostPosted: Wed Jun 06, 2018 10:26 pm    Post subject: Re: AFM Advanced Test Reply with quote

HECK YES! Can’t wait to run this on my spare in a few weeks when I get home.

Robbie
_________________
Learn to fix your vintage Volkswagen.
www.airschooled.com
Knowledge. Confidence. Reliability.
Back to top
View user's profile Send private message Send e-mail Visit poster's website Instagram Gallery Classifieds Feedback
Stuartzickefoose
Samba Post Whore


Joined: February 07, 2008
Posts: 10350
Location: SoCal for now...
Stuartzickefoose is offline 

PostPosted: Wed Jun 06, 2018 11:42 pm    Post subject: Re: AFM Advanced Test Reply with quote

Definitely love this test...
_________________
Stuart Zickefoose

2011 Jetta Sportwagen TDi 6 speed manual

206-841-7324
[email protected]
Back to top
View user's profile Send private message Send e-mail Gallery Classifieds Feedback
raygreenwood
Samba Member


Joined: November 24, 2008
Posts: 18336
Location: Oklahoma City
raygreenwood is offline 

PostPosted: Thu Jun 07, 2018 7:01 am    Post subject: Re: AFM Advanced Test Reply with quote

Yes.....very nice! From looking at several of these over the years....I don't own any of them right now.....the wiper boards were most probably printed in two layers with decent precision (for their day) of conductive ink deposition thickness.

Its also possible they used two different carbon inks in the layers for the two different needs.....more uniform resistance on one layer and better wear characteristics on the other. Then they laser trim....which has been bery common for a long time. Before laser trim....it was hand work in the assembly department.
You put the board on a test jig....apply power with a signal generator....and either trim resistance piecemeal with a scriber or rotary tool.

I do not know what polymer system they used back in the day in their inks or if they were a straight carbon or a blend like carbon/tin or carbon/silver or what percentage.

But if someone could ever find the original design detail and blueprint or find a brand new board that could be analyzed.....there are many shops that could reproduce these easily these days. Ray
Back to top
View user's profile Send private message Send e-mail Gallery Classifieds Feedback
Tcash
Samba Member


Joined: July 20, 2011
Posts: 12843
Location: San Jose, California, USA
Tcash is offline 

PostPosted: Thu Jun 07, 2018 10:34 am    Post subject: Re: AFM Advanced Test Reply with quote

Added to Tech Tips
Thank you
Tcash
Back to top
View user's profile Send private message Gallery Classifieds Feedback
Tcash
Samba Member


Joined: July 20, 2011
Posts: 12843
Location: San Jose, California, USA
Tcash is offline 

PostPosted: Thu Jun 07, 2018 2:49 pm    Post subject: Re: AFM Advanced Test Reply with quote

Image may have been reduced in size. Click image to view fullscreen.
Back to top
View user's profile Send private message Gallery Classifieds Feedback
furgo
Samba Member


Joined: September 06, 2016
Posts: 935
Location: Southern Germany
furgo is offline 

PostPosted: Thu Jun 07, 2018 6:14 pm    Post subject: Re: AFM Advanced Test Reply with quote

Thanks a lot for the great post.

Questions:

• Is there a reason you used such a small voltage (2V) for the test?
• What do you think the purpose of pin #8 is?

Some background to the questions:

Whenever I've done this test, I've always used a higher voltage at pin 9. I noticed when you connect an ECU, the voltage there is about Vbatt - 2V, so I either applied this voltage directly, or used the ECU to provide it. The end result of testing the voltage variation on the track should be the same for a low and high voltage there. I'm just curious as to whether you had a particular reason to test with low voltage.

A while ago, I measured the parallel resistor network on my spare AFM, and at some point I thought I'd simulate the circuit, but I never did. Your post prompted me to finally do it Smile

The pictures below show the AFM circuit diagram (without TS1), with the resistor network and the voltages at each wiper segment's junction point with their corresponding parallel resistor.

Those voltages would be read at pin #7 whenever the wiper reaches the given point on the track. Each full segment traveled corresponds to an angle of 15° of deflection on the vane.

• With an ECU connected and 12.6 V supply:

Image may have been reduced in size. Click image to view fullscreen.


• With your 2 V applied directly to pin #8. The simulation of my bus' AFM seems to confirm your measurements on the bug's AFM, so the PCBs are probably the same

Image may have been reduced in size. Click image to view fullscreen.


As per the need of pin #8, I've never understood why it's there. In some literature it's called "compensation (7.5V)", but I can't see what compensation it could be doing. The voltage to operate the AFM is provided by the ECU at pin #9. The ECU then seems to just tap into the voltage divider and the Air Flow voltage signal is taken at pin #7, referenced to ground.
_________________
'79 Westy, P22 interior, FI 2.0 l Federal, GE engine (hydraulic lifters)

Decode your M-Plate
Back to top
View user's profile Send private message Gallery Classifieds Feedback
Tcash
Samba Member


Joined: July 20, 2011
Posts: 12843
Location: San Jose, California, USA
Tcash is offline 

PostPosted: Thu Jun 07, 2018 7:39 pm    Post subject: Re: AFM Advanced Test Reply with quote

Is the 7.5v at pin 8 a reference voltage?
Back to top
View user's profile Send private message Gallery Classifieds Feedback
furgo
Samba Member


Joined: September 06, 2016
Posts: 935
Location: Southern Germany
furgo is offline 

PostPosted: Fri Jun 08, 2018 12:29 am    Post subject: Re: AFM Advanced Test Reply with quote

It wouldn't appear so, hence my question. For instance:

• Measured at the harness connector, with the connector unplugged, there is an open voltage of roughly Vbatt - 1.5 V at pin 9, and 0.67 V at pin 8
• Measured at the AFM, with the harness connector plugged in, there is a voltage of roughly Vbatt - 2 V at pin 9, and the voltage expected from the voltage divider at pin 8
• I could only bring pin 8 to be 7.5V by increasing Vbatt to 13.8 V (I did this one on the bench with a spare AFM, ECU, harness and power supply). The voltage at pin 9 was 12.36 V

If it were a reference, I would have at least expected it to have the reference voltage present with the connector unplugged. The voltage also varies depending on the battery voltage, which I wouldn't expect from a reference.

In addition to the Bosch manuals, often a good source of reference for the FI system is the Datsun 280-Z manual.

There I found this schematic of the AFM circuit, which is the same as the one in a bus (although most probably the resistor values are different):

Image may have been reduced in size. Click image to view fullscreen.


Notice how it seems to indicate that the air flow voltage output is not taken between wiper (7) and grond (6), but rather between 8 and wiper (7).

The main thing that this seems to do to the output curve is to flip it vertically, so that voltage DECREASES with INCREASING air flow. That would then better match the curve of AFM voltage vs. angle part (bottom left hand side) of the Bosch four-quadrant diagram. Taking the output of wiper (7) and ground (6) produces the opposite (voltage INCREASES with INCREASING air flow).

Image may have been reduced in size. Click image to view fullscreen.


On the other hand, this was probably a generic L-Jetronic marketing diagram, not particularly for buses, so I don't know how accurate it is, or if the magnitudes apply to the Type 2.
_________________
'79 Westy, P22 interior, FI 2.0 l Federal, GE engine (hydraulic lifters)

Decode your M-Plate
Back to top
View user's profile Send private message Gallery Classifieds Feedback
ve7kilohertz
Samba Member


Joined: June 28, 2015
Posts: 160
Location: Okanagan BC
ve7kilohertz is offline 

PostPosted: Fri Jun 08, 2018 7:24 am    Post subject: Re: AFM Advanced Test Reply with quote

Excellent, just excellent work guys! Glad to find some other electronics techs here.

I am putting together a couple of FI systems for 2 different engines, 1600 type 1 and 1700 type 4 or 914. I have started gathering all the sensors harnesses and injectors, and now have a few AFMs and ECUs and finding this thread is just perfect timing. I have the following AFMs and would be willing to pull them apart to contribute to this discussion. 280 200 006, 280 200 012 and 280 200 030. I would like to know what AFMs you have tested telford dorr and furgo. I think the bug is probably same as mine, 006.

Look forward to learning more about the ECU/AFM relationship and why some AFMs work with only certain ECUs.

Cheers
Back to top
View user's profile Send private message Gallery Classifieds Feedback
telford dorr
Samba Member


Joined: March 11, 2009
Posts: 3036
Location: San Diego (Encinitas)
telford dorr is offline 

PostPosted: Fri Jun 08, 2018 8:24 am    Post subject: Re: AFM Advanced Test Reply with quote

furgo wrote:
Questions:

• Is there a reason you used such a small voltage (2V) for the test?

My DVM, like most, has ranges 2 volts, 20 volts, etc. full scale. Wanted to use as much meter resolution as possible, but didn't think 20 volts would be acceptable to the AFM, so I used 2 volts. In short, purely arbitrary.

Quote:
• What do you think the purpose of pin #8 is?

(1) Feedback to an op-amp which drives pin 9, thus regulating the voltage at pin 8, and therefore the voltage across the resistor divider, or (2) a place to measure the voltage across the divider, used to ratiometrically determine the wiper angle, and thus the airflow. These are guesses - haven't reverse engineered the ECU circuit. Option 1 would be more elegant, engineering-wise, as it would make subsequent circuitry simpler. Option 2 would allow operation at much lower voltages, such as while starting, or with a low battery and bad alternator, but would require an analog multiplier circuit to determine the vane angle, adding complexity to the design. It has to be one of these reasons, as battery voltage varies too much to use without some kind of compensation.

Quote:
Some background to the questions:

Whenever I've done this test, I've always used a higher voltage at pin 9. I noticed when you connect an ECU, the voltage there is about Vbatt - 2V, so I either applied this voltage directly, or used the ECU to provide it. The end result of testing the voltage variation on the track should be the same for a low and high voltage there. I'm just curious as to whether you had a particular reason to test with low voltage.

See above.

Quote:
A while ago, I measured the parallel resistor network on my spare AFM, and at some point I thought I'd simulate the circuit, but I never did. Your post prompted me to finally do it Smile

The pictures below show the AFM circuit diagram (without TS1), with the resistor network and the voltages at each wiper segment's junction point with their corresponding parallel resistor.

Those voltages would be read at pin #7 whenever the wiper reaches the given point on the track. Each full segment traveled corresponds to an angle of 15° of deflection on the vane.

• With an ECU connected and 12.6 V supply:

Image may have been reduced in size. Click image to view fullscreen.


• With your 2 V applied directly to pin #8. The simulation of my bus' AFM seems to confirm your measurements on the bug's AFM, so the PCBs are probably the same

Image may have been reduced in size. Click image to view fullscreen.


As per the need of pin #8, I've never understood why it's there. In some literature it's called "compensation (7.5V)", but I can't see what compensation it could be doing. The voltage to operate the AFM is provided by the ECU at pin #9. The ECU then seems to just tap into the voltage divider and the Air Flow voltage signal is taken at pin #7, referenced to ground.

I've measured the resistance values on two units with exactly the same part number, and the values were quite different from each other, and from yours. But they both produced the exact same voltages at pin 7. The trimming process for these resistors must be pretty complex. I wonder if it was done on the bare resistor substrate, or on an assembled AFM? I noticed test pin pads on the substrate. One would think that if testing was done on the bare substrate, that it's manufacturing process would be able to hold much tighter tolerances, alleviating the need for such radical trimming. This makes me think that the trimming is done on assembled units, and is compensating for mechanical variations. It's a guess, however if it's correct, it would eliminate the possibility of mass-producing replacement resistor substrates. Crying or Very sad

===

ve7kilohertz wrote:
I would like to know what AFMs you have tested telford dorr and furgo. I think the bug is probably same as mine, 006.

Look forward to learning more about the ECU/AFM relationship and why some AFMs work with only certain ECUs.

Just the 006, as that's what my system uses.
_________________
'71 panel, now with FI
'Experience' is the ability to recognize a mistake when you're making it again - Franklin P. Jones
In theory, theory works in practice; in practice, it doesn't - William T. Harbaugh
When you're dead, you don't know you're dead. The pain is only felt by others.
Same thing happens when you're stupid. - Philippe Geluck
More VW electrical at http://telforddorr.com/ (available 9am to 9pm PST)
Back to top
View user's profile Send private message Gallery Classifieds Feedback
ve7kilohertz
Samba Member


Joined: June 28, 2015
Posts: 160
Location: Okanagan BC
ve7kilohertz is offline 

PostPosted: Fri Jun 08, 2018 9:31 am    Post subject: Re: AFM Advanced Test Reply with quote

Here is a block diagram of the ECU. I have 2 ECUs from 77 and 78, and it's all analog with no op-amps, unless they built a discrete one, but 741s and 4558s et al were in their heyday in 77.

cheers


Image may have been reduced in size. Click image to view fullscreen.
Back to top
View user's profile Send private message Gallery Classifieds Feedback
raygreenwood
Samba Member


Joined: November 24, 2008
Posts: 18336
Location: Oklahoma City
raygreenwood is offline 

PostPosted: Fri Jun 08, 2018 9:41 am    Post subject: Re: AFM Advanced Test Reply with quote

Questions:

Quote:
• What do you think the purpose of pin #8 is?

Option 2 would allow operation at much lower voltages, such as while starting, or with a low battery and bad alternator, but would require an analog multiplier circuit to determine the vane angle, adding complexity to the design. It has to be one of these reasons, as battery voltage varies too much to use without some kind of compensation.[/quote]


Many, many years ago when I owned L-jet systems and worked on quite few...I had one of the Bosch training courses that included D-Jet, L-jet and a few of its variants including LH and digifant among others.

This question of pin #8 was brought up by several in the class. I had not enough experience at the time with L-jet yet to know its significance. The answer by the instructor and several EE's in the crowd....were affirming the 2nd scenario in your answer ....that the voltage at pin 8 was there to provide a more stable voltage in the range required for low voltage situations when starting...both in cold weather, imperfect system grounding and in low battery conditions...because of the inherent issues with system output and low voltage.
How it works within the panel schematic...I have not dug into....but that wast teh answer that was circulated.


Quote:

A while ago, I measured the parallel resistor network on my spare
The pictures below show the AFM circuit diagram (without TS1), with the resistor network and the voltages at each wiper segment's junction point with their corresponding parallel resistor.


And bear in mind that the TS-1...meaning temperature sensor in the AFM...is variable ballast for the resistor stack up of the potentiometer.


Quote:
The trimming process for these resistors must be pretty complex. I wonder if it was done on the bare resistor substrate, or on an assembled AFM?


No it would not be done on a assembled AFM. Not that it could not be...but it would be costly and create a lot of defects.

Not every board that comes off of a print line survives or is defect free. And not all of them survive laser trimming. Installing a board on an expensive apparatus and then having to butcher it back out would be poor planning....and the plants that produce ceramic circuit boards are rarely the same plant/line as device assembly.

This is because printing....and more importantly curing... on these type of ceramic and/or composite boards...cannot be done in the same environment that uses some of the other materials that are in the AFM....adhesives etc....and must be in a very static and humidity controlled environment.

I have been in two of Bosch's ceramic board plants. One was an anti-lock brake board plant...the other one....I have no idea what the part was going into. Both employed both silver and carbon blend traces being printed and fired in lehrs.


Actually the laser trimming is not that complex. What is the MOST complex part....is printing the carbon trace to a uniform enough thickness....that it CAN make the trimming process simple by allowing the laser to ablate "resistance"...simply by subtracting uniform amounts of square mils or millimeters...instead of having to read...ablate...read...ablate....like the process that was (and still is in some places)...done by hand with a scriber.

So the circuit is placed on a jig....is read at one or more positions...probably one for each main step....and the ablation is simply done in one pass of the laser. The test jig is part of the process design of the product and is designed at the same time.
Up until recently.....say the last 7-8 years...every single circuit or board design would have its own dedicated checking jig and/or test unit connection harness that goes to the end of the print and/or assembly line to have full functional test and allow steps like laser trimming ans checking of surface mount components.


But with the improvement of X/Y pick and place robots and.... excellent vision systems....in many cases ...just employing pattern recognition, programming and placing the circuit on a target...can allow everything from pinpoint robotic soldering to laser ablation and drilling of vias...without complex and costly jig work.

Quote:
I noticed test pin pads on the substrate. One would think that if testing was done on the bare substrate, that it's manufacturing process would be able to hold much tighter tolerances, alleviating the need for such radical trimming.


Laughing Laughing ...Thank you for the vote of confidence for my industry!....but NO....not 40 years ago. You need to understand that these are screen printed. That does not mean that the process is inaccurate. But contrary to public knowledge of the process....which typically stops at T-shirts and posters....there are over 300 basic variables in the screen printing process...and THOUSANDS of interconnected interactions.

Example: In any printed circuit...you need to control film thickness and trace or film width/length/dimension. Resistance, capacitance, inductance...are volumetric with regard to the inks they are made of. We can control ink deposit to with a few microns...but it is not easy with conductive inks....because:

1. Your ceramic board surface texture will vary from lot to lot by as much as 2-25 microns. That texture must be filled with the conductive ink until the ink is high enough to create a uniform layer with a surface profile that is acceptable. That added volume to reach that level...creates either added resistance or capacitance..depending on what material you are dealing with.

2. Conductive inks have a certain amount of critical variation from lot to lot....particle/flake size, shape, texture. This affects the amount of solvent and binder that must be added to make a paste. That changes the rheology of the ink....that changes the screen recipe...that changes the print dynamic...that changes the drying/curing rate...that changes the particle dispersion....that changes the conductivity....that changes the surface profile...that changes the connectivity within the device.

Yes...all of this CAN be controlled within the printing process...especially now with much higher tech screens and photo emulsion masks and imaging processes than we had even a decade ago.

And no...you cannot do these with inkjet....at least not any time soon...for a reasonable cost. Laughing

These kind of process problem solving and improvements are exactly what I do in the printed electronics industry.....however....to solve for ALL of the variables and more listed above......and keeping mind that each new device that comes through a shop is different and requires the levels of tuning............its far simpler to just print a "decently" uniform layer of ink....and tune it with laser ablation.

Virtually all resistive circuits have some laser or scribe ablative tuning..because carbon inks are the most variable...and printing carbon is the most variable.... due to particle size and porosity.

Silver connecting traces are actually harder to accomplish depending on the design...because laser trimming is not cost effective to reshape traces that are poorly printed in the first place....not with silver conductive inks running between $900 and $1200 a kilo....and a kilo is about 1 pint... Wink



Quote:

This makes me think that the trimming is done on assembled units, and is compensating for mechanical variations. It's a guess, however if it's correct, it would eliminate the possibility of mass-producing replacement resistor substrates. Crying or Very sad


Actually no...it should not be hard to get done....there are many custom board shops in the US alone that do work like this every day. But it will take more than just a basic schematic like we have been showing here of the sensor steps.

If someone had a BRAND new board...or nearly perfect board that you can take to one of these shops...explain what it does...show them a complete part....let them map it and measure it....then pay the CAD costs....they could probably have parts for you within a week.

Also with all of the new board types...and new hard polymer ink types....you could probably also reproduce this on something other than ceramic.

The AFM board is nothing special in the printed electronics world. This same type of part has been common in everything from toaster oven dials to washing machine cycle selectors and lighting rheostats for eons....and still is.

However....the tough part...IS getting the board into the AFM...in the right position to be tuned right with vane movement and spring tension. I bet the jig for install and calibration was COOL!
Ray
Back to top
View user's profile Send private message Send e-mail Gallery Classifieds Feedback
furgo
Samba Member


Joined: September 06, 2016
Posts: 935
Location: Southern Germany
furgo is offline 

PostPosted: Fri Jun 08, 2018 10:20 am    Post subject: Re: AFM Advanced Test Reply with quote

ve7kilohertz wrote:
I would like to know what AFMs you have tested telford dorr and furgo.


I tested a 0 280 200 020, bus 2.0l AFM for a '79 Federal ECU. I also tested a 0 280 202 007 for an Opel Commodore (6 cylinders).

The voltage output was quite similar, but the resistor values were different. The return spring is set much stiffer on the 007 unit, and also it has a different max air flow (400 m³/h vs. 300 m³/h).

ve7kilohertz wrote:
Look forward to learning more about the ECU/AFM relationship and why some AFMs work with only certain ECUs.


I think if the AFM was designed for a bus, the only AFM/ECU incompatibility would be the presence or absence of the air temperature sensor (TS1) on the AFM box.

I *think* for instance, that your 0 280 200 006 AFM does NOT have a TS1, so you'd need an ECU that does not expect it.

telford dorr wrote:
My DVM, like most, has ranges 2 volts, 20 volts, etc. full scale. Wanted to use as much meter resolution as possible, but didn't think 20 volts would be acceptable to the AFM, so I used 2 volts. In short, purely arbitrary.


Ok, makes sense now.

telford dorr wrote:

(1) Feedback to an op-amp which drives pin 9, thus regulating the voltage at pin 8, and therefore the voltage across the resistor divider, or (2) a place to measure the voltage across the divider, used to ratiometrically determine the wiper angle, and thus the airflow. These are guesses - haven't reverse engineered the ECU circuit. Option 1 would be more elegant, engineering-wise, as it would make subsequent circuitry simpler. Option 2 would allow operation at much lower voltages, such as while starting, or with a low battery and bad alternator, but would require an analog multiplier circuit to determine the vane angle, adding complexity to the design. It has to be one of these reasons, as battery voltage varies too much to use without some kind of compensation.


Excellent, thanks. I'm not sure it's (1) though, as I don't think the current is regulated. The voltage at pin 8 varies with varying Vbatt. Unfortunately, it's not possible to reverse-engineer the ECU in this area. I tried a while ago, but the AFM inputs go first through some filtering and then straight into one of the Bosch proprietary ICs (i.e. black box).

telford dorr wrote:
The trimming process for these resistors must be pretty complex. I wonder if it was done on the bare resistor substrate, or on an assembled AFM? I noticed test pin pads on the substrate. One would think that if testing was done on the bare substrate, that it's manufacturing process would be able to hold much tighter tolerances, alleviating the need for such radical trimming. This makes me think that the trimming is done on assembled units, and is compensating for mechanical variations. It's a guess, however if it's correct, it would eliminate the possibility of mass-producing replacement resistor substrates. Crying or Very sad


Good point, I noticed the test pads too. I don't know. I guess there were two processes: 1) trimming, with a voltage measured at the test points and 2) placing the PCB in the right location, so that the wiper start and end points have got the right voltages. I had always thought that they were done in the same step, so that measurement only needs to be done once. Say if trimmed replacement PCBs were produced standalone, would it not be a relatively easy process to place them in the right spot, though? Am I missing something or is it not just a matter of checking that the wiper is in the middle of the track throughout its full travel, and that the initial (vane closed) and final (vane fully open) voltages are as expected? At leas the AFM remanufacturers manage to do that, right?

(I wrote this before Ray's reply, which probably invalidates my guesses there Smile but the question of installing the PCB still remains)

raygreenwood wrote:

And bear in mind that the TS-1...meaning temperature sensor in the AFM...is variable ballast for the resistor stack up of the potentiometer.


I need some more time to read your full reply (thanks Ray!), but this bit caught my eye at a glance. Unless I'm missing something, on a bus AFM the TS1 thermistor does not affect the potentiometer output. Yes, it's connected to pin 6, but that pin is GND (see the Datsun circuit I added on my last post)
_________________
'79 Westy, P22 interior, FI 2.0 l Federal, GE engine (hydraulic lifters)

Decode your M-Plate
Back to top
View user's profile Send private message Gallery Classifieds Feedback
raygreenwood
Samba Member


Joined: November 24, 2008
Posts: 18336
Location: Oklahoma City
raygreenwood is offline 

PostPosted: Fri Jun 08, 2018 11:31 am    Post subject: Re: AFM Advanced Test Reply with quote

Yes....TS-1 may not affect potentiometer outpit on bus (I will have too look at the schematic)....bjt it most certainly did on Datsun and other makes....where the connection of the variable resistant of the TS-1 is connnected to either input + or - on the potentiometer input leg..

In my reply....in reaponse to your last post.....make note that trimmimg of these boards is not that complex. Rqy
Back to top
View user's profile Send private message Send e-mail Gallery Classifieds Feedback
furgo
Samba Member


Joined: September 06, 2016
Posts: 935
Location: Southern Germany
furgo is offline 

PostPosted: Fri Jun 08, 2018 2:43 pm    Post subject: Re: AFM Advanced Test Reply with quote

raygreenwood wrote:

In my reply....in reaponse to your last post.....make note that trimmimg of these boards is not that complex.


Indeed, I had undestood that (actually, a while back in a conversation with you Smile). My question was rather on the installation of the board.
_________________
'79 Westy, P22 interior, FI 2.0 l Federal, GE engine (hydraulic lifters)

Decode your M-Plate
Back to top
View user's profile Send private message Gallery Classifieds Feedback
ve7kilohertz
Samba Member


Joined: June 28, 2015
Posts: 160
Location: Okanagan BC
ve7kilohertz is offline 

PostPosted: Fri Jun 08, 2018 3:56 pm    Post subject: Re: AFM Advanced Test Reply with quote

furgo wrote:
I think if the AFM was designed for a bus, the only AFM/ECU incompatibility would be the presence or absence of the air temperature sensor (TS1) on the AFM box.

I *think* for instance, that your 0 280 200 006 AFM does NOT have a TS1, so you'd need an ECU that does not expect it.


Correct, 006 is 6 pin with no TS1.

And, as I think is indicated by the several bus harnesses and intake plenums I have, that there is a separate TS1 in the intake plenum, a small 2 pin device same plug as injectors etc. and maybe these were just on early FI buses? I don't know but I should be able to make something work. I have one 030 AFM with 7 pins here, and a couple of the remote TS1s and will do some testing this weekend to see if the values track thru various temps. Just thinking now, my harnesses all have the 7 pin AFM plug, so the plenums I have collected must have been from earlier bus with separate temps sensor?? More research required.

I will also measure the AFMs this weekend, with the same 2.0 volts and post results. It's supposed to be very wet here this weekend so a good indoor project for me.

Cheers
Back to top
View user's profile Send private message Gallery Classifieds Feedback
Tcash
Samba Member


Joined: July 20, 2011
Posts: 12843
Location: San Jose, California, USA
Tcash is offline 

PostPosted: Fri Jun 08, 2018 4:16 pm    Post subject: Re: AFM Advanced Test Reply with quote

FYI
Fuel injection conversion shopping list - parts required FAQ

Thnx to mayor ratwell
Fuel injection swaps
http://www.ratwell.com/technical/FISwap.html#swaps
Back to top
View user's profile Send private message Gallery Classifieds Feedback
furgo
Samba Member


Joined: September 06, 2016
Posts: 935
Location: Southern Germany
furgo is offline 

PostPosted: Sat Jun 09, 2018 12:30 am    Post subject: Re: AFM Advanced Test Reply with quote

ve7kilohertz wrote:
And, as I think is indicated by the several bus harnesses and intake plenums I have, that there is a separate TS1 in the intake plenum, a small 2 pin device same plug as injectors etc. and maybe these were just on early FI buses?


Hmm... I don't know. I'm not an expert, but I've never heard of a remote TS1 on busses. I think D-Jetronic had it mounted on the plenum, so you might have gotten D-Jetronic parts instead of L-Jetronic.

- Do you have a picture or part number for it?

ve7kilohertz wrote:
the plenums I have collected must have been from earlier bus with separate temps sensor??


- Where exactly is the part mounted on the plenums you have? Do you have a picture of these plenums?

In any case, you might want to start a separate thread for the FI system(s) you're putting together.
_________________
'79 Westy, P22 interior, FI 2.0 l Federal, GE engine (hydraulic lifters)

Decode your M-Plate
Back to top
View user's profile Send private message Gallery Classifieds Feedback
ve7kilohertz
Samba Member


Joined: June 28, 2015
Posts: 160
Location: Okanagan BC
ve7kilohertz is offline 

PostPosted: Sat Jun 09, 2018 8:59 am    Post subject: Re: AFM Advanced Test Reply with quote

Furgo you are correct. I had another look at the sensors and plenum, and upon further investigation, the plenum intake runners are opposite each other, not offset like the bus, still black pressed steel, not aluminum like the type 3 I have as well. Measured the temp sensor and found it to be 400 ohms, then checked the p/n and it is for the D-Jet system. Damn. Well now I have to use the 7 pin AFM and hope I can get it to work properly with my ECUs as they both have connections to pin 27 inside and will be looking for TS1. So much to learn.

I'll have some time today to make the AFM measurements and report back here my findings.

Cheers
Back to top
View user's profile Send private message Gallery Classifieds Feedback
Display posts from previous:   
Reply to topic    Forum Index -> Bay Window Bus All times are Mountain Standard Time/Pacific Daylight Savings Time
Page: 1, 2, 3  Next
Page 1 of 3

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum

About | Help! | Advertise | Donate | Premium Membership | Privacy/Terms of Use | Contact Us | Site Map
Copyright © 1996-2020, Everett Barnes. All Rights Reserved.
Not affiliated with or sponsored by Volkswagen of America | Forum powered by phpBB