Afm Antics

[jim21070 2]

It's been a long time no post on here. Mainly due to non-car activity taking up a lot of my time and all quiet on the 205 front. It has been a paragon of reliability

 

Last week I gave it a good service and replaced its rusty fuel lines. The last part of the service was to check the CO level. Mr. Gunson and his Gastester told me it was up at about 5%. I put on my spare but unknown and tatty looking AFM which shows no signs of having been tampered with and this brought the CO down to about 3%. Still not good but the car runs excellently on this AFM. All other causes of high CO have been ruled out so the AFM looked decidedly the guilty party.

 

As my original AFM had been tampered with in the past I used it to conduct a few experiments, having nothing much to loose really if it all went pear-shaped. Firstly I gave it a good clean in the recommended fashion and looked into retracking it. However on testing it with a 9V battery and an AVO Meter (very high quality but vintage multi-meter) I found no need to re-track it. The voltage output rose and fell very smoothly and accurately whilst opening and closing the vane. I proved that even if the AFM was re-tracked it would make no difference by taking voltage reading at various points on the carbon track using a smooth brass probe and seeing they were identical to what the voltmeter read when the main wipers were on the same part of the track as I tested. One thing I did notice was that the temperature compensation thermister had a surprisingly low vale of about 900 ohms at ambient whereas the AFM retracking guide suggests it should be several K. Anyone know what the 205 one should read to be sure?

 

After the clean, I set the idle mixture half-way, refitted it and fired up. After calibrating the CO meter, I checked the CO. It was still up at about 4%. Running the AFM with the cover off I noted that the vane was open a bit on idle, perhaps open by about an eighth and not fully closed as might be expected, given the presence of the idle bypass passage.

 

Knowing the AFM had been tampered with but not retracked my suspicion fell on the vane bias spring ( the clock spring within the black cogged disc) adjustment. I gave it a tweak whilst watching the CO meter. In one direction the CO shot up and the idle speed went down. A tweak the other way and the CO fell away to practically nothing, in fact the meter began reading a negative value That is illogical and I'm not at all sure just what a negative CO percentage means. Anyone?

 

The point was that the spring made an enormous difference in CO reading over a very small adjustment range of no more that 5 clicks total. It did not visibly change the position of the wiper on the carbon track.

 

This got me thinking: Is a major cause of AFM troubles not in fact potentiometer tracking problems but a loss of temper in the bias spring? I would be interested in what those of you that know far more than me about these things reckon on this.

 

I checked the meter calibration a few times and rechecked the CO level several times to ensure consistency.

 

Finally I set the spring to give 0.5% CO with the mixture screw half-way. I have not had a chance to raod-test yet but at least I seem to have an "MOT" AFM if nothing else. I plan shortly to have a full MOT emissions test carried out pending a successful road-test to see how the HC levels are and if indeed it will pass an MOT like it is. Luckily the real MOT is not until late September.

 

Your opinions please...

[Rob_the_Sparky 9]

Temperature sensor: no idea but would love to know. It would not surprise me if this is a problem.

 

Spring tension: quite possible but without testing the mixture under load you do not know that you are not leaning out the mixture at the top end with potentially nasty results to engine health and performance. You could be correcting idle mixture due to a poor AFM temp sensor, ECU temp sensor etc. Don't know anyone who has done this with a sensor in place though to say for sure. Would need either a wideband to monitor it under load or a rolling road session.

 

I have come across people whoose AFMs have been "professionally" had the spring tweaked and certainly some have had running problems under load afterwards. Will be interesting to hear your results.

 

Rob

 

P.S. I'd take the Gunson's readings with a pinch of salt, having used one it was pretty unstable. Good as guide and for setting up but I certainly don't think it is terribly accurate.

[jim21070 2]

Temperature sensor: no idea but would love to know. It would not surprise me if this is a problem.

 

I agree. Problem is, to test another properly it means breaking the seal and disconnecting it. Anyone please got a few scrap AFMs in which they could measure the thermister for us?

 

Spring tension: quite possible but without testing the mixture under load you do not know that you are not leaning out the mixture at the top end with potentially nasty results to engine health and performance. You could be correcting idle mixture due to a poor AFM temp sensor, ECU temp sensor etc. Don't know anyone who has done this with a sensor in place though to say for sure. Would need either a wideband to monitor it under load or a rolling road session.

 

I took it for a good test this evening under a lot of varying conditions and it is running really well. It seems crisper and more lively everywhere. From cold start it only surged once before settling to fast idle and idled perfectly at 950 RPM when the SAD closed. I also noticed that after a long, hard run it immediately idles now back at 950 RPM whereas previously It would idle at 1200 initially and then slowly drop back.

 

I have come across people whoose AFMs have been "professionally" had the spring tweaked and certainly some have had running problems under load afterwards. Will be interesting to hear your results

 

I too was worried that altering the spring may have had an adverse effect as this happened to a pal who had his AFM "adjusted" by the MOT station to get a pass and afterward it was not at all happy, down on power everywhere and running roughly. He just replaced his AFM with another and kept it as his "MOT passer". This does not seem to be the case with mine. The improvement overall is quite a revalation.

 

 

P.S. I'd take the Gunson's readings with a pinch of salt, having used one it was pretty unstable. Good as guide and for setting up but I certainly don't think it is terribly accurate.

 

Quite, Rob! It is a poor old thing in my opinion but it gives an idea of what is going on. As long as its limitations are well understood it is useful. I have toyed with the idea of "borrowing" the wide-band lambda sensor out of it and building my own meter. Maybe during the winter. One thing I have found is that it is better if it is powered from a fully stabilised 12V Power Supply. I have a 10A one that runs it OK. I assume the stabilised supply helps keep the lambda sensor heater at a more constant temperature.

 

 

My next steps will be to run it for a few days and during that time check the ECU coolant sensor is behaving itself. Somewhere I recall that someone posted the resistance readings for it. I must go and search it out.

[Rob_the_Sparky 9]

TBH, I'd replace the ECU temp sensor if in doubt. It isn't expensive...

[Guest noop]

TBH, I'd replace the ECU temp sensor if in doubt. It isn't expensive...

 

Are the "how to do's" listed on this website anywhere with tools required etc? It's v.handy, but I'm no genius with cars and have even less room (no garage or drive!) to mess about with mine. But I'd like to learn, 1 step at a time.

[Rob_the_Sparky 9]

There are technical articles on the main site but really only for the more complex jobs. Not sure where to start with beginners guides (I learnt as a child passing spanners to my father!)

[jim21070 2]

Not sure where to start with beginners guides (I learnt as a child passing spanners to my father!)

 

Exactly as I learned, followed by nearly 30 years of running old cars on a shoestring. This slowly developed from being a necessity to a hobby and then a passion.

 

IMHO and TBH, delving into AFMs is not for the beginner or the faint hearted and ones first exploits should be on one that you don't mind scrapping at the end. Also I believe a CO meter is a must along with somewhere peaceful and reasonably sheltered to work in. Engineering on the roadside is no fun whatsoever.

 

I think the best way to learn is to keep reading this forum and start out doing simple, straightforward jobs like routine servicing and going from there. You'll become competent in no time at all, gain a great deal of satisfaction and save a packet into the bargain.

 

We're all here to help you every step of the way.