AFM output and shunting problems

[blitzracing]

As some of you will know, I’ve been messing around with the analogue inputs to the ECU as a matter of interest for some time now and recently got a couple of AFM’S to have a look at. Electronically they are pretty straight forward with only one IC (op amp) and a handful of other semiconductors, so not much to go wrong. On the various voltages front, there are various trim resistors, and one of them sets the static voltage output when there is no airflow. This voltage is then applied through the airflow range. The static Land rover spec' is .2 -.7 volts, which is a huge range considering how sensitive the TVR fuelling is at low revs due to cam duration, and from past experience it took only a shift of .2 volts to cause the engine to run lean at 1500 rpm, and cause some shunting. This may account partly for the reason why some cars shunt and some don’t as the original tolerances are so wide. The upshot of this is a change of a resistor value (R24 on the PCB) can control the AFM output, and it works out at around 1 ohm reduction gives a 1mv increase and increasing this voltage increases the available fuelling. The easiest way to tweak this resistance is to add a high value resistor across R24 (Example on the two AFM I have to test 46k ohms gives about 200mv more - this will vary a bit dependent on the original trim resistor value). Unfortunately these are all bench tests, and I do not have a TVR with the problem to hand at the moment to verify the change in fuelling, but may well account for the apparent variable low speed performances with different AFM’S. Any one here game to open up the AFM (Awkward but not impossible) to try tweaking the resistor if they have shunting issues?

Mark

[jayzcycles]

i would if i had any clue of what your talking about.when you say shunting is that iratic idle.

[That Daddy]

No,thats the slow speed snatching(around 1500 rpm)that some of these cars are suffering,emample is when your in slow moving traffic in say second gear say 20mph and on a even throttle the car lurches due to lean spot in the fueling.

[FlipFlopGriff]

I found that a proper original rotor arm solves the problem (or certainly reduces it a lot).
You want one with emobossed numbers on the top and the Lucas name also embossed, just simply getting one in a Lucas box isn't enough as they are st.
FFG

Edited by FlipFlopGriff on Friday 11th April 15:40

[paulathome]

Ehhhh???
Do you mean rotor arm or am I missing sumert?
Paul.

[Chimjunkie]

Nice work Mark. Anything inside that box which would allow the afm to be bored out and then recalibrated? Rather than remapping ecu..

[That Daddy]

FlipFlopGriff said:

I found that a proper original rocker arm solves the problem (or certainly reduces it a lot).
You want one with emobossed numbers on the top and the Lucas name also embossed, just simply getting one in a Lucas box isn't enough as they are st.
FFG

That may help,but thats not the actual problem Blitz speaks ofbut worth a stab if someone as said problem.

[FlipFlopGriff]

Sorry did mean rotor arm - gave us no end of problem with a non original one - they are £20+ though but worth the money.
FFG

[blitzracing]

Chimjunkie said:

Nice work Mark. Anything inside that box which would allow the afm to be bored out and then recalibrated? Rather than remapping ecu..

Interesting one. As I see it you have only 2 or 3 parameters to play with, one being the starting point, and the other being the rate of change. Im sure there is an upper maximum voltage point as well, which I know is in excess of 5 volts. By increasing the diameter of AFM, the flow rate through the hot wire port would drop, so you would have to increase the rate of voltage change to compensate and there is likely to be a feedback resistor on one of the Op amps that will control this. In the real world, it would be easy enough to put a couple of variable resistors onto the circuits, and using a fixed rate of air supply (ie a hairdryer on cold) get the voltages to match for modified against an unmodified as a starting point, so the corresponding mapping should not be far out. We are making an assumption that everything will be linear, and Im not an expert in gas dynamics so I dont know quite what would happen overall, but I dont mind experimenting.

Mark

[Chimjunkie]

blitzracing said:

Chimjunkie said:

Nice work Mark. Anything inside that box which would allow the afm to be bored out and then recalibrated? Rather than remapping ecu..

Interesting one. As I see it you have only 2 or 3 parameters to play with, one being the starting point, and the other being the rate of change. Im sure there is an upper maximum voltage point as well, which I know is in excess of 5 volts. By increasing the diameter of AFM, the flow rate through the hot wire port would drop, so you would have to increase the rate of voltage change to compensate and there is likely to be a feedback resistor on one of the Op amps that will control this. In the real world, it would be easy enough to put a couple of variable resistors onto the circuits, and using a fixed rate of air supply (ie a hairdryer on cold) get the voltages to match for modified against an unmodified as a starting point, so the corresponding mapping should not be far out. We are making an assumption that everything will be linear, and Im not an expert in gas dynamics so I dont know quite what would happen overall, but I dont mind experimenting.

Mark

You got a qualification in electronics Mark? You certainly know your stuff! If it is just a single Op-amp then indeed the feedback "gain" resistor could be trimmed or replaced with a multi-turn pot. Figure this one out Mark and you will have many customers i'm sure. Would not be difficult to set up a bench with a centrifugal blower (pulling air) and an inline calibrated afm as a reference. The blower could be run from a rheostat to set fan speed.

Go on! You know you want to!!

[blitzracing]

I would be really well sorted if I could get a full circuit diagram, but none of the AFM repairer's would play ball. There are 4 Op Amps in the chip, but I doubt they are all used. If I spend enough time I could reverse engineer the board, but being realistic I need to concentrate on just the bits that make a difference, not the various filters and stablising circuits. Fan wise I have simular ideas, I can make a fixture to hold a computer fan against the air intake or outlet, and feed it with different fixed voltages to give me some bench figures on a working unit. Air flow is a horribly fluid thing to try and measure otherwise. Ill keep digging away.

Mark

[Chimjunkie]

blitzracing said:

I would be really well sorted if I could get a full circuit diagram, but none of the AFM repairer's would play ball. There are 4 Op Amps in the chip, but I doubt they are all used. If I spend enough time I could reverse engineer the board, but being realistic I need to concentrate on just the bits that make a difference, not the various filters and stablising circuits. Fan wise I have simular ideas, I can make a fixture to hold a computer fan against the air intake or outlet, and feed it with different fixed voltages to give me some bench figures on a working unit. Air flow is a horribly fluid thing to try and measure otherwise. Ill keep digging away.

Mark

Good luck. Keep us posted. For air flow i would be more inclined to set up something similar to the real deal with hose each side of the afm so that a similar turbulence is experienced. Would also think that a heftier fan is needed? My thoughts were to have two afms in line one a test and the other a ref.

Will leave you alone now! It's the scientist coming out in me after spending the last 3 years in the building trade. Miss it sometimes

[MikeTVR]

I'd be fairly comfortable in changing resistors to set voltages. I'm an electronics engineer myself. I do have shunting problems at anything below 2000RPM but this is normally said to be due the HC cam. If you have info on ideal static voltage at least I would give it a go.
My Chimaera is a 4.0HC which seems to come with shunting as standard. Does your engine have the TVR high lift cam ?
Mike

[That Daddy]

MikeTVR said:

I'd be fairly comfortable in changing resistors to set voltages. I'm an electronics engineer myself. I do have shunting problems at anything below 2000RPM but this is normally said to be due the HC cam. If you have info on ideal static voltage at least I would give it a go.
My Chimaera is a 4.0HC which seems to come with shunting as standard. Does your engine have the TVR high lift cam ?
Mike

Mike, its not just the HC that suffer this problem(although the HC appear more prone)the poor low speed manners of the HC cammed motor just makes the situation worse,i have it very slightly(400 HC)but my brothers car is worse(non HC400)to the point of being awful some days(outside temp sensitive)with Blitz racings(Mark)input on here i feel my brother is gonna be a happier man this summer(just got to get the car over to my workshop)his car is out of hibernation shortly.

Edited by That Daddy on Sunday 13th April 10:44

[blitzracing]

That Daddy said:

MikeTVR said:

I'd be fairly comfortable in changing resistors to set voltages. I'm an electronics engineer myself. I do have shunting problems at anything below 2000RPM but this is normally said to be due the HC cam. If you have info on ideal static voltage at least I would give it a go.
My Chimaera is a 4.0HC which seems to come with shunting as standard. Does your engine have the TVR high lift cam ?
Mike

Mike, its not just the HC that suffer this problem(although the HC appear more prone)the poor low speed manners of the HC cammed motor just makes the situation worse,i have it very slightly(400 HC)but my brothers car is worse(non HC400)to the point of being awful some days(outside temp sensitive)with Blitz racings(Mark)input on here i feel my brother is gonna be a happier man this summer(just got to get the car over to my workshop)his car is out of hibernation shortly.

Edited by That Daddy on Sunday 13th April 10:44

My car does not have any fuelling issues at all, but its a 3.9 Ginetta, with cast manifolds, Piper 270 cam mildly ported and Chim Chip. Im just interested in the ECU has so few people know it really works, and Ive worked on a few TVR's with the shunting problem as part of this "how does it work" investigation. I tried one of the modified AFM's on my car and it over fuelled to the point the ECU could not control the mixture, so Ive lifted the voltages too high. I now have enough tube and a fan to run a stock AFM inline with a modified one to get some really accurate readings. I think a 10 turn variable resistor could be the way to go , so you could tweak the airflow output very accurately as long as it was done hand in hand with monitoring the lambda probe outputs at tick over. All I need to achieve is to run slightly richer at 1500 rpm under low air flows (caused due to the cams longer duration), as the cars run lean and the ECU cant pull it back into line, so the car snatches and altering the AFM output (dc offset) has proved most effective so far. It may not provide a total cure but it does improve things. Fuel pressure and temperature sensor changes dont effect the problem, as the ECU just moves the basic fuelling mixture over a much wider range, and the 1500 rpm fuelling "hole" still occurs. Any one thinking of a power check on the rollers, and a 5ltr motor, Id love to know high the AFM voltages go under full load. Its over 5 volts on the bench, but I need real world figures. Not seen over 4 volts on my car so far.

Mark

[That Daddy]

Mark although your spec is slightly off the TVR spec (only slightly mind,exhaust manifolds etc)and you dont have the problem but could quite easily,it just goes to show how no two engines with the CUX system are the same,as you already know i think this set/up teeters on that fueling ragged edge(closed loop,emission compliance etc)at low speed,but then i am just highlighting what you already know

[Chimjunkie]

OK, I've sourced a Jag Lucus 5AM afm for peanuts. Going to get this bored out as far as possible. All i need now is a 10 turn resistor a soldering iron with a bit smaller than a shovel and to know what resistor to change for the gain! Mark, its over to you

[blitzracing]

So all I need now is one of those "Guaranteed 100 % fantastic power increase" electric superchargers to make my own wind tunnel for calibration !

Mark

[taylormj4]

Hi Mark,
Great to see someone finally tackling this problem and prepared to share the process with everyone. I have tried to sort this one out in the past on my 450 and measured voltages and compared them to a few that Mark Adams had provided me. I called him to give feedback and thought he might be interested in working on the problem but never got any calls back. I guess he is a busy man these days.

I noticed a big difference in the level of snatching on Chimaeras that I test drove whilst trying to find a good one to buy. My Chimaera (1997) used to just snatch only on very light throttle, when trying to just maintain speed but would do this throughout the rev range, e.g. whether maintaining 70mph in 4th or 5th, or 30mph in 2md, 3rd, 4th or 5th. The snatching is obviously more violent in lower gears.

Opening the throttle removes the problem. I have tried replacing all the usuals and also played around with changing the orientation of the throttle position sensor, until someone told me that this makes no difference on the later cars.

The shunting has gotten slightly worse over time. Would be interested in trying any ideas that you have or providing values from my car if any help to you. I'm based in Cardiff. Would love to solve this one as it spoils the enjoyment of the car, especially as I use it as a daily driver and frequently have to use light throttles due to slower cars in front . I understand a lot of people get the problem solved by the change to the Tornado chip but would really like to solve the problem of the existing ECU rather than just replace it. I'd always be wondering what it was that had been wrong.

Edited by taylormj4 on Monday 14th April 12:30

[penno]

Hi Mark,

I have a spare 5AM AFM that was fitted to my 400 standard cam etc does give shunting at very low speeds 30 to 40 in 4th or 3rd gear. I was told that the car was running slightly lean and the cause more than likely would be the AFM thus this is why I replaced it. Now the interesting point I replaced the AFM with another secondhand unit and set the trim voltage as per RPI guidelines and the shunting has improved ie not noticable. But now the car doesn't "seem" to pull aswell as the old AFM but the drive is a lot smother. If you want to play with this AFM you can.
Just one thing to note if it's wet raining and cold then the shunting tenders to be worse.

Penno