Donations towards server fund so far this month.

 
£0.00 / £100.00 per month
Page:
Home > A-Series EFI / Injection > Megasquirt Traction Control

Rod S

User Avatar

5988 Posts
Member #: 2024
Formally Retired

Rural Suffolk




On 18th Feb, 2010 Paul S said:
Well, I'm not actually convinced that the 2917 has an internal zener. They have drawn one in one of the schematics, but I think it is suposed to be external.


That's because the datasheet is cr*p (I think).

The schematics on the datasheet jump all over from the different versions, hence my confusion earlier, but I now think Joe is right, the 2917 just needs the external resistor.

No more snow here yet but forecast really bad for Sunday night and Monday...... *frown*

Schrödinger's cat - so which one am I ???


Rob H

4314 Posts
Member #: 700
Formerly British Open Classic

The West Country

On 17th Feb, 2010 robert said:
is it worth making 5 volts at just above possible spin speed ,say 50 mph ? this would get a greater sensitivity for launch ,then have the system turn off above 50 mph .bit of a different philosophy ?
On 17th Feb, 2010 Paul S said:
Interesting idea, but we would need the code re-writing to add this facility.

You would also need a facility to stop the converters feeding the MS. It would be 12v at 120 mph and a toasty MS.


Could you not use a 741 Op Amp with a gain of one feed from a 5v supply between the frequency to voltage converter & the MS that way if the output of the LM2917 goes above 5v the 741 will hit the power rails and start clipping. Wow I never thought I'd ever want an amplifier to start clipping!

On 18th Feb, 2010 Paul S said:
Would a 7809, 1A Positive Voltage Regulator do the job better?

http://www.maplin.co.uk/module.aspx?moduleno=7937

Would I need any capacitors for extra filtering?


I always uses a couple of caps with 78 series regulators.

EDIT: Corrected a typo, no doubt there's several more in there

Edited by Rob H on 18th Feb, 2010.

Isambard Kingdom Brunel said:
Nothing is impossible if you are an Engineer


Rod S

User Avatar

5988 Posts
Member #: 2024
Formally Retired

Rural Suffolk

On 18th Feb, 2010 Rob H said:

Could you not use a 741 Op Amp with a gain of feed from a 5v supply between the frequency to voltage converter & the MS that way if the output of the LM2917 goes about 5v the 741 will hit the power rails and start clipping. Wow I never thought I'd ever want an amplifier to start clipping!


That should work nicely and allow higher voltages (hence better resolution) at lower speeds.

Schrödinger's cat - so which one am I ???


Joe C

User Avatar

12307 Posts
Member #: 565
Carlos Fandango

Burnham-on-Crouch, Essex

yes you need caps on the input and output Paul,

let me know your address and i'll sort you out some 9vregs and caps *wink*

probably got some zenners and things too if you need them.

On 28th Aug, 2011 Kean said:
At the risk of being sigged...

Joe, do you have a photo of your tool?



http://www.turbominis.co.uk/forums/index.p...9064&lastpost=1

https://joe1977.imgbb.com/



Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland


On 18th Feb, 2010 Rod S said:
On 18th Feb, 2010 Rob H said:

Could you not use a 741 Op Amp with a gain of feed from a 5v supply between the frequency to voltage converter & the MS that way if the output of the LM2917 goes about 5v the 741 will hit the power rails and start clipping. Wow I never thought I'd ever want an amplifier to start clipping!


That should work nicely and allow higher voltages (hence better resolution) at lower speeds.


I think that idea will be worth investigating if we find that we do not have sufficient resolution with 0-5v over the full speed range. I'm struggling with all this electronics mumbo-jumbo as it is.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


Brett

User Avatar

9502 Posts
Member #: 1023
Post Whore

Doncaster, South Yorkshire




On 16th Feb, 2010 jbelanger said:
Actually, since the code compares ADC counts (voltage), it would be possible to have different wheel sizes and/or a different number of teeth as long as you calibrate the frequency-to-voltage circuits to give the same voltage at the same speed.
Jean


i was thinking about this at work last night, would the sensor be capable of counting the crown wheel or a trigger wheel on the side of it?
or even how well would they work submerged in oil? the trigger wheel could be put in the speedo converter housing.
obviously either way would only be able to tell the front wheels was slipping not which one or by how for much for each, i am thinking of drag racing purpose's where lsd's are fitted and both wheels are spinning together

sorry if im barking up the wrong tree:)

Yes i moved to the darkside *happy*

Instagram @jdm_brett


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

Brett, the only potential issue is that you need the same numbers of teeth front and rear.

This is my collection of parts coming together:



A bit of machining to do and then tack weld the wheels on.

Edited by Paul S on 19th Feb, 2010.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


pinkyandnobrain

User Avatar

720 Posts
Member #: 2588
Post Whore

Pretoria South Africa

Paul have you checked the clearance between the trigger wheel and the rear wheel cylinder ? I had a problem with this when fitting a sensor for a rally tripmeter.

Mine only just clipped the dust cover but it was enough to cause problems.

"So wat we gonna do tonight Brain?"
"Same thing we do evernight Pinky!"
"Try to take over the world!"


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

I will check it. There is scope to machine the OD a little, but not a lot.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


Rod S

User Avatar

5988 Posts
Member #: 2024
Formally Retired

Rural Suffolk

I'm still thinking about this....

(something I often do instead of progressing forwards)

1 - I'm definately going to use Op Amps as voltage clamps (as Robert suggested) to allow better resolution at low speeds and ignore high speeds. I'll probably add some 4.9V zeners to protect the CPU if everything were to go pear shaped....

2 - Might use VR instead of Hall, the circuitry is simpler and can use the 8 pin package as the trigger offset isn't required with VR.

3 - Will definately look at using a sensor inside the gearbox for the front as per Brett - altering the R and C values for different teeth is straightforward, a pot for "R" would allow fine tuning.

4 - Don't like the 12 tooth wheels, I would prefer more teeth for better resolution al low speeds. Most OEM traction control systems seem to use the ABS signal which is derived from many more than 12 teeth..... but I can't find anything better than 12 yet that could be fitted to our antiquated design.....

Just rambling at the moment....

Schrödinger's cat - so which one am I ???


jbelanger

1267 Posts
Member #: 831
Post Whore

Montreal, Canada

Rod,

Using a Zener on an ADC input is not a good idea. The Zener will start conducting way before it's rated voltage and will skew the results: anything above about 4V (maybe less) will be unreliable.

Jean

Edited by jbelanger on 19th Feb, 2010.

http://www.jbperf.com/


Rob H

4314 Posts
Member #: 700
Formerly British Open Classic

The West Country

It's stupid question time again but do you really want to be measuring one front wheel and one back wheel? or would you be better measuring the two front wheels? as I would have thought the differential would lead to one of the driven wheels to start slipping before the other one, even with a LSD or ATB.

Isambard Kingdom Brunel said:
Nothing is impossible if you are an Engineer


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland


On 19th Feb, 2010 Rod S said:
I'm still thinking about this....

(something I often do instead of progressing forwards)

1 - I'm definately going to use Op Amps as voltage clamps (as Robert suggested) to allow better resolution at low speeds and ignore high speeds. I'll probably add some 4.9V zeners to protect the CPU if everything were to go pear shaped....

2 - Might use VR instead of Hall, the circuitry is simpler and can use the 8 pin package as the trigger offset isn't required with VR.

3 - Will definately look at using a sensor inside the gearbox for the front as per Brett - altering the R and C values for different teeth is straightforward, a pot for "R" would allow fine tuning.

4 - Don't like the 12 tooth wheels, I would prefer more teeth for better resolution al low speeds. Most OEM traction control systems seem to use the ABS signal which is derived from many more than 12 teeth..... but I can't find anything better than 12 yet that could be fitted to our antiquated design.....

Just rambling at the moment....


Rod interesting points, but I'm struggling with the electronics mumbo-jumbo and will stick with Patrick's initial design for now. Maybe develop it in the future, but I have a few points.

1. Surely it would be better to have TC on at all speeds. High speed losss of traction, for whatever reason, can be very dangerous.

2. Aren't VR sensors more prone to noise?

3. Running off the crwon wheel would be great as it will do the averaging.

4. Let's see how it performs at 12 teeth, Patrick's logs show that it is registering spin well enough.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


Rod S

User Avatar

5988 Posts
Member #: 2024
Formally Retired

Rural Suffolk

On 19th Feb, 2010 jbelanger said:
Rod,

Using a Zener on an ADC input is not a good idea. The Zener will start conducting way before it's rated voltage and will skew the results: anything above about 4V (maybe less) will be unreliable.

Jean


My fault (as usual !!!).

A lot of the MS-Extra circuits seem to use 4.9V or 5V zeners as protection but they are on digital inputs not analogue.....

So either rely on the Op Amp as a clamp or try to restrict the LM29*7 output voltage by other means ???

Back to the Datasheet.......

Schrödinger's cat - so which one am I ???


jbelanger

1267 Posts
Member #: 831
Post Whore

Montreal, Canada

If the Op Amp is rated for it, it should be ok. Remember that you want a rail-to-rail OpAmp since you want the output to cover the whole range.

Jean

http://www.jbperf.com/


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

Rod,

Just think about the scenario of getting to, say, 70 mph on a wet road and full chat and the traction control turns off because you've reached the 5v frequency.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


Rob H

4314 Posts
Member #: 700
Formerly British Open Classic

The West Country

On 19th Feb, 2010 Rod S said:
So either rely on the Op Amp as a clamp or try to restrict the LM29*7 output voltage by other means ???

Back to the Datasheet.......


Just had a look at the data sheet and as far as I can tell there is no minimum Vcc so is there any reason why you can't use a 5v supply to the LM2917?

Isambard Kingdom Brunel said:
Nothing is impossible if you are an Engineer


Rod S

User Avatar

5988 Posts
Member #: 2024
Formally Retired

Rural Suffolk

Rob (1) I was assuming measure all four wheels. The LM29*7 datasheet shows how to wire them for maximum, ie, if you measure both the rears it will ignore a lifted inside wheel. For the front(s) it would seem more logic to measure the average which again can be done electronically with a pair of sensors or mechanically if measuring at the diff/gearbox.

Paul,
1 - High speed, possibly more dangerous but I'm not aiming for massive power so would prefer to optimise it for leaving the traffic lights (or the pretty coloured lights on the drag strip for the rest of you !!!)
2 - VR may be more noisy for precise timing events but the LM29*7 uses very basic principles for its frequency to voltage conversion so I would think not much of an issue. All the ABS sensors I've unplugged are only two wire from memory so are most likely VR.
3 - Running off the crownwheel/gearbox, as Rob (1), gives the average, also with pot joints (my case) its hard to see how to mount a trigger wheel. Also, I've never seen an output shaft without play even with brand new bushes in the cover plates so keeping a uniform gap could be an issue, less so with hall as they have quite a distance range as you have proven, but VR would be a problem as it likes a nice small gap.
4 - 12 teeth is probably a good compromise for all speeds but mathematically poor at low speeds. Again, I was thinking of how to optimise it for the traffic lights.

Jean, I have no practical experience of using an Op Amp as a clamp, although the theory is sound, but I'll mock it up first and see.

Rob (2) - Yes, I had thought of that hence my referring to it as LM29*7, ie using the 2907 without the internal regulator. Just need to see what the losses in the output stage are.

Just need to find my old breadboard and reconfigure my little USB scope as a function generator to see how this works. The electronic components are cheap, it's the hardware (trigger wheels, sensors) that are pricey.

Schrödinger's cat - so which one am I ???


robert

User Avatar

6745 Posts
Member #: 828
Post Whore

uranus

nonsense post deleted.

Edited by robert on 20th Feb, 2010.

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM


Rob H

4314 Posts
Member #: 700
Formerly British Open Classic

The West Country

On 20th Feb, 2010 Rod S said:
I was assuming measure all four wheels. The LM29*7 datasheet shows how to wire them for maximum, ie, if you measure both the rears it will ignore a lifted inside wheel. For the front(s) it would seem more logic to measure the average which again can be done electronically with a pair of sensors or mechanically if measuring at the diff/gearbox.


I'm not sure is averaging is the way forward, as it may mask the fact that one wheel is starting to slip, would be fine if both the front wheels where slipping mind.

On 20th Feb, 2010 Rod S said:
Running off the crownwheel/gearbox, as Rob (1), gives the average, also with pot joints (my case) its hard to see how to mount a trigger wheel.


A bit of food for thought:

http://www.turbominis.co.uk/forums/index.php?p=vt&tid=214850

On 20th Feb, 2010 Rod S said:
Jean, I have no practical experience of using an Op Amp as a clamp, although the theory is sound, but I'll mock it up first and see.

Just need to find my old breadboard and reconfigure my little USB scope as a function generator to see how this works. The electronic components are cheap, it's the hardware (trigger wheels, sensors) that are pricey.


I just had a look in my bits box and I've got a few 741s so I'll see if I can knock something up this evening.

On 20th Feb, 2010 robert said:
nonsense post deleted.


I saw what you wrote, leave Mr Stick out of this!

On 18th Feb, 2010 Paul S said:
I'm struggling with all this electronics mumbo-jumbo as it is.


You're lucky I was born with "the gift"

Edited by Rob H on 20th Feb, 2010.

Isambard Kingdom Brunel said:
Nothing is impossible if you are an Engineer


robert

User Avatar

6745 Posts
Member #: 828
Post Whore

uranus

I saw what you wrote, leave Mr Stick out of this! *surprised* *laughing*

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM


Rod S

User Avatar

5988 Posts
Member #: 2024
Formally Retired

Rural Suffolk

On 20th Feb, 2010 Rob H said:
On 20th Feb, 2010 Rod S said:
I was assuming measure all four wheels. The LM29*7 datasheet shows how to wire them for maximum, ie, if you measure both the rears it will ignore a lifted inside wheel. For the front(s) it would seem more logic to measure the average which again can be done electronically with a pair of sensors or mechanically if measuring at the diff/gearbox.


I'm not sure is averaging is the way forward, as it may mask the fact that one wheel is starting to slip, would be fine if both the front wheels where slipping mind.


The way I look at it, whichever of the front wheels starts to slip, the non-slipping (or slipping less) one can never be slower than the (fastest) rear wheel so if you pick the fastest of the front wheels or the average, it will always be higher than the rear so you have your error signal.

With an open diff, if all the slip is on one wheel and you measure the fastest the error will be exactly twice than if you measured the average. With a good limited slip diff, the "fastest" should be fairly close to the average.

So it all depends on how much of an error the sofware requires, ie, how responsive it is. Hence I'll mock it up on a breadbopard and "test" it with a signal generator. The software seems to have installed on the MS fine so I can watch the results on TunerStudio once I've put the chips together.

I remember the other thread from a while ago, re-reading it it seems Jimster was the only one to make the same point as me about play in the housings messing up the sensor gap.

Schrödinger's cat - so which one am I ???


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

Rod, if you buy the latest minimag, you will se a picture of a toothed wheel on a pot joint and what is most likley a VR sensor.

Plus what looks like a load of swarf sticking to it *oh well*

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


Paul S

User Avatar

8604 Posts
Member #: 573
Formerly Axel

Podland

On 20th Feb, 2010 Rod S said:

1 - High speed, possibly more dangerous but I'm not aiming for massive power so would prefer to optimise it for leaving the traffic lights (or the pretty coloured lights on the drag strip for the rest of you !!!)


I'm not at all happy with it cutting out at any speed in our lightweight Miglia with potentially 200+hp (and my lad driving it!)

On 20th Feb, 2010 Rod S said:

3 - Running off the crownwheel/gearbox, as Rob (1), gives the average, also with pot joints (my case) its hard to see how to mount a trigger wheel. Also, I've never seen an output shaft without play even with brand new bushes in the cover plates so keeping a uniform gap could be an issue, less so with hall as they have quite a distance range as you have proven, but VR would be a problem as it likes a nice small gap.


The beauty of the 12 tooth wheel is that the sensor can be fitted either axially or radially. If the radial whirl of the pot joint is a problem, the sensor could be mounted axially.

Edited by Paul S on 20th Feb, 2010.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."


Brett

User Avatar

9502 Posts
Member #: 1023
Post Whore

Doncaster, South Yorkshire




On 20th Feb, 2010 Paul S said:


Plus what looks like a load of swarf sticking to it *oh well*


i saw that too, wasnt it the SC car? with the ecu from them. Maybe they are mustering up a kit for pot joints

Yes i moved to the darkside *happy*

Instagram @jdm_brett

Home > A-Series EFI / Injection > Megasquirt Traction Control
Users viewing this thread: none. (+ 4 Guests) <- Prev   Next ->
To post messages you must be logged in!
Username: Password:
Page: