Autosport Labs

Boy and girls
Has anyone out there used the method of using relays to switch the coils off and also a digram and what type of relay to use would be a masive help indeed.
Cheers
Rich

I have been trying to make a rev limiter, triggered by the rev limit output of mjlj. My design works by grounding the coil -ve, through a resistor to limit current. The way this is switched should avoid poorly timed and potentially damaging spark events that could be caused by a relay cutting power to the coils at a random time.

Unfortunately the rev limiter output from mjlj switches on and triggers the limiter whilst the engine is cranking. This obviously prevents the engine starting. In its current state the rev limit output does not appear to be usable for it's intended function. I plan to explore the behaviour of the other programmable outputs to see if there is a way round the problem, but haven't found time yet.

Great idea on grounding the -ve side of the coil. Try the beta firmware I posted in the files section, it should solve your issues during cranking.

Brent

Brent, MJLJ works well with the existing firmware and has completely transformed my car (Westfield, 2l Pinto, Dellorto 45s with tps). I have been a little reluctant to upgrade until you give the firmware the "all clear". Are you happy that there are no major bugs in the current beta version?

I've been running the beta for a long time now. I'm running it regularly in my race car.

The final version will be very close to the beta version. I would certainly give it a try!

Brent, how long will it be before the final version is released. I don't have a laptop so have to remove the unit from the car, power it on the bench and programme from my desktop pc. I would prefer to update the firmware only once. If you are planning to release the final version soon, I will wait.

I think the current version he has should be made final. You can buy 100 feet of serial from eBay for $20, and program from your house. You will need a laptop if you want to program the ignition map. There is no point in having megajolt if you can't program it.

It is already programmed. Changing things is not too difficult but I would prefer not to do it more often than absolutely necessary, hence the question.

Hi Dave,

I was also really interested to read your thoughts on using relays to ground the -tive side of the coil. I'd like to hear more about this especially as with the new (beta) firmware there shouldn't be the cranking issue.

Couple of key points:
1) Why ground out the -tive, couldn't you just cut the connection from the EDIS unit (using a normally closed relay) hence when the rev limiter was activating no spark would be triggered?

2) Assuming there’s a reason why the above is ludicrous, why/what are the considerations of using a resistor in the circuit?

3) Why do you think 'mis-timed' cuts could be damaging (or at least any more so than any rev limiter?

4) Have you got any thoughts on how the system could be made at all progressive? Obviously the other PGMs could perhaps be utilised through some kind of simple circuit to progressively cut ignition events with more frequency. Donno what you think of that?

Anyway, as said, I'd be keen to work on this a bit as it occurs to me that something ought to be pretty easy to rig up.

Chris

Brent - Taking a different slant - what is the main reason behind not implementing a limiter in the software itself? I know it’s been mentioned before but I've never seen anything conclusively saying this couldn't be done?? Providing there’s no major constraint I'd be keen to have a look at this as well as I am IT Consultant with a fair level of programming experience myself.

Chris,

There are a number of possible ways to implement a rev limiter with MJLJ and EDIS. The most practical are to i)interupt the +12V to the coils, ii) interupt the coil -ve, iii) ground the coil -ve.

If the + 12v or coil -ve is switched during the dwell period of one of the coils, this will probably generate a spark, which, if generated, must be more advanced than the intended timing at that rpm. This will cause detonation, possibly at max rpm and load. It is conceivable that this will damage the engine, particularly if repeated.

Grounding the coil -ve causes current to flow through the coil. When the electronic switch in the EDIS opens, to generate the spark, the current flowing to ground prevents the magnetic field collapsing, and inhibits spark generation.

An EDIS coilpack has a resistance of between 0.5 and 1 ohm which means a current of up to about 25 amps could flow through the coil if steps were not taken to limit it, which is the purpose of the resistor. My experimental limiter is presently using an 8 ohm resistor in series with each coil which seems to supress the spark effectively.

Of course, with all that current flowing, the grounding of the coil has to be released during the normal dwell period so that a mis timed spark event doesn't occur as the limiter disengages.

My experimental circuit is using a pair of thyristors to do this switching. When a thyristor starts conducting, it won't stop until the voltage across the anode/cathode falls so low that the current through it goes below a holding limit, at which point it switches off. Conveniently, the only time this occurs in the ignition cycle is during the dwell period, when the EDIS is effectively grounding the coil. This seems quite neat, as the switching logic is provided by the switching device itself, so no additional components.

The only drawback I can find to this circuit at the moment is the short duration trigger (about 150 msec) of the rev limiter output during cranking. If the engine stops turning during this time, the thyristor will continue to conduct, putting quite a large current through both coils which may cause thermal damage. This state will continue until the ignition is turned off. It would be possible to disable the limiter for a set time after starting the engine to get over this problem, but it is not an elegant solution and requires extra components.


To make the limiter progressive it would be possible to retard the ignition to 0 degrees above a limit set in the ignition map. Beyond that, you could use two programmable outputs and ground one coil at the set rpm, causing a two cylinder mis-fire, the other at a slightly higher rpm to limit the other two cylinders.

David,
Sounds like you have this pretty well thought out. Would you mind drawing a diagram of your circuit and posting it to the file section along with a parts list? I and others would appreciate it.

Thanks, Todd

One of those situations where the theory is slightly ahead of the practical design. Until today, it didn't work because of false triggering and the thyristors latching in the on state, which stops the engine. A rather extreme form of rev limiting!
I think I have resolved all of the problems but a bit more testing is required before I publish the circuit.

The triggering while cranking is clearly wrong, and I'll be working on patching the latest beta to fix this issue.. Hopefully I can get it done this week.


Brent

Why not reduce the time to 10ATDC. This should seriously stop the engine accelerating along with giving the driver a big warning.

Killing the coils will load the cylinders with unburned fuel that may cause poor running once the engine tries to fire again.

I like the idea of a progressive retard on the advance ending at 10ATDC in the basic firmware. Then, if you want the hard cut, you could go the next step with the coils.

Lee

I agree that ignition retard followed by hard cut would be the best solution. At the moment it requires the sacrifice of two columns in the ignition map to generate the retard. I haven't had any obvious problems when the engine starts to fire again, but testing has been quite limited.