DFI Module

[Daniel.l87]

This is a continuation to my questions in the Tuning topic but this is more specific to the GM DFI Module so thaught it better I put it here.

VX commodore, L67, I have heard that the blue base coils past 5500rpm start to struggle a bit...i dont know how true this is as i have had it out to 6300 and it appeared at least to be ok, however i wasnt monitoring the coil outputs at the time. I am looking to take the revs well past this anyway.

I have been through the Holden documentation and can't seem to find the info I need, maybe cause its not something you would really have to know normally, but does anyone have info on the maximum dwell of the standard coils, how they are triggered, Rising or falling edge? and also how they are charged, constant?

The reason I ask all this, I want to relocate and rewire the DFI module to drive a set of LS2 coils which should cope quite nicely with the high rpm and boost.

[VL400]

The ECU only has a single output called EST - electronic spark timing. Its a rising edge to start dwell and falling to fire the plug, the ECU manages the minimum burn time at about 600us. The DFI module manages which coil to charge. With the single ouput dwell can be fairly short at higher RPMs. I dont know for sure (but doubt it) if the DFI coil pack can extend dwell to begin charging the next coil before a previous has finished firing like a true sequential/wasted spark ECU will.

I cant recall if its LS2 and LS3, but certainly LS3 coils protect themselves on charge time. So if you command over about 4ms it will fire early. The delco ECU will put out nearly 7ms at idle and more than that under acceleration at low RPMs.

[Daniel.l87]

Excellent explanation, got a much better idea now and I think given that description i can make those coils work.

The LS2 smart coils i believe as you have described also have the a 'safety' function to stop excess dwell, some documentation i read this morning suggests that also but does not offer the actual figures. Maximum dwell apparently is around 6ms with nominal at about 4.5ms but it might not fault until well after this.

I have some great info on the LS1 coils which i will go through tonight but something worth mentioning is that in the Holden documentation on the DFI module, it states that the dwell is variable, controlled by the PCM and reliant on engine load & rpm....this would suggest there is a table?

If so & if this could be programmed, then there would be no issues implementing this. Alternatively, having this fixed at 4.5ms would be acceptable also as this is still far in excess of the OEM coils.

[VL400]

4.5ms nominal is better and 6ms max is better, will get it closer to what the factory ECU does. Can you post up or link to the LS1 coil info?

I think it would require a code patch to get it to be a fixed value or limit it in the ECU, there is no table, its a dynamic calc that follows some fixed items.

Did a quick calc on a 6 cyl and by 6500RPM your best case dwell time with 0.6ms burn is 2.5ms. At the 8K RPM your wanting its down to 1.9ms which is a big ask, the LS1 coil energy is getting much lower by then when testing it on the bench a few months back - my test was hardly scientific though, was just seeing what gap it could jump at various dwell times.

Dwell Calc.xls

(13.5 KiB) Downloaded 400 times

[Daniel.l87]

Don't know if it makes a difference but the ls2 coils are quite a bit different, they run less dwell and slightly shorter spark duration however they deliver triple the output, 120 vs 40. My understanding is that as a result they are far superior especially when you throw revs at them. I have 8 at home purchased for another project which I am looking to use so if you would like a sample unit I could arrange that if you wanted to do some more reasearch for your own benifit?

Happy to pass on the info, will do later tonight, as alot of aftermarket EMS systems use these coils theirs a fair bit of technical info on them.

[VL400]

I knew they were better but thats a huge difference, will grab some and have a play

[Daniel.l87]

Final question, as stated in my earlier posts, wiring these after the trigger on the DFI module, i.e to the posts which stick out and connect to the base of the OEM coils should work? Am I correct in saying the trigger is just a solid state relay designed to cope with the high amperage current required by the coil, that when switched off by the module, drops the earth supplied to the coil to trigger it? If so even though it would be better to connect to the module output directly prior to the trigger, realistically this should not result in much if any latency and the signal shape should the same?

I then would supply constant power to the LS2 coils via the EFI Relay and high amperage wire to the 'coil' side of the LS2 coils and earth this circuit to the block. The 'power' side of the DFI module coil output would go the positive side of the 'trigger' of the LS2 coil which effectively controls dwell given this is the charge time of the coil and the earth side of the DFI module coil output to the 'signal ground' of the LS2 coil to trigger it?

[VL400]

You will likely need to invert the signal and convert it to a 0-5v signal the coil requires. The DFI output gnds to start dwell and goes high to fire, one pin has ign switched 12v and the other the switching ground signal. The LS1 coil needs 5v to start dwell and drops to 0v to fire, its a logic level signal and only needs a few mA. I dont think its going to be possible to switch the signal ground of the LSx coil directly using the DFI output post.

[Daniel.l87]

Oh well that can be my next project to work out how to do that...doesn't discourage me in the slightest, it would be a great conversion if i can get it to work and solve a lot of potential issues in the future.

[VL400]

Yeah should make a difference

And just to clarify the part "The DFI output gnds to start dwell and goes high to fire" - it does not get pulled high (will just be a FET or something as an open collector type circuit), will need a pullup resistor on your circuit to get it going high. Simple circuit with a 74ACT540 should do the trick, they drive LS1 coils nicely