Originally Posted by
kur4o
I am back on the drawing board. Rewiring the coils to match the injectors will be pain, so I guess I will have to do it or live with the consequences. It will be interesting to do a force open loop in the event a injector or coil to fail.
No offense but I think you're overthinking things a bit. Granted - it would be better to have the ability to limp home, but the car's gonna run like crap. Seems like a lot of extra wiring for something that should never happen (unless a wire gets pinched).
Originally Posted by
kur4o
Did you thought about the ltcc to listen at the aldl stream and extract some data from it.
The only parameter I can think of that would be marginally useful that we couldn't easily measure with the controller is coolant temp. And I'm afraid the overhead of parsing the ALDL data could potentially cause starvation in the main loop which could cause missed or delayed dwell. It would be much easier to just read ECT with one of the unused ADC inputs. But I'm not convinced temperature compensation is necessary. Remember - the Bailey LTCC kit only connected to the opti and EST lines. It's unknown whether it compensated for operating voltage.
Originally Posted by
kur4o
d585 are known to self ignite on longer dwell time.
I fully intend to test and document dwell limiting behavior. In my initial tests I actually found that the D514a coil seemed to discharge a lot when it wasn't expected to. I didn't notice this with the D585 or the 8183. But I didn't test that function methodically - this is just something I observed casually.
Originally Posted by
kur4o
I am sure GM data is good starting point and there will be very limited tweaking over there.
I'm not as convinced as you are.
I spent a good bit of time one night last week looking over the bin dumps you posted back here [link]. It raised more questions than it answered. For instance looking at the main dwell table for the 98 f-body ls1 and the 02 ls6 they seem to match exactly.
For a frame of reference, here's the 2000RPM dwell row.
Code:
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
2000 26.00 26.00 26.00 26.00 26.00 26.00 26.00 17.29 8.60 7.29 6.00 5.24 4.50 4.04 3.60 3.30 2.99 2.99 2.99 2.99 2.99 2.99 2.99 2.99 2.99 2.99 2.99
Then there's a dwell multiplier on coolant temp that's also voltage compensated.
Code:
-40 -20 0 20 40 60 80 100 120 140
13 1.00000 1.00000 1.00000 1.00000 1.04004 1.11499 1.14502 1.17993 1.20996 0.99512
Just looking at the results of the 13v range, this gives a minimum dwell below 40c of 4.04ms, increasing with coolant temp to a maximum of 4.89 at 120c and then tapering back at 140c. Based on the megasquirt info here they're calling for a target of 5.6-5.8ms, yet in the GM tunes that never happens except at voltages below 10. Baffling.
This same range in the 2014 vette tune (13v@2000rpm) has a minimum dwell of 2.88 and max of 3.2, also increasing with coolant temp up to 140c where it seems to be reducing as a protection mechanism for an overheating condition. I presume this application would have been equipped with the 8183 coils.
Moving on to the 2004 Yukon tune that would have had the D585 coils the same range has maximum dwell at 20c of 4.59ms and then drops to a floor of 3.52ms at 60c and stays there. This is completely opposite of what the other temperature compensation tables do.
Then there's a dwell vs MAP multiplier table in the LS6 and the Yukon tune, but every cell is effectively set at 1 (0.9999999). Actually the Yukon bin dump has suspect data in this table - 8.1567373 from 20 to 70 kpa then 0.9999999. And there's no reference to dwell vs MAP in either the f-body LS tune or the 2014 vette.
Looking at the differences and commonalities, it's not terribly clear what the function of the temperature compensation is.
The only universally common data here is:
1) dwell is increased as voltage decreases
2) dwell is reduced with RPM
This last item is patently baffling to me. In a coil per cylinder setup you have the ability to increase dwell at higher RPMs, but these tables seem to mimic how a distributor setup would be mechanically dwell limited.
I'd like to think "surely there's a reason GM did this" but I'm not necessarily convinced.
Originally Posted by
kur4o
I was looking for various connectors and they are really overpriced on ebay. I have to dig some part numbers.
If you wanted to make your own harness, I recently bought a set (8) of the newer LS2 style connectors / pigtails off ebay for $25 shipped here.
I have a 120w 20v power supply (about 6.2 amps) for an all-in-one PC that I think will probably suffice. But I'll initially test primary current with a battery so I know what to expect. My guess is the D585 coils will have the highest primary current, but we'll see.
Bookmarks