[porschenstein]

Hi pazi88,

Yes - I'm inputting 1.7V as the battery input. This satisfies TS as the Vbat warnings go away - I'm assuming that's sufficient.
I've also set all the temp, voltage, load, etc compensations to 0 - which appears to be the default. And I setup a spark table with 10 degrees in all cells.

I did not setup the ve table - could that be causing a problem?

On the logic analyzer, I don't see a consistent firing order. I expected to see 1 thru 6 sequentially in time. However, each time I trigger a capture I get a different order (triggering on the falling edge of coil 1). This is more disconcerting than than the dwell. I also verified it's not an analyzer sampling issue by changing the sample rate.

Thanks.

[LPG2CV]

battery reference should be around 12v

[porschenstein]

I'm directly driving the processor board (no speeduino pcb) so the voltage has to be reduced to the processor ADC's input range. A value of 1.7V prevents the battery voltage warning which indicates the processor is good with it.

[LPG2CV]

I suspect this is because on the pcb, the relevant board circuit, reads the 12v supply to the board. The same supply is fed to the coils and injectors. By sampling the 12v supply, the mcu/FW is able to provide the corrections for injectors and coils.

Without the 12v, it can't work properly, and you will get weird results.

[porschenstein]

Here's a snapshot of what I'm seeing. On the bottom you can see the crank and cam inputs (@ 1k rpm). Notice how the coils (top 6 traces) are running much slower and not in sequence...

[porschenstein]

From above, the rpm is being correctly and consistently reported in TS based on the crank and cam signals. Somehow, that timing is not making it to the coil sequencing or the computed schedule is a fraction of what it should be.

At first I thought that maybe the crank (PD3) and cam (PD4) signals were swapped (or something weird like that), but I don't think the indicated rpm in TS would be correct in that case (???). I've verified they're connected to the correct pins.

I also turned on the option for "locked timing" at 10 degrees and that didn't change anything.

Is it possible that the parameters are not getting stored correctly (even though they read back correctly after a power cycle)?

[PSIG]

Post a current tune *.msq for relative settings. I would add a digital channel for TachOut, so you see what the ECM thinks one rev is, and relative to input (consistent decoding) and relative output. <shrug>

[porschenstein]

Hi PSIG,

I've attached the .msq file as requested. I hope you can figure something out because I'm drawing a blank. I don't know if will help but the coil period is about 3 times what it should be, and the dwell is about 10 times too long. I put the dwell on one of the TS gauges and it constantly shows 0. I'll hook up the tach out and capture a new output.

Thanks for your help.

[LPG2CV]

A heads up. This generally applies to fuel and spark. 100% = no correction. ie If Fuel = 100, then Fuel x 110% = 110.
However, Fuel x 0% = 0

For testing, set dwell compensation across the board to 100%. 100% = no correction applied. 110% = 10% increase. 90% = 10% decrease.

This may not resolve the issue you are having, as I still think you need to sample the 12v supply. Or, at least a supply to the pin that samples what the coil would receive. Probably best not to supply an unresisted supply to this pin.

Do you have a schematic for a STMF407 Speeduino board?

[LPG2CV]

Of course you do, there is a link to the github on the first page