Microsquirt v3 ign out woes

[jonfx4com]

Confused about v3 (plastic case) ign outs

Read this in the docs

"Spark Output (spkout_hi_lo) 0 equals spark when low (ground). V3 MicroSquirt®'s do not have a transistor for ignition control on the daughter card. As a result, if you are using a V3 MicroSquirt-II®, the Coil Charging Scheme may need to be reversed compared to some earlier units."


So, if it does not have a transistor in the output should I assume it's a 5v TTL output rather than an open collector?

Trying to fire my mx5 3 wire coil packs which need the trigger sat at 0v then up to 12 to charge and back to 0 to spark and not having much luck. The packs have inbuilt igniters which seem to have a pull down resistor of 415 ohms. I have added pull ups of 330 ohms but have no idea if I need them or if the output should or should not be inverted on a V3 microsquirt or indeed what the output circuit looks like. Any clues would be appreciated. Can't find any microsquirt V3 circuit diagrams yet so I'm rather lost with this setup.

One more question, is a V3 MicroSquirt-II referred to above a microsquirt V3, ie the new one in the plastic case or a revision of the metal cased microsquirt-II?

[jonfx4com]

Update

My coil packs are fried..

I scoped the ign out of the microsquirt and it does seem to be TTL. Tuner studio is set at ign out going low (normal) and with the engine not running and no pull ups the outputs are sat a 5v. Surely this is wrong? If the V3 does not have an extra transistor inverting the output then the output should be exactly as tuner studio shows? My logic is than spark going low means stay low, go high for the dwell period then go low to spark. Am I going mad or does the new V3 still do the opposite of what TS shows in terms of the spark output?

[grippo]

We are going to post the v3 schematics shortly, but I will attach them here to help solve your immediate problem.

[jonfx4com]

thanks Grippo

[jonfx4com]

Now working perfectly. I replaced the Mazda coils with the built in ignitors with a ford coil pack and a bosch dual ignitor. It is now running very well. Using FIdle to control the standard PWN idle valve. Had to make a mod here, added a flyback diode. The idle control had very little span until I did that.

Did not use the Mazda trigger wheel as I found a 12-1 replacement that fits directly in place of the OE wheel. The original crank sensor works just fine with the new wheel. This MSQ runs pretty much like the OE system and has fan control on the WU Led port and intake geometry control on the Acc LED.

The car is a MX5 (Miata) 1.8 2000 model.

Points to note:

I hear that the US models do not have a manifold referenced FPR but have fixed fuel pressure and the reg is in the tank assembly. Consequently this MAP may need some tweaking for those cars.
The original coil packs did not like the Microsquirt at all. They died without ever creating a spark. Since they were useless and very expensive to replace I never got to the bottom of what went wrong.
The fan wire is not where the Mazda drawings show it on the ECU connector, in common with many of the other wires!
The best way I found to make the harness was to buy an old Mazda ECU and cut the connector of the board, minimal costs and a nice install.
You will see a weird ignition offset number and what looks like a very excessive idle advance in the MSQ. I suspect that tweaking one will fix the other but I ran out of weekend. Meanwhile it starts just fine, hot or cold and drives well.
The NB lambda control works well and Tuner Studio did a good job of filling in the low to middle load parts of the table.

Next step is to fit the bigger injectors and then comes the turbo.

[jonfx4com]

Did some more thinking about my timing weirdness.

MX5 crank sensor, 12-1 wheel, The OE MX5 sensor is not true VR it has a straight square wave output. Normally sits at 5V and drops to 0v when a tooth passes.

How should I configure this in Microsquirt V3? It runs OK as it is with ignition inout capture set to rising edge but my advance figures look weird and I'm beginning to question what is going wrong.
As it stands I'm idling at an advance figure way higer than I would expect (25 Deg as opposed to the OE figure of 12 Deg) also the onset of detonation at moderate throttle and say 3000 RPM is around 24 deg. I'm wondering if I'm counting gaps not teeth?
Does rising or falling edge even matter with a 0-5v digital input as opposed to a VR? Will it just mean a timing error of the width of a tooth, corrected in the offset number?


All advice greatfully recieved.

[grippo]

Rising or Falling edge should not matter with a square wave vr input. The code triggers off whichever edge you pick and never sees the other edge. (The processor can be configured to see both edges, but it's not.) So it can't see gaps or teeth sections, it only sees rising edge to rising edge or falling to falling. However, changing the polarity will result in a timing difference equivalent to the gap (or a tooth section), but as you said, this can easily be compensated for by changing the trigger offset.