AbeFM

iTrader: (3)

I just set the number for "8". I think it should be ~9 or 10, but this wasn't an option when I set it up. It's worked well for a year or more, so I'm not too worried. MS people blame my lack of luck with closed loop idle on the set up, but the PWM Warmup works excellent. It's effectively what you did, only, it varies with coolant temp. Highly recommended.

Huh. Sure wish I knew why. I imagine it uses the same, edge detection code....

Dunno. I would imagine Mazda put the knock sensor in a good place for knock sensing. Certainly it's as equidistant from all cyl as one sensor can be, and height-wise it's ideal.

I had a knocksense (non MS version) which I sold and went over to this set up. No regrets. Just pulled some timing today on my lunch because of it, and the car was SO much more fun on the drive home. :-)

Do you use the "aggressive" settings?

NO NO NO NO NO!

NO CAS! Worst idea ever.

Sure, using any sort of a crank trigger (Mazda OEM included) you should be able to run at least 2* more timing, safely.

Good luck with the parallel set up. I hate it myself, but don't have your.. special constraints. :-) It'll open up more IO which is a plus

Zaphod

@Ben:

I just cross-checked your spreadsheet for the 99 and (apart from the fuelpump pin - which will be 4P for the European models - and I suppose for all with factory immobilizer) I stumbled across your injector wiring.

You have 4W, 4X (OEM ECU 3W,3X) for INJ 1 and 4Y,4Z (OEM ECU 3Y,3Z) for INJ2.

As I have my megasquirt running (and the "How to Megasquirt... thread shows) I have 3W,3Y on INJ1 and 3X,3Z on INJ2.



Who is doing it wrong?

Ben

iTrader: (33)

Car's metaphorically on cylinder blocks right now. I'll let you know next week.


All makes sense. I have a good mind to go ahead and pop on this trigger wheel.
What, you don't have freakin' emissions?

I honestly didn't need any more I/O. Plus there's a CANBUS enabled I/O expander board coming through the pipeline.

I just want the bitch to pass emissions.

Ben

iTrader: (33)

We're both correct. It's batch fired. One way might idle better than the other, but under any sort of load there is no difference. It idled fine my way. And it's getting converted over to sequential anyway.

Zaphod

Hi Ben,

one more question - I have seen, that you have in your spreadsheet that the trigger should be set to rising edge - in Abe's thread with the MSQ for his adapterboard he recently wrote, that it should be falling edge. What is correct. I have my "old" MS2 with Abe's circuit also set to falling edge.

Have you got your sequential set-up?

Greets

Ben

iTrader: (33)

It runs on rising edge. Have not tried falling edge. Falling edge is probably better.

Still have not picked up my sequential card. I have not had time.

AbeFM

iTrader: (3)

It's easy to switch it, aside from the timing jumpng by a few degrees. The issue is at low speed, one edge is unreliable (as when starting) and the motor will really misbehave.

Your inputs are inverting, on the crank?

Ben

iTrader: (33)

I had trouble starting but it didn't seem to be a synch error. It got a lot better after messing the the VE table and CSE.

I'm 1' of heatshrink tubing away from finishing the harness and the car running again. I also want to go ahead and pick up the sequential card. One day this week I'll have time time, hopefully...

Ben

iTrader: (33)

Finished my "boomslang" harness for parallel installation and all needed changes to the DIYPNP.

Tomorrow I hope to pick up a sequential injection card from DIY Autotune.

Also am installing a 10,000 uF cap between +12V and gnd on the adapter board. Figure it can't hurt. Need to run 2 wires from under the dash to the coils to get individual coils working.

Hopefully then it's done.

Matt Cramer

'99-'00 application docs and base map are up:

DIYPNP MegaSquirt by DIYAutoTune.com

10AEHooked

I'm not smart enough for this thread

Nice work though!

Ben

iTrader: (33)

DIYPNP is complete. I got the sequential board today and installed it. I loaded the new firmware to the DIYPNP and finally changed the settings to enable sequential. Unfortunately I had some sort of a software issue, and it took me several hours to get sequential set up in software correctly. Luckily, James on the MS2Extra site gave me a push in the direction I needed to fix the issue, and now we're good to go.

The seq card uses outputs PT6 and PT7 for injector drivers 3 and 4. I was previously using these outputs for VICS and cooling fan. Because the DIYPNP is going back in parallel with the stock computer, I let the stock unit take over cooling fan activation. The stock computer will also run the idle valve--though I hear DIY Autotune has closed loop idle working very well on the 99 now. MS will be controlling VICS still, but I moved it over to the IAC output. IAC is wired now wired to 4Q (if memory serves). IAC can drive a relay directly, so no need to run it through one of the internal relays. I removed the flyback diode from R5.

So tomorrow I'll reinstall it.

Zaphod

Great work... I keep the fingers crossed. Mine is also ready for going standalone... Gonna try today or tomorrow.

Ben

iTrader: (33)

DIYPNP is now installed in parallel with stock computer. Car is running.

No CELs from idling and rev'ing in the garage. If/when the rain stops, I will button the install up and try a test drive.

AbeFM

iTrader: (3)

So, are there enough outputs that you could do the full sequential AND maintain the rest? I've been more interested in this, I thought you could grab signals off the MSII card directly in a v3pcb, so perhaps you could get it on what you've got...

Ben

iTrader: (33)

With sequential injection and ignition, I believe there were only 2 spare outputs. I am using them for VICS and EBC.

I actually do NOT have sequential running right now, due to lack of a 2nd trigger. I'll address that soon enough. I had to give up using the factory sensors as sharing them threw codes.

Idle valve and cooling fans are run off the stock computer.

There is one more output on the card, but code is not written for it yet, so right now it's meaningless.

FoundSoul

iTrader: (5)

Ben's parallel install is pretty cool, and I'm sure some of you will decide to go that route, but for what it's worth the DIYPNP we just tuned in a 99 here was controlling idle very well. We idled it slightly higher than stock, but that's fairly normal with a standalone and helps it compensate for sudden load easier, which it was doing a good job of. I had that sucker pulling me up small hills with no throttle input, just slowly release the clutch in first and watch the idle valve open to compensate and away you go (slowly).

Ben

iTrader: (33)

I believe Dave is very happy with how his car turned out.

Thank you so much Jerry. And thanks to Matt and Russ for their tech support/hand holding.

Marc D

iTrader: (7)

Did you end up using the stock sensor, or going with that VR sensor and 60-2 wheel?

Abe, im sure you know we can do full sequential on our cars too. Again, Jean has a board that can give us full sequential. It cost in the range of 50 bucks though.

AbeFM

iTrader: (3)

It's not the cost I'm worried about... It's I/O...

Sequential ignition? Ah. There's something I would just plain lose. If you are running something like toyota COP's (cheaper than two sets of plug wires), you could save two I/O channels, AND have better performance. I would totally go this route.

You can't share the factory inputs? You should be able to if your input circuit is decent. The pull up is already there in the factory unit, so leave that out and share it. The current into the caps is minimal (leave it out if you're worried) and the current into anything active like an op-amp is so close to zero you couldn't measure it if you tried. If you want to do it as an experiment... Let me know.