@Ben - DIYPNP hardware latency?
 

Hi Ben, (I adress this to you as you might have the best access to the data),

have you found that you need to set the (ignition) hardware latency with the DIYPNP and the 99/00 Miata - maybe at your or the DIYautotune-shop-car?

I will try to scobe it at my engine, when it is up and running again in April. Just wanted to check out some other data...

Thanks

We've noticed that MicroSquirt C27 (filter on main trigger opto) can cause some latency issues. After removing it (which you can do carefully with a soldering iron by heating up and lifting just one side of it into the air), our test car ran better.

I have already done this due to Jerry's advice...

Great. Then no, I've not noticed any latency issues beyond that.

O.K. nice.. thanks

How "severe" is the issue? Is it worth "recalling" the cars I've modded for the update?

Jim

If those cars have properly tuned spark maps, then I would not go back and remove C27, since you've tuned around the latency as your commanded advance will be greater than actual. If you then pull C27, your timing will advance more at higher rpm, which might cause det.

The newest version of the microsquirt module is coming without C27 installed.

Thanks, I ordered a DIYPNP and a Microsquirt module the other day, do those have the C27 installed?

Jim

Double check them to confirm, but they probably will not have C27. If they do, then you can heat up one end and lift it up to remove it form the circuit.

Is this applicable to MSII aswell?

I warned about delays in the trigger signal in some thread about discussion of filters.
FWIW A "2 pole" RC filter will filter out high frequency noise better than a single pole filter for a given delay.

This is why it's good to have 2 values for setting timing errors in the software:
Basically for "y-intercept" and one for slope.
You set all timing to 10*, then you adjust the first value to get the correct timing with a light at idle, and the other at some high RPM say 5000 RPM. There is slight interaction so you repeat.

You'd want to remove C30 if it's there.

Jason, MS2 does what you describe.

Important to note, there is a difference between hardware latency and software latency.