This is where I got my info from supporting lower dwell times:
'99/'00 coilpack measurements / dwell - MX-5 Miata Forum

 

I tested both 2,500 and 5,000 today, couldn't tell any difference.

 

Several have scoped it. Mazda runs them at 5 ms. I understand Jason's findings. With the stock wiring and 5ms running dwell, the coils are not showing signs of going into current limiting on a scope.

However at 5 ms, after a certain RPM, the coil spark cycles will be on top of each other. Around 5k or so IIRC, I've watched it on a scope. If you spend a lot of time at higher rpm, then 4 or 4.5 *might* allow a bit longer coil life span. You'd possibly be giving up a minimal amount of energy at lower rpms. It might be a perfectly acceptable trade off.

You guys should be replacing your stock coils anyway.

 

Bumping thread. It's two years later and I want to see what dwell time everyone is running now. How is your cranking now with it and have you needed to change your coilpack.

I'm still running it at 2100, cranking takes about 3-4 cycles before starting and I haven't had to change my coil pack.

 

NB1 runs at 5, nb2 is at 2.1 I dont remember who it was but i remember asking one of the tuners why.. he said thats what the stock coils are and that the nba nbb are different.

 

Bull. NB2 is 5ms as well, I've scoped mine.

NAs where 8-16ms.

 

Would you really recommend that much dwell on stock NA coils?

 

I would definitely not recommend it, in fact you can't go higher than 8ms with the MS2.

 

Reviving this thread from the dead. I'm having spark issues on my 2001 that's all stock other than a wideband and GM IAT sensor. After a coil died, I replaced both with "Spectra" brand coils off rockauto. Now it starts and idles, but I have what I believe to be a high RPM, high load misfire. Spark plugs recently checked and gap confirmed to still be 0.044, even if they are a little worn.

Above, Reverant says 5ms for dwell which he scoped out on an NB2. I have an MSLabs MS3 and got a base tune direct from him. The base dwell settings are 8ms for starting, 3ms nominal.

So which is it? 2.1? 3? 5? What are VVT guys having luck with?

 

I use 3ms on my 400whp turbo build with the stock VVT coils. NGK BKR8E-IX plugs. Just because the stock ECU uses 5ms doesn't mean that you have to do it as well - stock coils die often on our cars and I believe the 5ms dwell time may have something to do with it.

 

Thanks for the fast feedback!

Not sure what changed in my car, but I'm pretty sure killing coils now. Moved the dwell in table mode first to 3.5, then 4, then 5 ms and the car ran better after each increase before they degraded further. Coils were <140*F, they shouldn't be burning up. They were the second cheapest on RockAuto though... That's for another thread I guess.