Just bought a 12-channel oscilloscope
 

And you know what I just HAD to do, right?

https://www.miataturbo.net/attachmen...ine=1329772892

i recognize the bottom two traces.

the ones above that appear to be evenly spaced like batch ignition.

the top looks like more "crank" triggers.

edit: you did it in the future!

Top plots are inj events on cyls 1 through 4. Stock ECU.

What model is it?

Looks like, top to bottom...
inj 1-4, spark a, spark b, ckp, cmp

Pretty cool to be able to see so much at one time.

and he's still got 4 more channels to play with.

I don't get the injection cycle. why aren't the squirts evenly spaced?

Crikey!

We have a pair of Agilent mixed-signal scopes at work (4A, 16D) and those are about US$8,000 each. That's a hell of a nice piece of gear you have there.


And I'm wondering the same thing as Y8s. I've never scoped the INJ channels on an NB before, but I find it rather interesting that they are pulsing each injector twice per cycle at what I observe to be slightly above idle (looks like ~1200 RPM.)

Also, just out of curiosity, was that taken from an actual running engine? Those injector pulses look abnormally clean, and the duty cycle seems way too high for idle (6.85% based on 11 pixels/squirt x2 squirts across a 321 pixel cycle.)

Sweet!

Holy frack

Insert 'whycan'tIholdallofthesechannels?.jpg'

Have a look at this video, it is on my custom stim:

InfiniiVision DSO-X 2004A (4-channel, 70MHz) with the DSOX2MSO upgrade for the added 8 channels. $3500 after taxes.

See my previous post. 2002 stock ECU on my custom stim. The lack of flyback on the inj channels should be obvious.

As said, that's what the 01+ ECU does. I've been thinking about why.

Let's suppose that Mazda "tacked on" the VVT algorithm to a non-VVT ECU in such a way that the injection algorithm was blind to intake valve timing, so they were unable to properly time injection events to actual intake valve opening. So perhaps they break up the fuel pulse into 2 half-pulses, 1 timed to where the earliest the valve could be open, and the second timed to the latest the valve could be open.

This is the only thing I can think of.

maybe it's an emissions strategy...?

Emissions might be the most likley reason.

I can't see how. Two separate squirts means twice the dead-time. Seems like it'd partially negate the purpose of going from batch injection to full sequential.

Reverant, I'd be very curious to see how the injector pulses change as you increase both RPM and MAF to get the ECU into a higher-load condition. Specifically, at what point do the two pulses merge into one?

I can't tell you why but it must be common for modern EFI systems to use multiple squirts per cycle. Here is the ECU timing diagram for an SRT-4 and you can clearly see that it squirts 3 times per cylinder every 720 degrees.

https://lh5.googleusercontent.com/-p...0/ngc_4cly.jpg

Oh no, I have a bad case of scope envy now....my little one channel MAC seems too small! Did you get a pressure transducer too?? BTW, Pico's website has some great diagnostic examples for digital scopes.

3 of the 1 Megasample memory depth, 4-channel 250 MHz scopes our lab was selling off, got sold at $250 each...

Well, I really wanted more than 4 channels, a large display (8.5" is awesome) the ability to control the scope from the PC using Linux.