DIYEFI.org Forum

Gday all

I'm the owner of a RB25DET R33 GTST and this will be my thread for doing out details of signals and whatnot, I'm in the process of designing a TA-based 6 cyl board to control it eventually, but it'll be a fairly slow process, so we'll see how we do.

Starting with what I captured on the scope tonight attached.

CAS is a Nissan/Mitsu 360 degree unit, signals as the ECU reads them are 0-5v logic. IIRC this sensor is an open collector type, so needs a pullup on these lines. (Related FreeEMS decoder thread: viewtopic.php?f=56&t=71 )

AAC/Idle valve is a 12v in, when you're on the throttle, it runs it at 50% duty, and decreases to about 20-25% as it gets to idle, the capture is with engine warm at idle.

The S2 R33 has ignitor in COP, from the trace, it looks like it feeds it in 5v (it's on a low current pin on the ECU), which starts the coil off at a low resistance, than as it charges, the resistance gets higher. At idle, dwell is about 2ms.

Injector is a high-Z, and needs flyback protection of some kind, stock ECU saw 85v on the scope, definately need some clamping.

I forgot to add some more details, had the stock ECU open up for some circuit tracing, ignition is fed by a weird old style module that gets 12v, so is likely a level shifted output.

Fire order follows pinout on ECU connector. (1 5 3 6 2 4)

You always say "level shifted output" but it's totally unclear to me exactly what you mean by that. Please explain in detail?

Level shifted means you go between differing voltage levels, usually not inverting, for example you have different parts that are say 3.3v and 5v, level shifting uses intermediate components to change voltage levels between these parts where those parts wouldn't normally be able to talk to each other. While these days a lot of 3.3v parts will tolerate 5v on their inputs, not all do, and same in reverse, not all 5v parts will recognize a 3.3v signal as a high. Sometimes this is as simple as a voltage divider, to something more complex, such as a FET driver (though some creative thinking sometimes gives you solutions like http://ics.nxp.com/support/documents/in ... n97055.pdf ).

In this case, I'm referring to the ECUs chip outputting a lower logic level signal, and being boosted to provide 12v output to the ignitor.

Yep, fully aware of what level shifting *usually* means ;-) Was more interested in what you meant by it: AKA, high-side drive. Did you prove this by disconnecting one and watching its unloaded voltage?

Yeah, pulled the injector and coil plug to cyl 1, unloaded it sends 12v out the trigger. I bought an extra probe today for the external trigger on my scope so I could record both CAS POS and REF on a stable trigger point. Dumped it all to CSV, and made sure I set it to maximum captured data rather than displayed, got one file per cylinder, with plenty of samples per transition so I can map out exactly how many POS per REF there is Going to put it all together and upload it as one big CSV (with a OLV screenshot too )

Hell Fire wrote:Yeah, pulled the injector and coil plug to cyl 1, unloaded it sends 12v out the trigger. I bought an extra probe today for the external trigger on my scope so I could record both CAS POS and REF on a stable trigger point. Dumped it all to CSV, and made sure I set it to maximum captured data rather than displayed, got one file per cylinder, with plenty of samples per transition so I can map out exactly how many POS per REF there is :) Going to put it all together and upload it as one big CSV (with a OLV screenshot too ;) )

AWESOME! Thank you very much, especially for the confirming 12V to ignitors thing, which I've strongly suspected despite everyone telling me otherwise for a long long time. Legend.

And here it is, each record is about 10µs apart. This file is merged from 6 different triggers and captures (though one capture managed to get about 2 and a half cylinders worth, but I went through each uniquely anyway so I could add the trigger points in).

The CYL-IGN is which cylinders ignition line I triggered off, trigger happened at the start of each of these and I made each 2ms long (as that's about how long the ignition dwells for), so it's close enough for a rough guess of when the coil is fired (though I should have triggered off falling edge rather than rising).

Hell Fire - you might be interested in this viewtopic.php?f=62&t=2257