3P's TCCS Disassembly/Analysis

[3p141592654]

The ECU processor computes and controls the single IGt signal as I noted above. There are three Denso chips separate from the processor that take the NE and Gx CPS signals and compute the A/B control signals to select one of the three coils. This allows the engine to run (limp mode) even if the main processor goes offline. Note that this also means that from the processor point of view, there is no difference between and GE and a GTE ignition system.

The IGf is checked by the processor during the 30 degree window after an ignition event for confirmation of ignition. Its a binary thing, it either happened or it didn't. 6 no shows in a row and a code is set. IGf plays no role in timing or dwell.

 

[Kai]

Aaaaah, right - gotcha! Geoff & I have designed a new igniter for the 7M that can control 6 COP's with the original ECU controlling them (simple design, still wasted spark), but we were confused by the 'dwell control' present in the igniter. Right now we're trying to figure out what the coil current draw is, and/or the inductance of the coils...if you feel like popping into flashchat, it's help - Geoff is great with electronics, but not so much cars

 

[grimreaper]

Would it be possible to post an excel style rpm vs load Ign timing chart? the graphs you posted are great, just difficult to nail down specifics.

Thanks!

 

[mecevans]

+1 for android. I use an app on my phone with my RC Tricopter to look at data and change PIDs on the fly over Bluetooth. Would be a cool feature to add.

 

[Nick M]

Note that this also means that from the processor point of view, there is no difference between and GE and a GTE ignition system.

So your board could make the ignitor for the GTE run a GE system? I was thiking of IGda and db.

 

[3p141592654]

Nick, I meant that from the microprocessors view inside the ECU the GTE and GE are the same. The IGda IGdb signals are generated by 3 Denso ICs inside the ECU. The processor just computes the timing signal (IGt). If it wasn't that way, the ECU would not work if the processor malfunctioned, and Toyota always makes their ECUs function even with the processor completely disabled.

I'll post the timing map shortly. I'm sure the standalones could use it.

 

[Rollus]

Hello all,

I'm still reading this topic.

@Pi, is there any ETA (even in years) for your board release? Is there a beta-testing campain?

Kindly,

RolluS

 

[bfr1992t]

Jon, I figured you didn't post the timing map in excel intentionally to encourage others to find it themselves. Posting or sending it would be most appreciated.

Kai, what is your goal? The stock ignition system is good to near 1000 hp.

 

[3p141592654]

RolluS, the hardware design is frozen and we have a number of boards out in the MR2 community where they have been performing well. Right now the bottleneck is on the software side of things. The current interface is reasonably complete for datalogging, but needs work on the tuning side of the story. That work is progressing, though slower than I'd like, due to work commitments mainly.

 

[3p141592654]

Attached is an image of the timing map. Haven't figured out how to attach files to this post. I would not use this map without a very capable knock-sense algorithm monitoring things. The stock ECU knock sensor is fast and pulls timing a lot during normal operation.

 

[Rollus]

Thanks for your answear Pi

Good luck, I know that the hardest part is the one you cannot touch with your hands, but only with your mind

 

[bfr1992t]

Attached is an image of the timing map. Haven't figured out how to attach files to this post. I would not use this map without a very capable knock-sense algorithm monitoring things. The stock ECU knock sensor is fast and pulls timing a lot during normal operation.

Wow those numbers are aggressive!

 

[Nick M]

The stock ECU knock sensor is fast and pulls timing a lot during normal operation.

So says 850 technician handbook. Although they call it a closed loop system, they don't call it "fast" per se.

 

[3p141592654]

If knock is detected, timing will be retarded on the next cylinder event (120 degrees of crank rotation), and the amount is based on the learned knock value for the rpm/load band and whether the detected knock event was severe or minor. I would call that pretty fast.

 

[Rollus]

1/3 of turn while running between 1000 and 6500 rpm seems pretty fast yes..

Will it be fast (or smart) enough to prevent detonation (ie for abnormal lean mixture caused by injectors or egr failure)?

++

 

[3p141592654]

Probably not. First it has to react to detonation, so at least one cylinder will fire lean, then the total correction is limited to a maximum value defined in a table of rpm and load, but typically no more than 15 degrees. That would not be enough to save the engine I would guess.

 

[Rollus]

Thanks for explanations your knowledge is just awsome (and you've worked hard for that)!

Did you get the timing map of JDM ECU?

 

[Kai]

Jon, I figured you didn't post the timing map in excel intentionally to encourage others to find it themselves. Posting or sending it would be most appreciated.

Kai, what is your goal? The stock ignition system is good to near 1000 hp.

As i said, an igniter to replace the stock one that can supply either the stock coilpacks/is300 coils OR *SIX* 1MZ-FE/1JZ-GTE coils (still wasted spark) - we're going for CPLD control with MOSFETS for switching. Mmm. beefy.

 

[Nick M]

If knock is detected, timing will be retarded on the next cylinder event (120 degrees of crank rotation), and the amount is based on the learned knock value for the rpm/load band and whether the detected knock event was severe or minor. I would call that pretty fast.

You will get no argument from me.

 

[bfr1992t]

As i said, an igniter to replace the stock one that can supply either the stock coilpacks/is300 coils OR *SIX* 1MZ-FE/1JZ-GTE coils (still wasted spark) - we're going for CPLD control with MOSFETS for switching. Mmm. beefy.

I honestly wouldn't bother making a custom ignitor if you're not also adding sequential control. The stock ignitor can drive the 7m coils and often the is300 coils just fine.

Are you programming dwell control in to the CPLD? It wouldn't be hard to eliminate wasted spark if you tap in to the G1 or G2 sensors in the CPS. This would be beneficial for some EMS's too.

Which mosfets are you using? There are some really nice purpose built units out there today.

Jim