i know i know. but ive been reading through that shit for days.

the only small thing you can try is setting the crank to RISING edge. JasonC SBB did some measurements and found that occasionally (cranking) the sensor gives bad trailing edge data. the rising edges are correct but the trailing edge are inconsistent. sometimes they come immediately before the next rising edge.

he also determined that it matters which edge you use for the cam on a VVT motor because at certain advance positions, the cam signal crosses the crank signal...

https://www.miataturbo.net/forum/t27028/#post321668

OK no difference, it pretty much ran the same. Attached is the zip of the log file.

I have a 53MB video I shot with movie mode on my digital camera. I uploaded it to Picasa since it is too large to post here or e-mail. The quality is not great, and I think the sound did not record for some freaking reason. Sorry, I grabbed the camera in haste and started shooting, not sure what happened.
Picasa Web Albums - John - Adaptronic
I show the dash (hard to see in the video), the AFR/VAC/TEMP gauges, and the WARI screen. The flashing red light is the low oil pressure warning light I installed. It flickers when the engine RPM is really low, like 300 rpm.

I used the a/c again to heat soak the motor. I turned off the a/c when it started running poorly. The radiator fans were still on though. After a minute or two of rough running, the coolant temp dropped, the rough idle cleared up and sounded better, and the engine revved more or less normally. So to me it is definitely something triggered by heat. It does not seem to be directly correlated to the coolant temperature, or even the MAT. Instead it may be something heat soaking in the engine bay that somehow messes with the Adaptronic.

It is too late for me to try the video again to get sound. But I'll try again tomorrow or this weekend.

One thing I noticed in your log is the battery voltage starts high and slowly drops down the hotter it gets. My battery voltage at idle is around 14.5v. Maybe your alternator is going south when heat soaked?

Could be. I have considered that. I wonder if Stein's logs are showing the same thing.

But I have had rough running at cruise and the corresponding logs show voltages above 14v.

Kind of perplexing. Maybe add some additional dead time correction for low voltage idle conditions?

It looks like inadequate injector voltage/engagement length--------->current to open properly. Thats what my gut just really tells me. If you try to idle your injectors at too lean of a condition they will act like that because pulsewidth gets too low and they become unstable. Sometimes they will open and bring you rich, then they will go way lean like out to 18-19 when they fail to open properly. I'm going to ask rob about this, and see if increasing the current could damage the injectors.

I've seen both. See my log screen shots posted on post 153 and 155.

On another note, I have asked the trim question at least four times and I have never gotten an answer.

When mine goes rough, the trim is always flatlined at 4. ALWAYS. When it idles, trim is working normally. I don't think that Travis has ever been able to get Rob to answer the question as to what may be causing that. I know that he has asked him. See posts #153 and 155 for reference.

I was wondering the same thing Travis. That is why I tried the correction table specifically for my injectors. I am still not sure it is right-on though, never got a response from Injector Rehab or Witchunter when I asked them for values, so all I have to go off of is the information I mentioned previously.

Even with the MS, with hi-res code, Chad and I both had problems getting our injectors to idle properly at anything above 13:1 or so. We both are using 550cc RX7 injectors. I messed with it for hours and finally gave up. So I tended to hand-tune the idle to be about 12.5:1 or so. That is with no loads. With an a/c or fan load, it was stable at higher AFRs like 14:1.

Because of this experience I have been tuning the Adaptronic cells by hand to idle at about 12:1, especially when there are no a/c loads. I let it get up to about 13:1-13.5:1 or so when there is an a/c load. I was trying to idle leaner with the peak-and-hold, but because of the rough running I gave up on that for now.

I would also like to know if running more current through the injectors would help. I was tempted to try this, but did not want to take a chance on damaging the coils. Plus, with the peak-and-hold scheme, I do not know how much current is pushed through during the peak mode, and for how long.

Hmm, it makes sense that we use the adaptronic to control everything that it can; but at the same time... I may look into using the factory ECU to control AC + Altenator as to free up some I/Os.

Travis, is there any drawback to letting the factory ECU control the alt and AC? Or will the lack of MAF cause issues? My thought in the matter is that if the factory ecu is still in there... let's let it do more work. :)

Anyhow, I will install the Adaptronic ECU tonight & this weekend; so I hope to contribute more than I have lately.

Cheers,
Prospero

He hasn't said anything about it, and I have asked... I'm not ignoring you my friend, I feel like I have to milk information. Maybe I'll call over there again. Ever talked to a real live Australian?? Hollywood doesn't make that accent up at all. :laugh: I suppose he probably thinks the same thing about Americans...

They are coming out with a newer pnp harness that gives you the option of choosing what the stock ECU does and what the Adaptronic does with jumpers to select. I don't have an upgrade path for you guys though, you would have to buy it outright. They did say that if the PNP harness was causing the problem they would trade them out though. They aren't going to leave us hanging on that.

My knowledge of how that circuit works is limited. I don't understand how you do better then applying 14.xx volts to the injectors? Maybe they have a regulator and apply 17-18 for .5ms and then drop it down to normal voltage, but then this ECU modulates in current, and intuitively I don't know how you just apply a certain amount of current to something. It makes sense because the work done by the magnetic field is done by current and not voltage, but the details are blah!

OK, snap... Travis what about this...

Looking at the advanced features in the Basic Setup tab, the Adaptronic is set up for full sequential. But is it not true that the Miata injector wiring has them wired as two sets of injectors in parallel, and not as four individual injectors? I think this is at least true for the NA, but cannot check the NB wiring. I do not have my wiring diagram in front of me.

If the injectors are also wired in groups of two for the NB, there is no way we can actually be running full sequential since we are using the stock wiring harnesses.

From what I saw at the Adaptronic site, there was mention of doubling the injector current when running two injectors in parallel. Now I do not know if the sequential firing offsets this somehow, but given the issue at hand. it makes me wonder if this is set up optimally.

No the NB should be wired for full sequential. I think they are wired full sequential all the way back to 94 when they popped in the 1.8L. I'll check the diagrams in just a second... Checked, there are 4 inputs. I also have the factory wiring diagram if you want it. Its too big to post though 8mb.

OK never mind then. No worries on the diagrams I have them at home, thanks.

You know, it occurred to me, that the most current that can ever run through the injector is just simply I=V/R since the injector coil is just constantly powered with system voltage, and ground-triggered by the ECU (Adaptronic). For the 550cc Denso, 195500-2020, the injector resistance is rated at 13.7 ohms. At 14.7 volts, that is about I=14.7V/13.7O= 1.07A.

So, basically, no matter what we set the injector current limit at on the Adaptronic, it will never be able to pass more than 1.07A. Even if the resistance is down a bit, say 12 ohms, then that would be 1.2A. Peak-and-hold cannot drive any more current through than this, because resistance is fixed, and the voltage supplied never exceeds 14.7 volts or so; It has no way of doing so, unless it has a way of creating a potential that is negative with respect to ground, which I doubt. Peak-and-hold is able to pass more current with low impedance injectors because they have, well, lower impedance.

Thus, I see no harm in setting the current limiter higher. at 0.9A, it is already close to the maximum possible current level anyway. Agreed?

Yea, thats what I was thinking, the only other thing it could be doing is bumping the voltage up for peak. I sent a message to Rob but he hasn't sent a message back...

I do not think that is even possible. The 12V signal to the injector is not provided by the Adaptronic, but comes straight from the stock harness, right?

No it comes from the Adaptronic. The grounds are to chassis. If you look at the ECU window you can see what each injector is getting as far as voltage. I don't think 1.5amps is going to burn them either way. Especially for just a test run... As heat goes up, resistance goes up and current goes down if we're already at the limit then that little extra heat soak might be enough.

Ahh OK I thought it was the other way around. Same deal either way. The Adaptronic would require a DC-to-DC converter or something like it in order to raise voltage above the supply level, at least for DC. I agree about the heat soak. I was thinking the same thing.

OK I bounced the injector current discussion off of a Engine R&D buddy of mine who knows a lot more about engines than I do. He agrees that there is no need to current limit high-impedance injectors. So I am going to try raising the current limit. No need to set it any lower than 1.9A (the max setting). The ECU Data window has a readout for injector current (below the voltage reading) so I can watch that to see how much current is flowing through the injector.

BTW, not that it matters for this discussion, but the injectors ARE fed +12V continuously by the vehicle harness, and the ECU is firing the injector by switching to ground. You can see this in the Adaptronic wiring guide. Also from the Installer's manual:
"..These [injector] outputs are current-controlled open collector outputs from the ECU. They pull low when the ECU activates an injector..."

The voltage readout on the ECU panel is showing the injector trigger wire voltage at the ECU. From the manual:
"The voltage sensed at the injector pin on the ECU. This would
normally read around 12V when the engine is not running. If this
reads 0V, that indicates lack of 12V supply to the injectors."

How are they doing anything with peak and hold then? The only way possible for that to work is for there to be some change in voltage... We need Joe Perez... Where is he lurking when you need him?