Even More Data 3.46

[grippo]

If you have a similar setup (sensor and wheel) on both cars then I would try the 10K resistor.

[TheMonkey]

I put a 10k resistor inline. The tach drops went away, and it revs right through that point now.

BUT... I seem to have intermittent hard starts. the cranks sensor VR voltage during cranking is +2v / -2v peak to peak. Would the resistor have dropped this? I think it was 3v peak to peak before the resistor went in.

What is minimum required VR voltage at the sensor?

[grippo]

A +/- 2V VR signal should be plenty. When you have the hard starting, is it because the trigger +/- is counting and the rpm jumps to 0 ? If this is the case, have you tried raising the cranking and after start pulse tolerance ?

[TheMonkey]

Al-

I'm pretty sure it was IAC settings. Needed a bit more throttle during cranking. I'm always surprised how subtle changes can make dramatic differences.

Things looking real good.

[TheMonkey]

Actually, I think that what's going on here, is erratic IAC behavior from the cranking voltage.

I was kind of driving myself crazy continually adjusting the idle stop screw after warmup when IAC should have been all the way closed. Then, I found that the IAC was not in a consistent behavior. I understand the settings very well, but just not behaving right. So, my hard starts were from adjusting the stop screw too low when the IAC still was letting air bypass, and then not enough during cranking.

Remember, the beta unit I have has the original power thingy in it that will be replaced with the upgraded one in the production version to fix the cranking resets. Right now, oil is warm enough that my unit does not show a reset during cranking. But... this IAC problem is even happening when there isn't a full power reset during cranking. I hope the new power unit in the production version of the Sequencer fixes this.

EDITED TO ADD:
I just pulled the IAC to confirm step range. it is exactly 205 steps to fully extended (closed bypass for full warmup). So, I set starting retraction to 250, and a linear idle step curve to max at 205 when fully warmed up. The problem is... that when it is fully warmed up, the idle is sometimes still too high. If I put my thumb over bypass hole, idle drops to where it should be. After it's running and idling too high, if I change the warmed up steps to be 350, it extends more, closes the gap and idles okay.

The point is this: the IAC is not fully extending when the computer thinks that it has. This happens both in 'always on' mode, and in 'only when moving' mode.

Do you think this is from cranking voltage even though there is not a reset on the data log?

[TheMonkey]

I unplugged IAC completely and shut it off in TS.

The intermittent hard starts continued. Sometimes perfect, sometimes a real problem with backfires. Sometimes would start, but won't stay idle. Other times fire right up.

So, I suspected the bad ignition signals again. Seemed unlikely, but hooked up the scope, and ugh....

Are you kidding... the phase shifts are back. These intermittent hard starts are from funky ignition signals again.

See the following picture. The red is the raw VR signal (after 10k resistor). The blue is the logic signal sent to coil #1. The very first blue ignition signal is accurate timing and duration of 5 ms. The one immediately following should not be there and has shifted to the next cylinder phase and dwell duration of 14ms. The 3rd signal is the proper dwell duration, but still remains in the wrong cylinder count.

This is the same exact phase shifting that happened in earlier code on the BMW that was kluge fixed by trying different delay/offset.

When I captured this trace, it was a huge backfire. Cranking advance is set to 20 degrees.

[TheMonkey]

I'll continue testing when new hardware run is available shortly. Right now, probably troubleshooting 1st run hardware unecessarily.

I'll be ready when it arrives.

[Peter Florance]

I put a 10k resistor inline. The tach drops went away, and it revs right through that point now.

BUT... I seem to have intermittent hard starts. the cranks sensor VR voltage during cranking is +2v / -2v peak to peak. Would the resistor have dropped this? I think it was 3v peak to peak before the resistor went in.

What is minimum required VR voltage at the sensor?

We should talk. I've modified some V357 boards to be more like V3 boards for this problem. I think it's a better fix than the 10k external. V3.00 VR circuit is rock solid with BMW crank sensors.

Or you could put NC relay contacts across the 10k and trigger it with starter signal (shorts out resistor during cranking).

email me as this forums email notification doesn't work for me and I'm too flighty to remember to check back.

[Peter Florance]

As I'm testing on a very similar setup as Scott, I thought I update here would be good. I've driven the car a lot since November. A few autocross event and even a short dyno session. I learned one thing: at least with my setup, I can't really run low z injectors in all types of driving. Street driving is fine, as is a 50 second autocross run. I tried to make my dyno pulls in 4 gear and the drivers overheated and went into thermal shutdown. 3rd gear pulls were short enough that they worked.

I'm adding 3.3 ohm resistors in series with injectors (these worked for me on an MS3X installation; same hardware drivers) and hoping they don't make the car hard to start (opening time doubles at low voltage AND added resistors).

But the car runs great 99% of the time. Hoping for 100% when I add the resistors.