Speeduino is now on Github Sponsors (Rather than Patreon): https://github.com/sponsors/noisymime
Help with building your Speeduino, installing it, getting it to run etc.
User avatar
By Eric H
#45252
So I've been trying to the the race car up and running properly, we had a dyno session cut short because of a bad sync loss issue.
If re-done the power and grounds, separated the power and signal line runs, then fixed the real main issue which was a dangling ground wire in the center console occasionally shorting out. :oops:
Now we get an occasional sync loss between 4K-6K RPM, and I'm wondering if there's any advice.
The setup:
UAC4 board, with a DSC.
VW ABA engine, full sequential setup with wasted spark (NEW VW unit) the stock VR sensor and shielded wiring. 60-2 crank wheel with a single pulse on the distributor hall sensor, also the stock wiring. The distributor is only feeding the ECU the single pulse signal, there's no spark through it.
The DSC Extra 0.1uf filter cap across the second input to filter the input signal more.
Both outputs on the DSC look extremely clean on my oscilloscope, no glitches or noise. However the output signal level is only 3.3V.

One thing I've noticed is the distributor signal on the composite logger is often weird. Some frames it's missing, some it's fine, and sometimes it's a very narrow pulse. Is this typical? The main signal looks great and the tooth logger is very clean.

The only thing I can think of is there is so much vibration that the DSC or arduino is losing contact somehow because the RPM and load isn't consistent.
Anyone have any ideas?

Thanks.
Attachments
(1.33 MiB) Downloaded 24 times
User avatar
By Eric H
#45262
The tune has both triggers on falling.
Sometimes it shows fine under the composite logger, other time I get an impulse, not a good pulse., and sometimes the single cam sync signal is missing entirely. This is what I'm asking if it is common. During all of this the signal outputs of the DSC looks great on the oscilloscope.
In the original post the tar.gz file there's both a regular data log and a composite log if you want to look in more depth.
Attachments
CompWeird.png
Tooth weird
CompWeird.png (42.83 KiB) Viewed 746 times
CompGood.png
Good tooth log
CompGood.png (42.74 KiB) Viewed 746 times
User avatar
By Eric H
#45270
Digging through the missing tooth decoder, the only way a sync loss happens is at line 397
 
if( (toothCurrentCount < triggerActualTeeth) && (currentStatus.hasSync == true) ) 
{ 
        //This occurs when we're at tooth #1, but haven't seen all the other teeth. This indicates a signal issue so we flag lost sync so this will attempt to resync on the next revolution.
        currentStatus.hasSync = false;
        BIT_CLEAR(currentStatus.status3, BIT_STATUS3_HALFSYNC); //No sync at all, so also clear HalfSync bit.
        currentStatus.syncLossCounter++;
}
Which is when the decoder has missed a tooth. I'll have to carefully recheck the wiring,
That this only happens under load and never when just revving the engine is concerning.
User avatar
By Eric H
#45281
OK, Here are some tooth logs, idle, 2K, 4K, 6K.
Looks clean to me.
Attachments
tooth6k.png
tooth6k.png (120.32 KiB) Viewed 651 times
tooth4k.png
tooth4k.png (116.62 KiB) Viewed 651 times
tooth2k.png
tooth2k.png (113 KiB) Viewed 651 times
IdleTooth.png
IdleTooth.png (109.46 KiB) Viewed 651 times
User avatar
By PSIG
#45289
Random: Misfires, preignition, bad timing, etc, can cause loss of sync when Speeduino is confused by the signal positions. Check your logs for oddities or the effects of them, such as a spike in your RPMdot, indicating when and where you got one. You can expect a spike at sync loss, but if you find other (usually smaller) ones, they may point to an issue that only causes sync loss at a certain magnitude.

Verify your jumpers are set for both signals "VR" so they both go through the DSC. This prevents the 2nd channel floating and possibly interfering if the Hall signal is bypassed (direct).

The added cap can slope the signal voltage, causing lower signal amplitude. If you don't need it, don't use it.

Some VR/wheel combinations are crazy voltage. If the VR voltage gets too high, again it can cause losses. Add a 4.7K resistor to the input and check losses.

No, it's not common.
User avatar
By Eric H
#45294
PSIG wrote:
Wed Sep 09, 2020 11:30 am
Random: Misfires, preignition, bad timing, etc, can cause loss of sync when Speeduino is confused by the signal positions. Check your logs for oddities or the effects of them, such as a spike in your RPMdot, indicating when and where you got one. You can expect a spike at sync loss, but if you find other (usually smaller) ones, they may point to an issue that only causes sync loss at a certain magnitude.

Verify your jumpers are set for both signals "VR" so they both go through the DSC. This prevents the 2nd channel floating and possibly interfering if the Hall signal is bypassed (direct).

The added cap can slope the signal voltage, causing lower signal amplitude. If you don't need it, don't use it.

Some VR/wheel combinations are crazy voltage. If the VR voltage gets too high, again it can cause losses. Add a 4.7K resistor to the input and check losses.

No, it's not common.
The DSC has the VR & Hall going through it. The cap is on the hall distributor input, not the vr input. The VR input is stock.
What voltage would be considered "too high" from the VR sensor?
The logs look good until a single cycle RPM drop from the sync loss. The timing is very conservative, good enough to drive around before the dyno.

I did more of a deep dive into the weird distributor pulse in the log viewer.
It turns out that MLV has bugs that if the signal level is zero at at the start or end, the display doesn't work correctly. I'll report them when I have time. I added a rough composite log viewer to my log viewer from another project and the distributor signal looks fine.

I'm also working on installing my 0.4 into my 97 m[…]

Water pump controller via boost port

Simplest option would be a separate controller tha[…]

ATmega2560 vs ATMega1280

Unlikely... It has half the flash space as the 25[…]

Question about sensor groud

There is only 1 ground in Speeduino designs genera[…]

Still can't find what you're looking for?