trying to fix Sync loss around 2000rpm

[NMBRPL8]

Gday, long time listener, first time caller. Have installed Microsquirt on my Kawasaki W800 and thrown a turbocharger in the mix for good measure. Have it starting nicely, idles and revs fine to around 2000rpm at which point it loses synch, cycles the fuel pump relay and settles back down again. If i hold the throttle there it is like hitting a hard rev limiter (which ive checked and its not of course) I have tried to eliminate as many variables as possible, running fixed timing, very little if any ASE and WUE, but it comes off of those anyway once it warms up. Have turned off as much as possible anything else that would affect it, from the forums and documentation I suspect its a noise issue upsetting the microsquirt and hope to track it down and eliminate said noise.

Now, i know log files and an .msq would be super handy, ill get those up when i can but I dont have any currently in front of me, i bought the unit about 3 years ago and only just fit it all up and tuned it, so one of the things i tried was updating to the latest firmware, to which i restored the tune but next im thinking I will start fresh on the updated firmware and set it all in again manually. I have used the search button, ive been reading and trying things, i have the bike at my workshop so it will be Monday before i can try more things, its currently 1:31am sunday morning, so im doing the usual and jumping the gun here to ask you for your random wild guesses over the internet.
Things on my to do list so far:
Wiring. I realised the uS is sharing a power feed with coils and injectors, that cant be great for noise. Seperate wires, fused seperately and whatnot but in the harness they share the same feed back from the ignition switch. Will fit a relay off the battery to the uS and trigger it from the key.
Earths are run together, crimped, soldered, tested and nice and solid to the frame of the bike, with wire going from the same point back to the battery directly and to the engine cases directly. Any issues with this setup earth wise?
VR pickup. This is the stock pickup on the stock trigger wheel, variable reluctor wires to VR1 as per wiring diagrams, shielded wire through the harness but it does have 10cm ish of stock unshielded wire - worth shielding? I have not tried flipping the polarity, adding capacitors or resistors inline, or playing with software based noise filtering. I welcome any and all pointers here. I did try it on rising edge and falling edge, seemed about the same, factory setup is rising edge so left it on that. The factory flywheel is a 24-2 setup, 100 deg offset timing has been verified with a strobe and responds accurately to timing set in software. Running it on fixed advance at this point.
Mucking about with capacitors on the VREF, and/or coil (single coil wasted spark, original coil driven with a mitsubishi J121 igniter) and/or injector power feeds - again just in case of noise issues?


I realise there isnt much we can do without hard data, datalogs and msq files and even being able to tell you exactly which hardware and firmware versions I am on. I will get those monday and post them up. I will likely spend a bit of time monday trying to figure out what blatant rookie error I have made and hopefully correct it. If from this limited info anyone can tell me where ive goofed, or what to check, please fire away! Thank in advance
Scott

[dontz125]

You MAY have a ground loop. You mentioned MS to frame, frame to engine, frame to battery. There may be a heavy engine to battery for the starter; if so, you can have weird things happen. The preferred grounding scheme is MS to engine, engine to battery, and leave the frame out of it.

[NMBRPL8]

Much appreciated, thank you i have added it to the to do list. I niavely thought since its such a short run compared to a car I could get away with using the frame as an earth. I did earth everything together on one bolt, so even though the frame is earthed everything also has wire to run through regardless of the frame. So as it stands its heavy gauge wire from battery to frame & heavy gauge wire from same point on frame to engine, MS two heavy earth wires crimped and soldered and bolted to same point on frame as heavy battery earth, fuel pump to same point on frame also. Sensor grounds to sensors only of course. I will remove the frame from the equation entirely - actually since I am redoing the entire harness I wont actually use the frame to earth anything across the board, its only a few extra wires to earth all lights etc in the harness anyway, no need to use the frame really besides convenience. No lights or charging system currently connected trying to get it running right before adding possible interference back in.

What is preferable ground wise, I saw it asked in other threads when i searched, but found no answers. Is it preferable to earth at the battery where possible, or is it better to earth at the engine itself? I would have assumed at the battery, but happy to earth to the engine if it helps.

[NMBRPL8]

OK just been out and got some new 125a copper wire will remake the earths and be sure to only go between engine and battery and nothing else. Stopped in at jaycar but they were more or less useless on the capacitor front, since its running just fine up to 2200 ish rpm hopefully its not a bad signal issue, but there is a more helpful store closer to work so will check that out after some rewiring and logging.

[NMBRPL8]

SOLVED

I think. Silly me, not reading all the pages correctly. I had forgotten abour the shunt resistor solder pads inside the case, so as a test i wired in an external 10k potentiometer across the VR pickup wires, fired it up and dialed in resistance until it revs cleanly with no resets. Magic, its fixed! will remove external pot and solder the jumper pads inside the box and hope it works even better.

My coworker was soured on the whole idea of megasquirt after he struggled to help a friend pin down a reliable crank trigger on a car setup years ago, on megasquirt 1. In the end he sold it and went with a haltech. My coworker has been stressing on my setup here, and over my shoulder telling me i should have bought a different setup, MS is too sensetive, too adjustable, too hard to get working. Well, while im a long long long lway from perfect, its wired up and running with fairly minimal fuss and in about the worst conditions i could pick for it, and im pretty darn happy and impressed with it