MicroSquirt.com read only forum (Jan.15,2018)

On my last data log I ran the engine for approx. 15 min's and experienced 2 resets in that time period.

Al, has Bruce looked at the jpg's I sent of the +5 Vcc noise? Some feedback would be nice.

Thanks

Marc

Marc,

Bruce hasn't had time to do anything - he was knee deep in making presentations for the Megameet in Atlanta all weekend and has probably been locked up at work all day. He will reply when he gets a chance. I see the following steps: You are pretty much maxed out as for what to try. Bruce may want Kris to take out the power lines from the harness, use caps on the coils, etc. If that doesn't help and two people have the same problem, then he will probably want to try it on his engine and track down the source of the noise. That won't happen this week - we are too swamped.

As far as the VR sensor problems, if they are coming from the missing tooth gap, usually happens above a certain rpm like 3 or 4000, but could happen above a lower rpm, I can send a code fix for the ECU only that will compensate for that. Or it may be coming from the Vref noise and go away when that is fixed.

Al,

OK, I'll carry on testing Lance's code and getting the new engine ready to fire. Thanks for the feedback.

Marc

I'll also look at how I can re-run the +12V wires and check to make sure the supply is clean but the way it's wired up now is just so convenient and clean. Looking at LS1 wiring looms from places like SpeedSceneWiring and EFIConnection, they include the supply wires too but instead of having 1 loom up the centre of the engine and branching off, they branch it off at the rear of the engine to supply each bank.

If I can get some time with my friend I might see if we can look at some ways to include more filtering.

ca434sbc4 wrote: Al, has Bruce looked at the jpg's I sent of the +5 Vcc noise? Some feedback would be nice.

Marc,

I am totally slammed right now... but a quick look at the noise, the question I have - is there noise always there? I would start with just power hooked up, nothing else, is there noise? Then slowly add loads and see if the noise is still there, add them systemically in order to identify the culprit. Do the same with Vref loads.

A few things come to mind - noise back from +12V or noise feeding back on Vref into the box. The testing Jim did where he had noise attributed to the MAP sensor would lead back to the Vref (its used there and the TPS)...

Keep on testing as much as possible..

- Bruce

Also VREF is tied to the bootloader wire(s) for those that hooked them there - in the original doco's it said to do that for the beta units.

krisr wrote:Also VREF is tied to the bootloader wire(s) for those that hooked them there - in the original doco's it said to do that for the beta units.

Yep, for the beta units this is correct. Both processor lack the series resistance that keep noise from damaging the processor, its the only line not protected with resistance. We learned this the hard way with early Microsquirts but the sequencer was already in PCB stage before we figured it out. Hooking them to Vref keeps them happy, and the ports are inputs and no real appreciable current flows. But if there are P/S unregulation situations I could see this adding to the resets.

Anyway, for these version just try to see if there is any sensitivity/correlation of Vref to resets...

- Bruce

Bruce,

I'm considering adding a 0805 1K pullup on the red & green bootloader inputs and yanking the wire out of the connector. I'll use the jtag +5 as a source. It will remove 1 potential noise input source.

Too bad the uSquirt won't accept a MAP,&TPS signal via can bus from the MS2 box. I could yank these inputs and still be able to run the engine to test for resets live.

Marc