EMI Noise issues with MSI causing ECU resets

I have been having some issues with resets on my MegaSquirt board, it's a MSI V3.57 MSnSE.

I figure I'd tell you guys what I'm going through right now as Mr Taco is considering going with MS for his 2JZ build and would start out with MSI.

From my thread on HBZ:

flatblack280 wrote:

I'm running a MSI V3.57 PCB using a 83 ZXT Dizzy for spark on a turbo F54/P79 L28E, Intercooled, L28ET 260cc injectors.

I didn't bring my laptop to work today so I can't upload the MSQs or datalogs yet.

Problem #1:

On surface roads, the car will run fine, the Fuel maps still need a bit of work but it will rev fine up to 4500. [I didn't want to go above that because my wideband is acting up, see Problem #2].

My problem is on the interstate, of if I am going over 60 MPH, so in 4th or 5th, When I try to give it gas it will start to sputter and when my wideband was still reading it would go lean. Occasionally if I give it enough throttle, it would sputter so hard MS would reset. On the datalog you can see the MAP go from ~ 3400 RPM and a little over 100 KPA immediately drop down to ~30 KPA when it would reset.

I'm fairly sure it's not my fuel maps, it'll run rich at the same RPM and MAP at lower speeds, it's only when I'm at the higher speeds.

My injectors only had a 60% duty cycle when this was happening, and I've got 36 psi fuel pressure at the fuel rail gauge at idle.

I tried jiggling the wire harness around when it was idling to see if anything was loose but nothing happened.

I'm confused. I'll upload the Datalog and MSQ when I get home tonight.

trwebb26 wrote:

what plug wires do you have? (no, I'm not crazy).

http://forums.hybridz.org/showthread.php?t=120033 may help...

letitsnow wrote:

Post up the datalog, or look specifically at the batt v graph in MLV, I'd bet it's really rough, kind like this one. This one specifically was cured by doing away with PWM and installing dropping resistors.

flatblack280 wrote:

I'm using the N/A Spark plug wires...

I'll be getting some Turbo ones this week now. Thanks for the suggestion.

I'll have to read that thread more thoroughly later in the week when I'm not swamped with school work but I think my timing is right with the Dizzy/MS

Hmm that is very probable. What did you do exactly with the dropping resistors?

Here's a screen shot at the time of the resets.

letitsnow wrote:

You're definitely getting some resets because of voltage spikes, I can't say if they're all because of them though.

With dropping resistors, you basically just put some high wattage resistors inline with the injectors and turn PWM off. What is the resistance of your injectors? You probably need 1 for each bank, around 25w and 3.5ohms, it depends on the resistance of your injectors.

Where are you getting power for the MS/relay board? If it's on the alternator, I'd move it to the battery.

flatblack280 wrote:

Alright. Which dropping resistors did you use?

I've got switched 12V and straight 12V running directly to the battery.

I've got the turbo injectors for now. I'll be upgrading to 35# Ford SVO injectors soon.

letitsnow wrote:

The graph I posted was a TBI 305 in a firebird, it only has 2 very low resistance injectors which changes the required resistance.

If you have 3 4 ohm injectors per bank use a 25w 3.5-4ohm resistor per bank.

I like this one if you mount it to a heat sink(like the case of the MS or something). It's pretty big, I doubt you could fit it in the case, so you'd have to put it on the outside or mount it to something else.

Or this one if you don't want to mount it to a heatsink, but it's about twice as long.

Either will work for turbo injectors, I'm not sure about the ford injectors, if they have a resistance lower than ~2 ohms each, use the latter one and mount it to a heatsink.

Even if it doesn't completely cure it, it's about the cheapest thing to try at under $10. Just be warned, mouser will send you multiple catalogs that are at least 4" thick.

flatblack280 wrote:

Cool! Thanks for the info. I guess I should track down the source of the nasty voltage first but it can't hurt to order those, they are cheap.

trwebb26 wrote:

Please do yourself a favor and buy magnecor wires... I had SO many problems with resets until BRAAP got me straightened out by getting some new wires. Magnecor wires are a 100% must for a Z megasquirt install IMO.

letitsnow wrote:

Some of that nasty voltage is because of the PWM, dropping resistors would cure some of it up. Notice how it gets worse when you get on the throttle, it's because the injectors are at a higher DC and the pwm circuit has to deal with more current/emf.

flatblack280 wrote:

I'll get in touch with Paul tomorrow about some for my S30, my N/A wires are crappy and old and I'm sure are causing some problems.

Where are yours mounted? I'll order two of the 25s tomorrow. I've got a busy week ahead...

letitsnow wrote:

I don't have any in my s30(yet), my uncle mounted them to something under the dash in his firebird(it was his car that had the issue, not mine), really any cool piece of metal that has a flat spot big enough to mount them will work, bare metal with some heatsink compound is preferred, but painted would be alright. The MS case is pretty ideal, but I think they'd make it look slightly too 'mad scientist' for me, especially if it's visible, like mine is.

The only time these have a chance to heat up is under extended high duty cycles, like a track day or top speed run.

flatblack280 wrote:

Alright, I just ordered the 50W 4.0 OHM dropping resistors. I'm seeing if BRAAP can get me some magnecor wires and see where that gets me.

Flexicoker wrote:

sorry, I only skimmed through most of this post, but I was having similar problems. Magnecore wires didn't cure it, redoing all the grounds didn't help...

I turned PWM off and drove around for a bit and it seemed like it helped with the voltage spikes, so I installed dropping resistors... problem completely solved. Unfortunately i don't remember what resistors I used at the moment, but that is my first recommendation before you spend $100 on plug wires. You can turn off PWM and drive around for a bit, as long as you keep the duty cycle low, and should not have a problem with overheating the injectors.

flatblack280 wrote:

Sweet, thanks so much - I got the dropping resistors Let it snow recommended, I'm going to try out the PWM off thing tonight. The car runs fine otherwise.

So first I'm going to play with the PWM settings [Pulse Width Modulation - it's the injector driver circuit] and after that try Magnecor wires. If that does not work I will be looking into my dizzy settings.

Chris and I were also talking about seeing if my alternator needs a condensor or not.

This is also a good hack to the V3.XX boards to make it more robust:

http://www.diyefi.org/forum/viewtopic.php?f=6&t=285&start=10

I will have updates soon!

awesome, that sounds promising. hopefully this cures it, because you'll need to pass 65mph if you're going run a 2j supra down

mr taco wrote:

awesome, that sounds promising. hopefully this cures it, because you'll need to pass 65mph if you're going run a 2j supra down

You mean 85

Alright so I tried the "PWM off" thing, which as soon as I hit the throttle it blew out my injector drivers because I had no protection against the nasty voltage my alternator was putting out.

My car is stranded under the old Shell Gas Station hang over, but I'm fine with that until it stops raining.

The injector drivers are Q1 and Q5 - those redish boxes on the heat sink at the top of this picture:

http://www.megamanual.com/ms2/v3components.gif

I'm buying a set today from DIYAutoTune and hopefully that is my issue. That will be easy because those parts are through-hole like the V3 board. Also Russ at DIYAT said "I have replaced surface mount parts before on personal projects.. it can be done, just tricky and small... make sure you don't have coffee / rockstar / monster jitters when you try to do it!!! LOL"

You know where my shit-ass soldering iron is..

Whats the plan for the alternator situation?

mr taco wrote:

You know where my shit-ass soldering iron is..

Whats the plan for the alternator situation?

Well, you helped me install it yesterday :P

I was at the shop yesterday, picked up some random parts, got my 240Z bumper so I could fab up some mounts for it [my early 260Z mounts are hitting the FMIC] and while I was there I grabbed a condensor off of my L28ET alternator.

Hopefully that will tone down the batt V.

Whooo!

My FET injector drivers came in today, I'm going to go pick them up after work today, put them on the board, and see if the board will fire up on the old map.

If it's fixed, I'll put the PWM circuit mod in as soon as it stops raining and we get some sun.

FETs are installed, I'll try and get a run in tomorrow.

Chris gave me the software for my wideband, hopefully I can get it working properly and my sensor isn't fubar'd.

You can see the change in my battery voltage from up there ^ to down here v - much nicer.



I got resets when I swapped in the Dropping Resistors, but it was really random.

I drove to school today and decided to take the interstate to see when the resets would happen. I went through the intersection for the long stretch to get onto 540 by my house and did a 25-120 mph pull. NO resets, just really really rich.

I'm still going to order some Magnecor wires but I feel like this issue is coming to an end! yaaay

amazingly I actually knew what I was looking at there. lol. Thats great to hear mat. I can't wait for you to get that thing tuned, then we can race and you can see how far you have left to go