ha... ms problems... help me sort this out.

[740ATL]

SUNDAY

Out of the blue... I started the car (without changing anything) and gee... the idle seemed kinda high... almost like the idle valve wasn't working properly... that's strange.

... warmed up... idle stays around 1400 (supposed to be... normally 900rpm). The car has been running great for a couple months now.

Then... I go out driving.... punch it... rev-boost-rev- BAM!... huge backfire....not at redline... happened again and again and again, then it died... and wouldn't restart.

No blown fuses, nothing melted in the box, MS still fires up, just won't start.

Pushed it home, yanked the plugs... drenched. Blowtorched them, restarted... rough... but restarted. Idle still high.

reflash, same thing.

Let it run... e-fan comes on... damn thing resets the ECU when the fan shuts off.

TODAY.

I disconnected the idle valve, turned it off in MS, and now I don't get any resets at idle when the efan comes on. Haven't tried driving it yet... a bit timid on having to push it again... The idle valve Q5 was upgraded to the tip122 a while back.

I'm attaching a datalog of the whole boost-reset debacle. Note the crazy rpm spike (which I haven't been having at all) occurring just before a reset and the CLT sensor value plummeting at a reset. No real pattern here as neither of these things happened at any of the other boost-resets (7 all told).

I'm investigating adding 1n4001 diodes between each actuator/relay... the boost controller has one already, I just added one the efan and I am about to add one to the idle valve.

I've checked the grounds (all grounds ground to same point), resistance is 0, voltage (when running) is 0.001v between the db37 ground and the ground point.

I'd appreciate any advice you could throw my way... if only to help me sort out my thoughts.

regards,

Mike

MS v2.2 029v

PS>.. overboost was set to 255kpa

[MrBill]

What are your voltages during all of this? Sorry i am a simpleton and dont know what to do with the file:(
actually i could try but im lazy
either way, a click that temporarily draws so much amperage to shut off MS seems possible with low voltage issues, as you said you were having problems before, maybe theyre not fixed?

Edit: oh an excel doc...
Well im not sure man, why not try to straight wire everything, grounds all to one place, power wire to the battery, and see what happens.
then go through and disconnect one thing at a time until the problem goes away>

[benflynn]

check cam timing and actual piston pos

unless this is the 16v, then i guess you would know if you sheared the small gear


also can star disabling all you can, start from as basic as you can, start pulling clt, iat, tps conn and see if it stops

sounds like a bug though, only ms problem ihave ever had was the weird barro thing

[The Aspirator]

DEFINITELY put a cap on the idle valve, I remember how happy you were when you figured that out a few months ago for other reset problems. And it seems like the idle valve is what's causing your issues. I made sure to buy a good dozen of those caps, I'll use 'em everywhere!

The only RPM spike I see in that whole log is near the end when everything gets stuck on for like 30 seconds. And the coolant spikes seem to get worse and worse as the log goes on. Humm, stumped me!

[740ATL]

Yeah, definite weirdness in there.

I yanked the idle valve hoses, turned it off in MS, and was not able to reproduce a reset. As soon as I "turned it back on" the idle went wahoo. So either the valve is wonky, or something is amiss with the circuitry. hmm.

Looking back at the day it wouldn't restart, I think the reason lies in the fact that the car continued to dump fuel in the chamber even after it died...

I did a few things yesterday, like add in a 2nd alternator ground, beef up the daughercard ground (22ga cable to 12ga cable), and run a dedicated line from the battery to the MS box main supply. Went from 13.2v to 14v...

No e-fan resets (haven't been able to produce one).... dunno about boost because I've been a little afraid to drive it for fear of having to push it again. :)

we'll get there.

Mike

[JohnMc]

I found I had to isolate the MS box from the main e-fan relay with an intermediate little cube relay. MS grounds to turn the fan on, and the main relay was hot all the time. And this was apparently feeding power back into MS some odd way when the key was off. Very weird behavior until I added that little cube relay to cut the connection between MS and the fan relay when the key was off.

[The Aspirator]

Looking back at the day it wouldn't restart, I think the reason lies in the fact that the car continued to dump fuel in the chamber even after it died... Had that happen to me once, I accidentally loaded an .msq file from a car that wasn't using spark into my car while it was running. Spark stopped, tons of fuel, car died and literally flooded the cylinders/intake/intercooler. Fun. Change your oil if it was really flooded.

JohnMC, that's really interesting, I'll have to keep it in mind! Thanks for the tip.

[740ATL]

Had that happen to me once, I accidentally loaded an .msq file from a car that wasn't using spark into my car while it was running. Spark stopped, tons of fuel, car died and literally flooded the cylinders/intake/intercooler. Fun. Change your oil if it was really flooded.

JohnMC, that's really interesting, I'll have to keep it in mind! Thanks for the tip.

Good call on the oil John... I'm going to dump it and put some new stuff in. Oil and filters are cheap. This is just one of those things with MS I guess. I don't have nearly the same depth of experience as you or Kenny or others... so I guess this is just a baby step. I still love it even if it beats me up occasionally. :-P

I found I had to isolate the MS box from the main e-fan relay with an intermediate little cube relay. MS grounds to turn the fan on, and the main relay was hot all the time. And this was apparently feeding power back into MS some odd way when the key was off. Very weird behavior until I added that little cube relay to cut the connection between MS and the fan relay when the key was off.

John, are you saying to have an additional relay between the existing e-fan relay and the e-fan circuit in the MS box? I'm interested.

Mike

[JohnMc]

Yes. I forget the exact oddities that were occuring, but I think my issues were mostly related to shutting it off, then starting it again. I initially ran the wire from the MS box (wired up an output on the duaghter card) straight to the low side of the two-speed 940 fan relay. I noticed that some LED's were remaining lit when the key was off, and MS was supposedly powered down. I think it wouldn't start until it was physcially unplugged.

My hunch was that the fan relay is always hot, even with the key off. I wired it up like that so the mechanical temp switch on the radiator could cycle it on once or tiwce if it was really hot when shut down to prevent a possible post-shut down boil over. Worked fine when MS was running, but when shut off I think current was flowing back through the MS box in some odd way. Since I wanted to leave the fan relay on all the time, I needed to isolate the MS box from it when the key was off, so I just added a small relay in the line. When the key is on, the relay connects the MS output to the fan relay, when the key is off the connection is broken.

Doesn't sound *exactly* like what was happening with yours, but the e-fan part is the same. I'd just unhook your e-fan wire (temporarily) and see if your symptoms clear up. I could see some nasty voltage spikes coming off an e'fan when it shuts off if the circuuit isn't isolated proplery. I know AC compressor clutches give off *really* nasty spikes, most cars have diodes on the clutch wire now to help suppress it.

[sethfc]

mike,
i know you know me - and how i cant help you =p BUT i say run w/o the idle valve. make life more simple ;)

[740ATL]

Thanks John. I appreciate it!

I've installed a 1n4001 diode on the efan between 12v const. and the ground to ms. So far so good. No resets.... all of this has been done with the idle valve disabled. I've gone around installing 1n4001 diodes to everything now... boost control, efan, I'll throw one on my tach output circuit, etc. I'm trying to get this car as close to daily driver reliable as possible.

Mike

mike,
i know you know me - and how i cant help you =p BUT i say run w/o the idle valve. make life more simple ;)

Seth... the idle valve makes a huge difference on cold starts.... for me it's really the difference between starting the car and driving right away vs. waiting 8 minutes for the thermostat to open and pissing off the neighbors. Even warmup mode only was a godsend over no idle valve.... Mind you, it still starts without the valve, but cold idle is rough at 600rpm with no valve.

With the valve, cold idle is 1200rpm, tapering down to 900rpm. nice nice.

Mike

[benflynn]

I'm trying to get this car as close to daily driver reliable as possible.



weird, my ms has been problem free for 2yrs now

[shaved240]

Hope you get the problems sorted out. Seeing all this weird, random problems makes me wonder how reliable ms is actually...

[740ATL]

Hope you get the problems sorted out. Seeing all this weird, random problems makes me wonder how reliable ms is actually...

MS itself is extremely simple. People using it for basic fuel and ignition control (single coil + dizzy) are using it every day in their daily driver.

All of msns-e is experimental... there are still features on http://megasquirt.sourceforge.net/extra/index.html that have the experimental or test label.

The real issue is with what people do with it... specifically the issue is with their wiring... I've noticed over time lots of dirty connections that were sapping voltage. I've noticed poorly soldered connections, inappropriate shielding, etc. All were my mistakes.

The problem for me is that I'm not an electrical engineer... if something goes amiss... I'm often at a loss for how to put 2 and 2 together, whereas someone with more experience would see a problem immediately.

Mike

[dl242gt]

Another thing that can be important is to not always ground things to the same ground point. For example the idle motor shouldn't be grounded to the same ground as a ground for the MS box. You can get noise feeding through the ground plane and this is reduced/eliminated with separated grounds.

If you remember the stock kjet. There are separated grounds on the intake manifold because to have them all together will cause noise and it will affect the other circuit.

[The Aspirator]

Really Dave? Interesting, never thought of that. In a few weeks I'm about to re-wire my whole MS and I was planning on putting every ground onto a common ground block. Humm, maybe I won't know, although it'd be sooooo handy. How about MS and all relays onto one ground block, then all moving parts like the idle motor, boost control solenoid, etc onto a separate ground block in the engine?

Then again, ground is ground, isn't it? Yes some spots give better signal strength than others, but if you're getting feedback from somewhere wouldn't it find it's way through the whole car? Just thinking out my butt here.

[740ATL]

I dunno what I did, but I'm not getting the reset anymore. I had to turn off the idle valve completely in the software as it simply would not regulate idle anymore. I'm not sure if that means the idle valve is bad (boost killed it?) or the idle valve circuitry is bad (on the board). This is even after adding the 1N4001 diode between the two wires of the idle valve (12v and ms ground).

Also, for whatever reason, the low speed on the 960 efan will not turn on... even when I ground it to the body. It doesn't inspire me with confidence, because this happened at teh same time as the idle valve wonkiness but the high speed still works.

sighs.

[The Aspirator]

Well when I re-wire my car I'll try out the idle valve and e-fan control, as well as all the other goodies, so we'll see if I'm getting those problems too. Hope not!

[JohnMc]

I had a long struggle with my idle control until I realized the Bosch valve was actually shut at something like 35 - 30% DC. It was also wonky because I didn't have a real TPS to turn idle control on and off, so it was active all the time.

[dl242gt]

Yes, John. That would be the way to do it. You can put the electronic device grounds together. I'd put anything that is electric like an idle motor on it's own ground away from the grounds for the electronics. The ground for the fuel pump is another ground to not have near electronics.

The diodes are nice for taking out any reverse voltage spikes that may occur. You are putting in reverse biased standard diodes so they absorb spikes, correct? It would also be a good idea to add a cap across some of the circuits that may be noisy. They absorb voltage fluctuations and smooth the DC voltage. Although I would think that the Megasquirt designers would already have that in the circuitry.

On the Crane HI6 ignition that I have to get around to installing they advise an electrolytic cap be installed on the heavy red wire if the wire is lengthened from the length that is factory supplied with the Unit. An MSD might need the same thing to keep down noise.

[tequila_gundam_no_chaser]

I cannot comment on MSNS-e, as I've only been running it a scant month now, but MSNS worked quite well for me (MS with distro). It had no fast iidle control, but I successfully ran the car for 2.5 years, without issue. Never had any problems, ran great despite how bad the harness looked. I still have the harness today.