Trouble starting / Running like crap
 

I finished the COP wiring harness and stuck em in the car. Had to try forever getting it to start, I think that might be because the IAT settings are STILL way skewed... Its SNOWING right now and it shows 54 degrees IAT. Car struggled like hell to start, then started but ran like complete and total ass.. AFR showed 8:1 on wideband. Let the car sit with lc1 powered on and megasquirt still continued to show 8:1, closed megatune and fired up the LC1 logworks, it showed 19:1. Went into Megaconfig and changed the Lambda settings to Innovate PLX 0-5 volt 10-20:1 AFR, saved, went back into megatune and it shows identical readings to the LC1 log viewer.

Car won't start. Still shows its really rich on cranking, catches and acts like its trying to start.. but nothing. :confused:

*sigh*

Nothing?? Nobody??

You are taking into account that the wideband takes ~30 seconds to warm up right?

needs more fixing of the AIT resitance values....AIT is a big factor on fueling...

Sensors sensor sensors. Fix those first, then move onto other things.

Wanna get the car to start, pull the AIT and let it default to -40. I bet you it will start damn near instantly.

Shit.

I unplugged the MAF and IAT went to neg 40. Cranked and it didn't even catch, so theres no ignition. Pulled the 1 and 2 coils out, stuck a screwdriver in them close to a ground and cranked. No spark. Have +12 at all four coils inputs, cont to ground is good. I unplug the harness from the car and bring them inside because its 18 degrees outside right now (I'd give my left testicle for a GARAGE instead of the hole in my house thats called a garage..) and what do I find:

http://www.jugrnot.com/cops.jpg

I'm confused on what to do now, I modified all the pins like in Scotts writeup, but they won't stay secured in the connectors. :(

push it in hard with a precision screwdriver and proceed to epoxy, super glue, jb weld, etc,etc.

i didnt mod my pins (y8s came up with the novel idea), and simply siliconed the ends of them for peace of mind

then you didn't do exactly like his writeup said. or you inserted them upside down. they should and do grab. push them in till you here a click. they shouldn't pull back out

Pressed the pins in hard with the coils plugged in. Checked to make sure they were wired correctly for the trigger, they are. Have no resistance between the car side plug of my harness and each corresponding pin on the coils connectors. No spark. Came inside with the MS and hooked it up to Stim, shows injectors are firing (knew that much) and checked for frequency on both outputs for ignition. Both seem to check out okay. My test point for ignition output was at the LEDs, removed power from the unit and checked for cont between D19 and D17 between pins X11/X12 of DB37-25 and 27 both check out.

What should the resistance between the +12 and ground be on each coil? Double checked all of my ignition settings in megatune, all seem correct. No inverted spark, LED17 is Spark A LED19 is Spark B, trigger angle is still 65 which is how it ran with stock coils, 0 addition, time based cranking, 10 advance angle, hold ignition 1, fixed angle -10, 0 trim angle, dwell set for Dwell control, 4.0ms cranking and 2.6ms running, minimum discharge .5ms.. WTF is going on? :(

I test using my harness approx 27k ohms between the +12 and Ground on each coil, plugged in one at a time across the board. Also test between trigger and ground of each same way of 348 ohms.

And I just found cranking dwell should be 3.5 cranking and 2.5 running, I had 4 cranking and 2.6 running.. did I kill the ignitors again? :(

Nah, they should be ok considering some people are running like 4.6 ms running and 5.6 cranking dwell when using them on stock ecu.

Checked another measurement.. Resistance between the output of the coil and +12, and not getting any kind of reading at all.. I thought there was supposed to be some high reading between the two and that was how you tested it??

They get hot when they are being abused. No mistaking it. :-)

Might've found my problem. Scotts post on wiring shows this:

http://www.boostedmiata.com/FAQ/cops/cop_pins.jpg

I just found the 2004 Corolla wiring diagram for engine control which contradicts his wiring?

http://www.jugrnot.com/engine%20control-2.jpg

Scotts indicated 1 is ground and 4 is power, toyotas manual says 1 is power and 4 is ground... Would having those reversed fry the coils? If I switched them, think it'd fire?

:confused:

Theres no winning with this shit. Looks like I'm out 40$ for coils and will have to spend another god knows how much to get another set.

So am I reading toyotas diagram wrong, or is it just incorrectly labeled?? I'm getting really frustrated here and can't find anymore information other than that

Judging from your wires, it looked right to me, red is power and black is ground, right?

Check this:
http://abefm.smugmug.com/photos/242055166-XL.jpg

Looking at the left most coil:
white/black skinny wire is ground (can go right to the head)
green/red is signal (are you "inverted"? I use "normal", going low)*
Yellow is Tach Out. You won't hook this up
black/red is, counterintuitively, +12 switched

That's left to right, looking towards the coil from the connector end.

When I had the power backwards, it got hot and didn't fire (I couldn't believe the ground was the thinner line). When the polarity is wrong, it also gets hot. Touching the coils is the best way to tell if it's hooked up wrong.

This is totally fucked and I'm seriously pissed off right now.

They're all wired up just like Scotts diagram, red power, black ground, brown is the trigger. Spark is NOT set for inverted. Why the fuck was it running before and won't even fire ANY OF THE COILS NOW?!

Maybe because, for some reason, they're all burnt out.

can you revert to your stock coils to determine if the coils are burnt out for real or for some reason a setting is preventing spark?

FWIW, i didnt make any of the diagrams, i just copied and pasted into a thread without 15 pages of chit-chat.

It doesn't make sense. All settings are the way it was running good with stock coils, except the dwell.

Nope. The reason I went to COPs was because I loaded an old MSQ file I saved when I was doing the VB921s with inverted spark set so I blew up the factory units. That msq was deleted the day it happened, and I check for inverted spark everytime I start the car to make 100% sure its not set.

You're running these god damn things, do you have them wired up the same as your diagram and my wiring???

yep. and wired newbsauce's that way too. no issues. :confused:

Yeah, doesn't make sense. I checked the voltages at the wiring and everything in my wiring and it all checks out like its supposed to.

Is it seriously possible the coils were just bad to begin with? I'm really finding that hard to believe because it did run.. really badly, but it actually ran for five minutes or so!

what PN are your cops?

All four are 90080-19015 and have hand written by salvage yard I assume "2001 Toyota Corolla" written on them. Got them off ebay.

Looked up like 3 sets on carparts.com that I'm gonna order when I hear back from them, but I'm seriously concerned that something I did damaged them. Just can't figure out what though.

Try throwing them on a corolla and see if they work.

If I had a corolla, I would. :(

So I just found ANOTHER wiring diagram from toyota:

http://www.jugrnot.com/Wiring%20Manual-All_Page_10.jpg

What the fucK!?!?!?!??!?!!??!?! Why are these showing we're all wired up wrong???????

i wired brad's with the stock harness and plugs. ill have him check the wiring colors to it's relative location based on our numbering.

but if it was running, i dont see how it was wrong..




edit:


http://www.boostedmiata.com/cops/cops005.jpg


#1 is W/B #4 is B/W


so their wiring is reverse number from ours. so it's kosher.

Not really any help, but I spent almost a full day looking at wiring diagrams - this is what I learned:
Toyotas are all different from each other. The color wires mean nothing, in one car, white/black is positive, in another it's ground.

The pins on the coils do seem to be consistent, though. If you swap power and ground, it should get hot in a hurry. If it doesn't. there's definitely something wrong. :-) Check across the pins at the coil? I think you already did this. Check for a spark signal, too.

So my wiring is up to par, MS is setup correctly.. They had to be bad from the start?

It's pretty weird that all four would be bad at once.

Oh, do this (I did):

Hook known good power and ground to the power and ground pins.

Then, tap 5V (or is it ground? I'm pretty sure it's 5V) to the fire pin. If it shocks you, you drop the coil and wet yourself, then you know you're good. While I didn't do the wetting myself bit, I was able to hand fire them. I switched the MS to "interverted" so I had a steady supply of +5V, and I used left over pins from the boomslang operation to do the connections.
-Abe.

I have a pretty high dollar lab precision power supply that'll do 0 - 30 volts but only puts out 2.5 amps. I hooked up one of the coils with my timing light pickup around the wiring harness. Try to give it five volts into the trigger and nothing happens. Gave it the same 12 volt source as the power source for the coil and the coil clicks. gave it a bit more voltage and the timing light fired off once but I didn't get shocked. The coil continues to click like a relay when I give it voltage to the trigger?

Good news, Bad news.

Picture is worth 1000 words:

http://www.jugrnot.com/IMG_0233.jpg

I keep forgetting that I have dozens of old computer power supplies around here that have common grounding, +5 volt 30 amp outputs, +12volt 15 amp, -12 volt 30 amp, etc. The bench power supply I have has isolated grounds and I forgot.

Megasquirt *appears* to check out but it could still be the problem. Its somewhere between megasquirt and the factory coil harness plug on the car, so I've got the coils narrowed down.

Heres a good question. I have a relay board and cable, so I could hook this up on the bench. What are the chances I could use a spare CAS I have laying around, the relay board, and some wires to fool megasquirt into thinking the engine is running to trigger a spark event?

Okay. HELP!!! :confused:

All four coils will fire on the bench with me triggering them manually. BOTH Spark outputs have been verified functional with Stimulator running via. Oscilloscope having an offset 5 volt on/off.

Wired up A coil to the trigger output I made a jumper for from the stimulator to test with scope. Hooked just ONE coil up, test fired manually, powered up MS with the trigger hooked to ONE spark output. Nothing! It WILL NOT fire the coil! Any of them I tested, but when I do it manually giving it 5 volts.. Fires!

The cars wiring is good, the coils are good, Megasquirt seems good.. What is going on??? :confused::confused::confused::confused:

Bueller???

two things:
1) How are you test firing? You want to apply 5V for, oh, a few milliseconds, so, as fast as you can, while giving it +12 volts. If you're sure you're doing it under the same circumstances it should work if the MS does it or if you do. Again, I hooked up the same 12 volts from the car, set the output of the MS to "going low", but really, just take any 5V from the MS and put it on that line, and tap it on and off to the trigger pin on the coil.

2) I dunno about the holy MS-I, but in MS-II/extra you can test outputs and get the coils to fire away as much as you want, at any speed, etc, you can imagine.

Anyway, I'd stop with the pc power supply (though it's good to know the coils are good) and just do it.

You're not going to see anything with a timing light, don't bother. It can be done (I have) but the timing is wrong, for one thing, and two, it's hard. If you really want to, take the negetive power to the coil, jumper in a LONG bit of wire, several feet, and wrap that around the pick up a bunch of times, as many as you can and still get the clamp shut. That should fire it, but you don't need to since you can just dangle a spark plug out. If you want to set timing, let the coil hang out, jam a aligator clip in it, through the timing light pick up, and then on to a spark plug, which you ground. That will work fine, I set my timing that way and my car runs AT LEAST one day in three. :-)

Anyway, your two next steps are making sure the MS can put out the right thing and making sure the coils work from your harness on your car with your power and the MS's 5V does the trick.

I guess I just have to assume maybe the ignitors inside the coils are maybe bad? I know for damn sure I can't press that little button in time measured by milliseconds. Guess I'll source another set of coils and just keep my fingers crossed they end up working, at this point I really don't know what to do.

The bench test with an oscilloscope shows the output trigger is 5 volts. The coils need five volts to fire, everything should be on the up and up.... But its not.

Have you checked continuity between the coil firing pin and the ms pin?
You would just need an ohm meter, some wire (to extend to the coil from the ms) .

The coils are pretty robust, i doubt they would just die.

don't toss your coils. You said they worked on the PC power supply, so you know the coils are good. The trouble is in your wiring on the car or MS settings, I'm sure of it - somehow.

Get bare wires, ideally, pull the ones out of the toyota plug. Hook each one into the coil. Verify on the back of the pins (they will stick out a bit) that you have ground, 0V, and +12. Pull the trigger wire off. Then go in the car, switch the MS to "inverted". go back out. You should have ground, +5V, +12V.

Don't sit around too long with the trigger line hooked up.

Now take the 5V wire, and just tap it on the pin on and off, as fast as you can, even just one time. It should throw a spark - if the voltages are all correct.

If you don't have 5V, let me know. If you do and it doens't spark, and you checked +12 and ground are right, in the same spot as with the computer case, let me know.

My guess is if you do it like that, it'll work. If not, it might be a software thing.

Assuming it does work, there's lots of playing to be done in the output test mode.

I'm sure we can get this working.

My way checks both that, and everything else, too. Same idea, but you don't need long wires. :-)

I was thinking more along the lines if he had the right wire from the stock harness wired in via the patch harness into the ms. Isn't he parallel?

I don't care what he is, really. If he does what I said, and just measures what's at the coils, then it's got to be wrong, or they'd be firing. He already showed the coils are good.

Doubting a test you did will only have you walking in circles. :-)

I mean, heres what I did.

Took the PC Power supply. Ground, +12 volts, +5 volts. Hooked a momentary contact push button switch up between the trigger on the coil and +5 volts. Used the wiring that I installed on the car even to test this with, so I know for a fact the wiring should be 100%. When I press that button I get spark, Every time without a delay. Immediately unhooked the wire from the switch and hooked up to Megasquirt, running off the SAME +12/GND from the power supply.. Touched trigger output from MS to the trigger wire I just tested. No spark.

Any chance you can do me a favor? I realize that voltage and frequency are different things.. I don't have an electrical engineering degree (Yet, anyway) but I'm just curious about something.

Hook MS Up to the stimulator and stim for wheel. Crank the RPM to max (I read 36000rpm) and measure with a DVM the voltage output of each spark channel. Now, While I understand that the Oscilloscope claims its a 5 volt trigger the MOST my meter shows is 2 volts, and pulse frequency is different than solid voltage... but I just want to satisfy my curiosity... I'm going to head back to my brothers office tomorrow and use his boss' $70,000.00 microprocessor controlled calibration scope to 100% check the voltage output instead of the 400$ one I've been playing with. I'm starting to maybe lean twards MS's spark outputs aren't on the up and up. As for software, I went back through and checked everything spark setting wise 4 or 5 times with the software mods I got from DIY and they all match. Maybe the next step is reflash the firmware and check it again? Long shot, but I'm willing to do anything right now to get this running.

You get a spark when you press the button, or when you release the button?

Generally, if you can't do it with a $8 voltmeter, a $25,000 won't do it either. Except you might break the meter and have to sell your car to get it repaired. Oh, it's 70. All the beter reason not to mess with it. :-)

Odds are very good that you're looking with a digital voltmeter at a time varying signal. Basically, your voltmeter is seeing a pulse-width-modulation control. The easy way to think about it is it've averaging .3 seconds of 0V with .2 seconds of 5V. DMM's are almost always horibbaly slow. Don't worry about it.

Please try what I said. The MS puts out 5V, if it doesn't, that's what you need to fix. If you don't understand what I'm saying tell me, I'll draw you a picture. This was VERY easy for me to do in my garage in like ten minutes and it answered all my questions. You don't need to throw money at the problem, you need to learn how it works then it will be easy. You have all the tools you need (though a very small screwdriver would help.

Worst case, you can take your signals off the OEM Mazda plug and run wires over. It's really very easy, and then you'll know what to blame. Unless your goal is to drive around with a power inverter plugged into the cigarette jack, and a PC power supply in your footwell. :-)

LASTLY, if your power supply and MS weren't sharing grounds, then the 5V won't matter, and SHOULDN'T trigger it. This is a free EE lesson: Voltages are relative. Don't dink around, hook it up to the MS and let's make it work.



It doesn't matter. The coils are good. An equally useful question: Is it a normally open or normally closed button? :-)

Okay update again. I was about to type out another long post then just said F-it and wanted to try another power source.

Upgraded to a 50amp 12 volt power supply. Hooked both MS and single coil up, started power and touched the coils trigger wire to the output lead I soldered onto the stimulator. Spark spark spark spark spark spark spark! BOTH WORK!!!!!! That computer power supply just didn't have big enough balls for MS to kick off the coil.

Still won't start on the car. Finished making this really ghetto bracket, bolted the coils in, tried to start. Nothing. Have +12 volts, cont to GND on ALL FOUR coils, then checked the trigger with my multimeter while cranking. The trigger to all four coils has signal, checked both voltage and frequency with my DVM.

HOWEVER.... When I'm cranking (and its around 8 degrees ambient temp right now...) the +12 volt on the coils drops to EIGHT FARKING VOLTS!!! I'm 100% positive they won't fire at 8 volts! WTF!? I slap on my battery charger with 100 amp start booster going, still won't maintain higher than 9 volts during cranking. Kind of looks like I need to run a dedicated fused and relay switched power source from the battery to the coil packs??

Well, this is the last time I'll bother with this. Try hooking it up like I said.

My guess is they will fire just fine at 8 V. My car starts when I'm at 9V while cranking.

I've told you a test you can run that actually MEANS something.
<edit>
You've already almost got it, just touching the wire on at 5V will tell you all you need to know. But please just use the MS for all of it. You won't have to run wires or anything weird. I did add a stronger ground when I was having issues, but the only thing that ever helped was tapping 5V like I said.

I basically did EXACTLY what you wanted me to for the test! Thats using Megasquirt to trigger the coils, with the spark output A and B FIRING THE COIL! I Had NOTHING to do with the coil firing, that was the spark output FROM megasquirt making the coil fire on the bench! If that test means nothing then I give up and you can have my entire car and everything I own!

Here... I drew a picture of what I did.

http://www.jugrnot.com/IMG_0286.jpg

Megasquirt was using the 2222 driver from the LED to trigger the coil, Not me with some mythical benchtop +5 volt power source I pulled from my ass. I also tested ALL five volt sources inside mgasquirt and they're all 5.02 volts..

are you changing your ignition settings in MS back from Stim mode when you have it in the car?

(It is a stupid question, but it could happen)....

Stim Mode? Uh, oh, I think I'm out of my depth here. You wacky MS-I people.

Anyway, not trying to be a total ass... But you should do it all on the car, just to be consistant. It doesn't take all that much power to fire the coils. And since doing it with your fingers will mean long dwell times, they will charge even if you're a volt or two short.

http://abefm.smugmug.com/photos/247046110-L.jpg

It's very similar to what you did, but just do everything on the car. I'd unplug all coils but one (so you don't fry them) then "invert" the spark output so you have 5V all the time. Don't run the car, and just tap it. If it works, you KNOW it's a software thing.

That's the most important test. You know the coils are good, so it's either wiring or software.

^^^ what he said would definitely prove if all is well (makes sense now) :)

Okay, I think I understand what you're saying..

Yeah, MS1 doens't have output test mode.. kinda wish it did meow.

Its supposed to be a lot warmer tomorrow, like 39 degrees so I'll see what I can do. Came up with another test that I'll try after yours. If your test checks out on the car, I've got a spare CAS laying around.. I'll plug in teh CAS and turn ignition on, turn the CAS and the coils should spark....... I hope... This seems like a bad ground or bad battery, but my batterys 100% charged now.

I'll figure it out, I'm fuggin determined. Thanks for the help everyones given so far!

I saw someone do that, a drill motor got the cas turning 250 rpm, my car cranks right there, so I guess it's a good speed. The MS-II works down to about 50 rpm. And don't forget to switch the coils back after my little test. I'd hate to see more coils burn cause of bad wiring. But don't be too worried about it, they can get pretty hot before they fail. Just touch them once a minute or so to make sure they aren't getting hot. Good luck!

I seriously give up.


Manually turning CAS gives me fuel pump, fuel pressure, injectors firing, IAC valve going... NOT FUCKING SPARK!!!!!!!!!!!!! Thats with 12 fucking volts at the + input of the coils!

And its colder than balls, spending enough time to unhook my engines CAS, hook up another one, remove one coil from the spark plug and turn the ignition on my fingers are numb and I can't stay outside any longer. God I hate the midwest, this is bullshit. It was supposed to be 39 degrees today. More like 29. ------ weather bastards.

Two things:
1 - did you try the test I said? Have you actually seen it spark with the car providing all voltages and signals (except you tapping the signal line like I said, but the voltage is provided by the MS on the MS pin with the MS circuits, etc)

2) Use a drill! You'll wear out your fingers like that. :-\

It sucks it's giving you so much trouble. But you need to know it's firing with the MS and your wiring 100% for sure first.

When you know that, it's a software issue, which you can fix in your nice warm home. Good luck.


I don't know how this applies, but I had a very similar issue, with a bad burn on my MS-II. It didn't make sense, but the plugs wouldn't fire. This is exactly the sort of thing the hardware test I've suggested 5 or more times would confirm.

Okay, I did your test first thing.

On the car I Unhooked all but one coil. Removed from the harness the trigger for 1/4. Inverted spark in megatune, checked for 4 volts at the coils connector, plugged in one coil. Key on, engine off.. pushed the trigger wire back into the harness and it sparked. Will note also, along with the spark the fuel pump triggered, injectors primed, and megatunes connection reset.. Dunno wtf that was about.

Right now I've pulled it back out of the car to reflash the firmware. God I hope this fixes it, because I've no freaking clue after that.

Yeah I'm lost. Reflashed firmware, still won't fire the coils during cranking or turning the CAS manually. Watching RPM on megatune while turning CAS with my fingers and its spinning the engine at 300 or so RPM, which is pretty close to cranking.

Adjusted cranking dwell for ignition, made no difference. Checked again with one coil hooked up inverting the spark, sparks everytime.

Should I mention that it sparks when I touch the trigger to the pin, and not when I remove it??

That seems weird, I think it should spark when you remove it - but I was so nervous doing mine that I can't remember when it sparked.

It does sound like the hardware is CAPABLE of running it. I know my car was acting weird and adding more grounds helped (though in the end I got rid of most of them). Try this: Unplug your wideband if you have one - that's less load on the line to the computer. And add a ground right to the head on teh ground line back from the coils. I doubt it'll do a lot.

As a last desperate measure, try setting the "cranking speed" down to like 50, see if it behaves.

It looks like either a power/ground issue, or a software issue. If it's software, I won't be much help since I don't know MS-I very well. But the good news is your wiring and your coils are fine.

I've got a weird update.

1/4 are wired together, 2/3 wired together. Waste fire.

If I have 1 and 2 hooked up, turning the CAS externally by hand, they'll both fire. If I hook up 3 and 4 none of them fire. It looks like maybe the TRIGGER signal isn't strong enough to light off these coils?

I have a pair of VB921 FETs I can install.. Wonder if that'd help?

You shouldn't have to. I did change the drivers for those LED's to 2n2222's instead of the much smaller parts that are in them. I guess you could check the current on that line, but as I understand it, the coils use field effect transistors which don't use any current at all.

Do 3 and 4 work alone?

At least you are making progress.