Cant rev past 6500

[hybrid]

So you may have the dreaded "spark blowout" that some turbo guys talk about.

[nukkinfuttz]

well, I would argue that point. I don't believe I had spark blowout... I was detecting erratic VR signals and losing sync (without the MS declaring a sync loss, look at some of the pics I posted)

Here is the way I think of the situation... in order to jump a gap, you need xx,xxx volts per inch of gap IN ONE STANDARD ATMOSPHERE. As a consequence of discharging an arc, there is a broadband electrical emission, and its power is directly related to the discharge voltage. So, as my voltage needed to jump the gap increased (due to cyl pressure going up from boost/load), then the ignition system was having to produce maybe double and triple the voltage that would be required with no load, then the Broadband emissions (RFI) would also increase. It looks as if there was a certain threshold I would pass and this noise was then being counted as VR pulses, which disturbed the engine sequence.

Ive had spark blowout before, but this does not feel like spark blowout. and the VR sensor trends going bonkers showed me it was something else.

closing the gap reduces the overall voltage required to spark, but I think this was more of fixing the symptom than the root cause.

[nukkinfuttz]

a spark blowout would have showed up in datalogs as a lean spot. The odd thing here was I was able to maintain a clean AFR through this. meaning, I was sparking fine... just at the wrong time.

that might also explain the 15 head gaskets Ive blown.

[jsmcortina]

a spark blowout would have showed up in datalogs as a lean spot. The odd thing here was I was able to maintain a clean AFR through this. meaning, I was sparking fine... just at the wrong time.

What evidence do you have that you were sparking at the wrong time?

James

[nukkinfuttz]

VR sensor trends show the Teeth counter go all crazy at the 6K rpm mark. They are attached both in .csv and screenshot to this thread.

my knock sence showing knock at 6k every single time, even with the sensitivity all the way up (Lowering the advance bin changed nothing)

audible popcorn sounds above 6k (Lowering the advance bin changed nothing)

able to hold WOT at 25PSI of boost on a flat road in 2nd gear and NOT be able to accelerate above 6k (Lowering the advance bin changed nothing)

A GTECH inertia trend showing HP drop off at 6K

Megatune "HP" trend showing HP fall off at 6K when at WOT at 25PSI

not sure what else I would call it if I can hold WOT at 25PSI and not be able to accel past 6K while maintaining AFR

15 or so blown head gaskets... on a built motor running 2.5 points of compression LOWER than the factory build, yet using a stock ECU spark map (so you would think you would require MORE advance with less compression... then dropped 1* per 2PSI of boost as the map increased above 100KPA) always while maintaining excellent AFR of 11.2:1

knocked the ring lands off a set of Ross pistons... while I had great AFR.

[nukkinfuttz]

Notice a RPM Clip at 12 PSI, good Afr. Now, how in the world could the RPMs NOT go up any higher on a flat road in 2nd gear on a 12 second car? Well, its not a lack of fuel(good afr), so it must be improper spark timing. I have posted previous screenshots and .csv files of the tooth logger showing crazy counts at this level, but there is what it actually looks like. Feels like this too.

Notice I have good AFR
Notice I have 0 sync loss counts
Notice my RPMs "jump" from like 3.5K to 5.5K instantly, but have 0 sync loss counts. (how could that be?)
Notice I have 0 as a "sync loss reason" code.

In other words, the MS is counting teeth erroneously (as noted by RPM instantaneous jump, and .csv tooth datalog), yet on the high level fails to flag a sync loss (even though one must have surely occurred with that RPM jump) nor did it set a "reason code".

[jsmcortina]

What does a timng light show?

James

[nukkinfuttz]

I cant view timing while driving... and the problem doesn't happen unless im at WOT and full load. catch 22.

[dontz125]

Unfortunately, some things (like WFO / 100% load issues) can only be sorted out on a dyno.

[Six_Shooter]

I have to wonder if this is due to how the MS reads the GM 7x wheel.

On OEM applications there is a 60* additional offset to the timing signal, as compared to a typical dizzy application using the same basic ECM and bin file. This is, according to some sources, to switch between reading the initial 3 notches that are coincide with the home notch being 10* after TDC #1/#4 (on V6 application, which I think is 70* ATDC #1 on a 4cyl, IIRC), to the 3 notches that would be spaced far away from the home notch. This according to some is to eliminate the home notch as a possible trigger and avoid false triggers at high(er) RPM. This additional offset may be built into the code when GM 7x is selected but I have seen any mention of this anywhere.

So I have to wonder if switching to a differently laid out trigger wheel would solve this issue.

I also think that maybe going to a DIS module from the OHV 2.0/2/2L (Up to about 1994) might work better. Those engines used the same GM 7x trigger wheel, but used a pair of dual tower coils, and not the integrated coil that the Q4 used. The MS would need to be set up for GM HEI 7/8 using the DIS settings, which were probably similar to what you had when you used the original Q4 DIS.

[nukkinfuttz]

I have to wonder if this is due to how the MS reads the GM 7x wheel.

On OEM applications there is a 60* additional offset to the timing signal, as compared to a typical dizzy application using the same basic ECM and bin file. This is, according to some sources, to switch between reading the initial 3 notches that are coincide with the home notch being 10* after TDC #1/#4 (on V6 application, which I think is 70* ATDC #1 on a 4cyl, IIRC), to the 3 notches that would be spaced far away from the home notch. This according to some is to eliminate the home notch as a possible trigger and avoid false triggers at high(er) RPM. This additional offset may be built into the code when GM 7x is selected but I have seen any mention of this anywhere..

Can you try that explanation again? I didn't follow, but I definitely want to understand your thoughts on this.

So I have to wonder if switching to a differently laid out trigger wheel would solve this issue..

I have previously switched to a 36-1 trigger wheel from DIY and mounted a Hall Effect pickup... this combo did not fix my problem and was hard to keep the bolts tight on my crank pulley, so after I found out MS2 would natively decode GM7x, I switched back to the internal relator wheel on the crank with stock VR.

I also think that maybe going to a DIS module from the OHV 2.0/2/2L (Up to about 1994) might work better. Those engines used the same GM 7x trigger wheel, but used a pair of dual tower coils, and not the integrated coil that the Q4 used. The MS would need to be set up for GM HEI 7/8 using the DIS settings, which were probably similar to what you had when you used the original Q4 DIS.

the quad 4 and the twin cam had the same coils. What was different was the ignition control module. On the older quad 4, it was "intellengent" and would produce its own spark timing of 10*. Once the ECU decided it wanted to play, it would put the Ignition into Bypass Mode, then also begin sending it a new commanded Ignition Advance signal which it would follow.
The later twin cam ignition module abandoned the "intelligent" part, and it because just a giant transistor switch to drive the coils. The ECU would always be in control, there was no bypass. A pair logic level coil signals was sent out to the ignition module, where they were stepped up via transistor and coils to produce the spark.
When I first abandoned the quad 4 ignition system, I switched to BIPs driving the stock coils (which would be the same thing as the twin cam arrangement) and the problem still remained.

Yup, this is a tuffy.

And I agree, the only way to "see" what the timing is doing is going to be to stick my head in the wheel well with a timing gun while someone else floors it on the Dyno. Not too excited about that idea.
however, if my Trigger Tooth logger shows This when under WOT and full boost/load, there is really no point in seeing it my my own eyes... the input signal goes crazy and the engine loses power, and sometimes blows. no reason to doubt the spark is happening at the wrong time. not only do I lose my proper tooth count, I also start decoding a Cam Input signal for some reason... and I don't have a cam input at all!

bad Teeth.jpg

[nukkinfuttz]

and here is what "good teeth count" looks like when im just cruising with no issues...

Good Teeth.jpg

[Six_Shooter]

Now, I'm mostly speaking from a V6 application, that for some reason is the exact same wheel as the 4 cylinder application. In the 6 cylinder application the notches coincide with TDC events of the cylinders, however only one notch coincides with TDC of one of the cylinders in the 4 cyl application, IIRC it's TDC#1 (home notch being 70* after #1 TDC). Anyway, there are 6 equally spaced notches on the wheel, only 3 of them are used for actual RPM reference by the ICM (Ignition Control Module). The consensus is that 3 of the equally spaced notches starting with the notch next to the home notch is used for starting (and base timing), after the bypass signal is applied, the EST signal gets an additional 60* of timing added to this signal to switch the the other 3 equally spaced notches, to move RPM reference away from the home notch to avoid false triggers, at least this is what some information seems to indicate. I know that using the Delco ECM with the 60 degree V6 and OHV 4 cyl applications require an additional 60* added to the EST signal, otherwise the engine does not run well.
It's because of this that I'm wondering if the MS is having a hard time deciphering the home notch from the others at high RPM.

However you said that you tried a 36-1 wheel and you had the same issue, so that leads me to believe it's something to do with your car and set-up.

The other ICM I mentioned was not a Quad 4 or twin cam application, but OHV 2.0L and 2.2L, which looks like the 60 degree V6 DIS, with only two coils.
I'll see if I can find a picture.

It's just a thought to move the ignition duties from the MS to the ICM that may eliminate some issues.

Another thought comes to mind, something I'm dealing with on a project of mine (Arduino based), that if there's noise on the power side it could cause false triggers. So far adding some input filter caps on the power supply has almost eliminated the false triggers, I just added some more filtering tonight that I have to wait until tomorrow to test. So what I'm saying is if you add some a couple filter caps to the power inputs of the MS it might help eliminate any noise. Also for a test run you could try disconnecting the alternator for a run (the source of power line noise in a car) and you will be running off battery which will have no noise on it. If the problem goes away, or lessens, then you know the power input is noisy and needs to be dealt with.


One more thought about power supply, again something I think I'm looking at in my own car is that there may be inadequate power supply to the EFI, or part of it. In my case I SOMETIMES have the ECM drop into LHM (Limp Home Mode, Not an MS, Delco currently), where the timing drops to base timing and all coms are lost for a second, and MAYBE even injection and/or spark, but it happens so intermittently and for such a short period of time it's hard to pinpoint, but due to a starter engagement issue I had it's kinda brought me to this possibility. Anyway... do you have everything running off the ignition switch, or do you have power feeds to various parts (coils, injectors, etc) run through relays to get good solid power to them? My plan is to run the injectors and ignition off their own relays to take the load off the ignition supply. In my case it's 42 year old wiring, and for some reasons I didn't add these relays when I first installed the system.

[shaheryarkhan]

ok guys for a month or so i didnt have time to get back at the project and resolve the issue. so we started working on it from last week and found the perfect solution for the swift gti motor and may be it'll help others too. as i previously mentioned few months back that the swift gti g13b motor has 2 wire vr sensor in its distributor with 12 teeth and it use to revs fine untill boost kicks in and under boost load ther was lots and lots of sync loss problem they we did everything to fix, but nothing helped and things get even worse. on boost load ignition spark was messing up the vr signal even with ferrite beads and expensive sheilded wires.

so after finding no hope with the cam vr sensor, we decided to move to the crank trigger setup. for that we needed a trigger wheel and a decent hall effect 3 wire or 2 wire vr sensor. in pakistan, we could not find any trigger from any make or generation model car to fit on our crank pulley. so we decided to make our own trigger wheel. but that was also a big problem cuz it was to expensive to get it cut and make it specially for our engine crank pulley or buy it online and had it delivered to pakistan. so in the end decision, we decided to try honda 70cc motorcycle rear sprocket as a trigger wheel. honda 70cc 125cc motorcycles uses 40, 38 and 36 teeth sprockets. so made a cut, less one tooth using lathe machine to fit the 36T sprocket on to the crank side alternator and ac compressor pulley and electric weld it to fit it. and as for the sensor, motorcycle sprockets have narrow teeth then the crank trigger wheel and any vr or hall effect sensor would have hard time sensing it so we give it a wild try and used swifts own vr sensor from the distributor. as it was the same sensor we did not change any jump points on the ecu, we just took it out of the distributor and mount it on the engine block 1.5mm space away with the sprocket trigger wheel.

we now tested it with 22psi of boost upto 7600rpm with out any hassle. motor runs fine and smooth like never before with same ignition coil and distributor. starts in seconds on half cranking.

Picture486-1.jpg

SP419.jpg

[SQLGUY]

The symptoms sound like the problems a lot of us had with the first couple of batches of Microsquirts. I thought the issue was finally fixed, though. For mine, I ended up ripping out the stock VR conditioner circuit and installing an NCV1124 instead. IIRC, I had to add a clamping rectifier in addition to the input resistor. Anyway, after this replacement, my 6'ish KRPM limit was gone. And, yes, on mine, too, the tach signals just went nuts after hitting the limit. Red line on my bike is 9500, and it revs just fine to that limit.

Here's the original thread: http://74.124.198.224/~micro/viewtopic.php?t=23135

[nukkinfuttz]

So I’ve tabled this project for about 5 years... just needed a break. I’m starting to mess with is again. I have placed my MS2 on the jim stim and ran it slowly up thru the revs. I noticed at right about 5800r’s my LEDs controlling the BIP’s have a significant brightness change as they tick over 5800r’s. I threw the scope on the singal to the BIP’s from the LEDs and find the square wave starts “jumping” around in its pulse duration on/off time. This is probably changing real spark timing on the engine even though it’s not reflected in any Datalog.

Next I changed the setting away from GM7x and to basic trigger on both the stim and my MS2. I find slowly reving DOES NOT ever alter the period of on/off time like it does when in GM7x mode! Nor do my LEDs all the sudden dim in brightness. So
It would seem something is going on with the GM7x code at around 5800 that significantly alters real engine spark timing and/or dwell.

First pic is of GM7x below 5800 threshold. Looks good and clean.
Second pic is above 5800 threshold, notice the falling edge is jumping back n forth in time period! That’s the problem!
Third pic is of basic trigger at above 5800, no problems, nice and clean

[nukkinfuttz]

And here is a video...

https://m.youtube.com/watch?v=qy7n_OOLU ... e=youtu.be

[jsmcortina]

Does the stim's GM7X pattern match your engine's pattern exactly?
The wheel decoder and spark control algorithms use "every tooth" so if the position of the +1 tooth was incorrectly coded or different, then the timing could vary when that tooth was used for timing.
This was an issue that was recently addressed on Renix where the wheel pattern coming in was inconsistent between vehicles and caused timing issues.

James

[nukkinfuttz]

Well JSM.... I tried shooting the pin 24 tach signal
Out of the jim stim while it was not even plugged into the MS2 to get a look at the pattern it’s producing... kinda looks like junk on my scope. Which is funny because the Tunerstudio tooth logger shows me 6 long duration teeth and 1 short duration tooth. Tooth logger would have you believe all is well, but the scope looks pretty erratic.



https://m.youtube.com/watch?v=LjUtQFG16 ... e=youtu.be

[jsmcortina]

That's a scope artifact. Try changing the way the scope triggers and make it display a single sweep rather than overlaying multiple sweeps.

For an even repeatable trace (like spark output) it is valid to use overlay to look for jitter.
For an irregular pattern (like GM7X) the scope will likely get confused by the +1 tooth.

James