Loses cylinder(s) at high boost/rpm

[letitsnow]

I put my IC on the datsun over the weekend(2.25" pipe, evo 8 IC, mystery oem BOV(maybe subaru?), t bolt clamps and silicone couplers everywhere). I had to reloacate the IAT sensor in the new piping. That covers the most recent modifications.

 

At ~7psi everything is dandy and it pulls to 5500 just fine. At 12 psi, 4600rpms and it feels like it looses a few cylinders, when you let off the gas it's running very badly on 3-4 cylinders. It clears up shortly after and runs on all 6 just fine. From boost onset to ~4400 the car pulls very hard, it doesn't break up or lose cylinders at all.

 

I just changed the plugs from bpr6ey's to br7es's thinking that the non-projected tip might do better with the higher boost, no change. The ignition system is an MSD blaster 2 coil fired directly by megasquirt, NGK wires, and the plugs previously mentioned gapped at .028.

 

AFR's are 11.5ish at 12psi and 11.5 falling to 11.0 at 7psi. One of my datalogs shows it holding 5000 rpm for 1 sec at 10 psi like it was on the limiter, boost was 12 falling to 11 before it hit 4900ish. From 4000-4500 took 1.1 sec for reference.

Edited July 19, 2011 by letitsnow

[letitsnow]

found the distributor cap was pretty cracked, replacing it seems to have helped a ton, I think the rest is in the tune.

[letitsnow]

Nope, still losing them. Gapped plugs down to .018, that seemed to help a little, maybe. Maybe my coil is getting weak, maybe plug wires, maybe time for 36-1 and LS1 coils?

[Cannonball89]

Sounds like the problem I recently had. Try reducing your spark duration settings. I have the msd blaster 2 coil triggered directly by megasquirt and my spark duration setting is currently at 1ms, dwell time is set at 3.5ms. If spark duration is too high, it will drastically reduce dwell time of the coil at high RPM which will cause the coil to not be charging long enough to produce a spark powerful enough to ignite the denser mixture. You could probably even reduce spark duration to below 1ms if needed. It doesn't really take a long time for the spark to initiate a flame front, but it does require a lot of voltage to jump the plug gap, which is why dwell duration needs to be about three times longer than spark duration at any given RPM.

 

For single coil applications, Let's say X= number of spark events per second, and if X=(RPM/60)(# of cylinders/2), and if Y= milliseconds between spark events then Y=(1/X)(1000). Do the math and see how little time there is in between spark events with a single coil. Megasquirt will scale the sum of your max. dwell time settings and your max. spark duration settings until they fit within the window of time available for whatever RPM you are at. So as RPM increases, dwell and spark duration decrease. The trick is to have dwell significantly higher than spark duration so that dwell never drops below unacceptable levels, which will very upon how much boost you are running. It is important to understand that the max. dwell time setting dictates how long the coil will charge and the max. spark duration setting will dictate how long MS will wait before initiating the next charging cycle, and as RPM increases these numbers will be reduced by whatever percentage is needed to fit into the timeframe between ignition events.

 

I am running 12PSI with a .030 plug gap on NGK BR8ES plugs using the same coil you are. So I don't think you need to run a wasted or sequential ignition system yet unless your power band is above 6500 or you plan on raising the boost signifigantly higher.

Edited July 26, 2011 by Cannonball89

[Cannonball89]

Also your pretty rich, Leaning it out some might make the mixture easier to ignite (maybe, although some say that a rich mixture is easier to ignite).

Edited July 26, 2011 by Cannonball89

[letitsnow]

Spark duration is set to .1ms, the manual says that's fine for 99% of installs so I went with it, I'll give a higher number a shot.

 

Last night I checked out the wires while it was dark out, I'm getting a bunch of glow/corona at the plug ends. I'm going to try a bunch of dielectric grease on them to see if it helps any.

[Cannonball89]

0.1ms is pretty short, I think you mean 1.0ms?

[letitsnow]

Nope, .1ms is straight out of the manual. http://www.msextra.com/doc/ms1extra/MS_Extra_Software_Manual.htm#dwell

 

Adding dielectric grease seems like it helps some, so I guess a set of wires is what I need.

Edited July 28, 2011 by letitsnow

[letitsnow]

Bad news, I did a leakdown test.

 

This is pressure remaining, from 100psi or thereabouts, cold engine. Home made leakdown tester, .040" orfice(std size).

 

1: 60psi

2: 40psi

3: 43psi

4: 25psi

5: 68psi

6: 70psi

 

As a check I did one cylinder on dad's firebird, 90psi.

 

This thing felt really strong for leakdown between 30 and 75%. It ran a 14.2 at 98 with ~9psi last weekend.

 

I do have to wonder if I broke it or if it was broke when I got it, it has never idled as well as I thought it should.

Edited August 11, 2011 by letitsnow

[2eighTZ4me]

Are you getting any blow by out of the valve cover? When I broke piston skirts and ringlands due to excess boost, it was smoking like a chimney out the top of the valve cover.

[letitsnow]

Yes, I drove around with the hood off for a little bit after I put the intercooler on, at idle it was fine, any time I went near boost there was a pretty decent fog coming from the little breather filter I've got on the valvecover.

 

 

 

 

I've got the engine out and apart now, the pistons are.....perfect. They are .50mm over N/A dished pistons(everything else is L28ET), one of the previous owners put a set of L28ET rings on it, the top ring is 1.5mm, the N/A ring groove is 2mm.

 

I haven't updated this thread much, so since my last update I've: swapped to MS2 and retuned, new wires, boost leak check and repair, and went to the track. At the track, I ran a best of 14.2 at 98mph on ~9psi, 2.2 60', every time I added boost I lost mph. I also had a best 60' of 1.9 with an open diff and street tires, that tells me this thing is very nearly a 13sec car with almost no cylinder sealing and not much boost.