10 second 190,000 mile low boost 3.0 build

[87turbodance]

I don't exactly understand why Megasquirt won't allow you to have your timing less than the offset. Now I have to research :P

Megasquirt 1 and 2 use calculated periods of time to guestimate spark advance. Megasquirt bases all advance off of a reference point which is your trigger angle. Megasquirt will then calculate how much time to wait past the trigger angle to fire the spark. It cannot fire a spark before the trigger angle because of this fact. The stock distributor acts as a 6 tooth wheel run at crank speed.

If you want to simply things immensely, drill 6 evenly spaced holes in the outside of your damper 60 degrees apart with the first one aligning with a hall sensor when the crank is at 60 degrees btdc on any cylinder. Thread the holes and install 6 1/4" ferrous metals studs/bolts in the damper. Don't forget to loctite them

Configure MS for 6 toothed wheel with a trigger angle of 60 degrees

Then transfer the distributor wiring to the hall sensor and continue to use the distributor but let the crank sensor do all the timing work. Just adjust your distributor however you want to achieve the best spark range. ie set it up so that the rotor is close to the correct spark tower with a crank angle range of 50 to +10 degrees or whatever works out. When you have the distributor timing and the timing pickup separated it makes things soo much easier.

In my experience I could achieve a 60 degree spark angle range without cross talk in the distributor cap. Based on some quick math, I believe you will need to go wasted spark if you need range larger then 60 degrees because of limitation with using a distributor cap.

Bottom line, when you separate the distributor cap from the MS trigger input angle you gain the freedom to choose an ideal trigger angle and an ideal distributor cap position without making sacrifices. With the stock setup and MS you have to sacrifice one or the other to achieve something that works because both are tied together in the same component - you turn one and the other turns too.

[c2xejk]

I would recommend you turn on the enhanced accell enrichment algorithms. They do a much better job of creating a stable AF. Before I turned them on I would have crazy AF spike while cruising at a steady throttle.

[Ondonti]

Here is what I came up with in a visual format using the data I found with a distributor set close to "normal" oem position . Visual works better for my brain.

[Ondonti]

Timing drops to -10 on my 2 step (where it is actually set).
No more being stuck at 16 degrees or above. This makes high boost safety a reality.

I am showing the base spark map command and the actual spark and the idle corrective changes to timing.

[Ondonti]

Went to the track, even dragged reaper1 down there to take a video. He mentioned that the car sounded bad. I made zero test hits and didn't even drive the car after going through the timing system. I should have.

Made one pass on old 24.5x8x15 slicks. Bogged first burnout attempt and killed the engine. Restarted, burnout. 4800rpm 2 step at the line and eased out the clutch. Bit hard for old slicks. After grabbing, almost nothing happened for 1320 feet. I was misfiring badly all the way down and was down at least one cylinder on the return road. Cleared up but the miss did not go away.

So I might be crossfiring the distributor like Ed Kelly talked about or just have a bad plug/wire. Not very happy. 18.9 seconds at 47mph. Worst pass ever in any car. Above 6000 rpms the car just stopped accellerating in 2nd gear so I went to 3rd and the miss got worse so I gave up and hit the brakes. Pulled off at the finish line so I didn't pass the timing booth to get my slip. Caught it all on video. I couldn't really tell my car was running badly at first because I just assumed the car was much slower then my turbo Spirit.

I have to figure out why its misfiring so I can keep my current trigger angle setup that is boost friendly. I don't know why it would be crossfiring the distributor at 20 or so degrees of timing since that is about what my car runs at WOT for low rpms. It ran the same junky feel all the way through the rev range.

My only hint is that afternoon the car would not start and I had to fiddle with the distributor harness to make Megasquirt synch up with the engine. I have no idea what I fixed beyond wiggling and cleaning some wires.
I don't really think the video is worth anyone's time so I might not upload it.

[Vigo]

Yeah, sounds like a pretty boring video. On the bright side, you did get all the way through a burnout without breaking any axles so in that respect the duster's track performance may be improving.

[Ondonti]

Had my 3 year old help replace the rear view mirror that got knocked off and then since he didn't want go to inside yet (bedtime) we took out the spark plugs.
#5 was fouled with fuel. Not sure why yet. None of the plugs look like they should (very dark) but I know they were very sooted up when I was messing around with idle tuning etc and had checked one. Car has not actually run in a condition that would clean them off.

[Ondonti]

cleaned off the plugs, filed them a bit for fun with my son, then put them back into the car. Torched the one wet plug till it changed colors the other night. I figure the problem is deeper so I am not throwing a new set in yet.

Tried to start it and it has synching issues. Wiggling the distributor signal wire might be helping, it might not. I definitely got it to start and then wiggled the pigtail and it died instantly (as it had a few times while messing with it). Tried to drive it around knowing its not working and barely made it around the block while passing a cop. Lots of cutting out.

Swapped out the known bad distributor pigtail that had been previously repaired. Had some problems and figured out that I just threw the distributor back in without clocking it properly. DOH!

[Ondonti]

I messed around with the distributor LED output (on and off signal send by optical sensor inside the distributor) and ended up throwing another distributor in that I had in the shed. I set the megasquirt back from rising edge to falling edge and was more careful setting things up and actually used an adjustable positive stop in spark plug hole #1 to determine the exact top dead center of crankshaft rotation for cylinder #1. Set my degree wheel to that. I then messed around with spinning distributors by hand that were plugged in but removed from the engine. Everything worked with both distributors.

I can't exactly describe what I did but after messing around with other things like having the distributor off one tooth in each direction, I turned the crank to 10 degrees btcd (before top dead center) and also the cam gear to its proper rotation (up, its up every other crank rotation) and was able to get the distributor from the shed to turn on/off around 10 degrees btdc with the distributor in a normal position.
Previously doing 87turbodance's distributor sync setup I was forced to retard the distributor almost all the way. This must have been causing crossfiring at high timing numbers. I locked down the distributor in the appropriate location and I am going to walk away without figuring out what mistake I made before. I was previously setting it up on a long period of sleep deprivation. I was considering reclocking the optical disc inside when I couldn't get the signal to go on/off in the proper crankshaft rotation window.

[Ondonti]

I had to change the above settings from falling edge to rising edge otherwise timing was something like 50-80 degrees advanced. That got timing in the 17ish range. Changed my trigger offset from 75 to 81 and rev'd up the engine to stabilize the idle and it was about 10 degrees indicated. Turned back on "table advance" instead of fixed 10 degrees and timing is behaving properly.

I also activated the accell TPS WOT curve so megasquirt doesn't dump so much fuel at low rpms when I dab the throttle. Obviously I don't need 100% throttle at 800 rpms to get atmospheric pressure inside the intake manifold so the table is something more like 15-20% and rising with rpms.

[Ondonti]

Went for a drive while taking a datalog and I could barely get around the block. At least this time I remembered to grab the registration/insurance papers and put the license plate back over the kill switch before I drive past cops scoping my block.

I am getting accel enrichment fuel injector pulsewidth "bombs" where no matter how little I push the gas pedal, if I even cross the "engage" point for accel enrichment (based on TPS or MAP sensor % change over time) it will bomb 6ms of fuel injector even though in those situations I only needed ~2.0ms for normal running. That is a 400% increase in fuel and of course that causes a massive bog/misfire. I have a lot of situations where it should only be adding 1ms fuel and it no matter what dumps 6ms of fuel. Pretty crazy go to through the datalog and every time its 6ms of fuel when I try to accelerate.

What is frustrating about this is that there is a reference table for how much fuel to add and it doesn't seem to be referencing the table, just adding 6ms of fuel for 0.1 seconds no matter what. Other people had similar problems related to having their threshold too low but mine is at good level, the pulsewidth just doesn't reflect what my tables show. Another version of this problem is caused by MAP sensor noise (I dont have noise causing spikes). I am getting this problem even just when just TPSdot (% change over time) barely crosses the activation threshold. My accel is 50/50 MAP/TPS right now.

I never had this issue with the older firmwares. I don't understand why my car is so annoying lately.

[Sundance 6g72]

have you tried straight MAP based enrichment by setting all the TPS values to 0? Ive been running MAP based for a while and I have no complaints. Mainly because I did not have a TPS but I dont ill even bother switching back to TPS when I get my ECU installed. MAP worked well once setup. I just rev the car with quick throttle stabs to set the values. Just enough to prevent the car from going lean. I was able to keep AFR in check for small throttle changes and for roll racing. No lean spikes when stabbing the gas in 2nd at 45mph.

[c2xejk]

I have also seen those weird events. First, the MS log files do not contain all the data. So you can have things that happen that the log file doesn't see. I would start by double checking all the wiring. Then consider some additional filtering of the TPS. The size and duration spikes imply there is a significant amount of noise happening on that TPS input.

What helped most for my smaller rich pulses was to switch to the EAE (Enhanced Accel Enrichment). That really smoothed out the Inj Pulse Widths and AF. Got rid of the spikes that would pop up for no apparent reason as I was cruising at a steady state.

Took a little bit of effort to get it tuned in, but ultimately it was well worth it...

[Ondonti]

Yes I know I need to go and try 100% tps and also 100% map as well as drop the values to near zero and other things.

I am still quite bothered by the fact that its not following the chart. I also noticed that my old firmwares did not record some of the information I am looking at (actual accel enrichment) so I don't know what my car was doing back when it functioned properly.

I don't understand how noise would cause this (even though its spoken of a lot on the msextra forums) since my noise does not trigger events. The most noise I see is 46%tpsdot and that happens once during about 1 full minute of iding, then 3 events of low 30's. I have to hit 100% to trigger an event.

[Ondonti]

have you tried straight MAP based enrichment by setting all the TPS values to 0? Ive been running MAP based for a while and I have no complaints. Mainly because I did not have a TPS but I dont ill even bother switching back to TPS when I get my ECU installed. MAP worked well once setup. I just rev the car with quick throttle stabs to set the values. Just enough to prevent the car from going lean. I was able to keep AFR in check for small throttle changes and for roll racing. No lean spikes when stabbing the gas in 2nd at 45mph.

So for whatever reason, the TPS accel enrichment just doesn't work on my 3.3.1 firmware. I was looking at anything above 10% TPS, the accel enrichment does not activate, no matter how fast the tps opens over time. So, hold it at 11% and slam it farther open and nothing. But, wiggle it between 0 and 10% and it flips out. Has to be a megasquirt issue since TPS logs perfectly and TPSdot is a calculation based on the logged TPS.

I did a test switching between 100% TPS and 100% MAP accel enrichment. 100% map was behaving and giving enrichments that match my table for % over time. Obviously I needed to work on the numbers a little. 100% tps would now instantly dump 12ms pulsewidths of fuel no matter how little I pushed the throttle as long as it was above the threshold. No linear function following the table.

Modified the MAP table and went for a drive. Datalogged the drive. With a very very quick throttle stab I could get the motor to stumble but I couldn't capture what was happening in the datalog. I noticed that AFR follows MAPdot much more closely when driving than free reving....or thats just because my rpms were a little higher.
Car has throttle on grunt and it makes me happy.

[Vigo]

Well, at least you got proof of something workable to go along with your aggravation of the system not doing what it says it will do.

[Sundance 6g72]

firmware upgrades can be a pain, id just stick to map. The stumbles usually are lean or rich spikes (lean in my case). I was able to tune them out.

[Aries_Turbo]

FYI, ive had MS firmware get mysteriously corrupted before.

reflashed with clean firmware and then reloaded all the settings and all was well. this was on a MS2 on a 5.0L mustang but it shouldnt matter.

Brian

[Ondonti]

FYI, ive had MS firmware get mysteriously corrupted before.

reflashed with clean firmware and then reloaded all the settings and all was well. this was on a MS2 on a 5.0L mustang but it shouldnt matter.

Brian

My firmware has already been corrupted once earlier this year. It is very frustrating. That instance was due to low battery voltage while making changes.

There is already a 3.3.2 "stable" update so the 3.3.1 will probably not be going back in. The changes in 3.3.2 do not seem to matter for my issues so I am not going to obsess about it.

[Sundance 6g72]

just remember when changing firmwares you need to manually copy your settings in rather than importing them.