Spark hardware latency, or timing light issue?

[aw113sgte]

Single cylinder COP, using "wasted spark COP)
"dumb" COP, driving it with BIP373.
Polarity is verified as correct.
Timing set as static 10degrees BTDC
I am using inductive pickup on the high voltage output of the COP.

Pickup light measures a stable 10 degrees at idle. When reving, timing retards. I had to add 100uSec in the spark hardware latency to keep the timing at 10 degrees.

Now should there be any need for the hardware latency with a COP setup? I have some doubts about the timing light as I was running a relatively normal timing map and I was getting massive detonation under boost.

[elaw]

Is this a "smart" timing light with a dial or electronic method of setting advance, or a "dumb" one without such a feature?

"Smart" ones can read wrong under a number of circumstances. Best bet if that's what you're using is to get your hands on a "dumb" one and try that.

[aw113sgte]

Is this a "smart" timing light with a dial or electronic method of setting advance, or a "dumb" one without such a feature?

"Smart" ones can read wrong under a number of circumstances. Best bet if that's what you're using is to get your hands on a "dumb" one and try that.

It's a "dumb" one. No dialback feature. This one: http://www.amazon.com/INNOVA-3551-Induc ... B000EVYH72

[billr]

COP is no different than any other type of ignition, there always will be some turn-on/off time for the circuit/device that controls current in the coil primary. If I did my math right, that 100usec equals 3.6 degrees of retard at 6000 rpm; is that correct? Does it seem to be a retard that is linear with increasing rpm?

[aw113sgte]

COP is no different than any other type of ignition, there always will be some turn-on/off time for the circuit/device that controls current in the coil primary. If I did my math right, that 100usec equals 3.6 degrees of retard at 6000 rpm; is that correct? Does it seem to be a retard that is linear with increasing rpm?

Yes the retard is linear with increasing RPM. Problem is, the latency could be in the ignition system or in the timing light. Obviously both will have some latency but I don't have any way to measure them.

[billr]

Try another light, borrow one from a friend or rent from a parts store. Post where you are located, maybe another member is located close enough for you to check your engine with their light and their engine with your light. Post the MSQ, I have no clue right now what rpm and boost levels you are talking about (and note levels where detonation is "massive"). Post a run log.

[aw113sgte]

Try another light, borrow one from a friend or rent from a parts store. Post where you are located, maybe another member is located close enough for you to check your engine with their light and their engine with your light. Post the MSQ, I have no clue right now what rpm and boost levels you are talking about (and note levels where detonation is "massive"). Post a run log.

Well since I don't have a knock sensor (I would be there doesn't seem to be a decent setup for the Microsquirt now that the KnockLite isn't produced). This is a motorcycle so everything is very loud, I did hear a faint rattling at WOT, shortly after that I lost compression and ended up with a piston with signs of detonation (local melting of the top) and smearing of the aluminum on the skirt.

wr250x-supercharged.msq

2015-08-15_13_modified.zip

[billr]

Okay, now that I know it is a single-cylinder engine, a few questions:

Why are you using "wasted COP" instead of "single coil" ? Why are the spark dwell and duration rather low? Is the 12-1 crank wheel actually on the crank, or is it on the cam?

Also, spark timing seems pretty aggressive!

[aw113sgte]

Okay, now that I know it is a single-cylinder engine, a few questions:

Why are you using "wasted COP" instead of "single coil" ? Why are the spark dwell and duration rather low? Is the 12-1 crank wheel actually on the crank, or is it on the cam?

Also, spark timing seems pretty aggressive!

Wasted COP, seemed the most common used. Would it even make a difference in function? I'm really not sure of the difference.

12-1 is on the crank flyweel. A direct connection.

Spark timing I got from looking at a lot of high performance motorcycle engines. Then I decreased timing under boost .5PSI/lb. Maybe that's just too much timing.

[grom_e30]

you are not subtracting much timing under boost, normal safe starting point is 1-2 deg per psi of boost

[kaeman]

looking at your ve table I was wondering about the fuel range. I noticed in the 5000 rpm range that 50 kpa bin is using 100+ ve value and that 170kpa is only using 135 ve. I would expect to see a much larger change in fuel requirements you only raised the fuel about 30% with a 120 kpa increase in airflow? I am wondering if the engine shouldn't be requiring more fuel that that. I am also noticing that your battery voltage is very unstable... jumping around between 11.6 and 14 volts. that might need addressing before the other issues can be tackled. the large fluctuations in battery voltage will affect dwell, fuel and spark.......as well as making other sensor readings unreliable.

[aw113sgte]

looking at your ve table I was wondering about the fuel range. I noticed in the 5000 rpm range that 50 kpa bin is using 100+ ve value and that 170kpa is only using 135 ve. I would expect to see a much larger change in fuel requirements you only raised the fuel about 30% with a 120 kpa increase in airflow? I am wondering if the engine shouldn't be requiring more fuel that that. I am also noticing that your battery voltage is very unstable... jumping around between 11.6 and 14 volts. that might need addressing before the other issues can be tackled. the large fluctuations in battery voltage will affect dwell, fuel and spark.......as well as making other sensor readings unreliable.

About VE values, I'm not sure that's how they work, it's not like pulsewidth which directly indicates the level of fuel. If you look at the AFR datalog, I'm not running lean under boost. Most of that table was made with the VE analyze tool.

As far as battery voltage...you have a good point. I have no idea why it thinks voltage swings that wildly. I put in a buffer on the dash because it jumped around so much. I don't think the voltage is actually swinging that much, I think it's how the microsquirt is reading it. I'll have to see if the voltage is affecting fueling...I think I can look that up ion the logs.

[aw113sgte]

you are not subtracting much timing under boost, normal safe starting point is 1-2 deg per psi of boost

I was doing a lot of reading with the turbo busa guys and they often didn't pull any timing at 8 PSI and if they did, only 1/2 PSI/lb boost. All engines are different though!

[kaeman]

if the microsquirt is showing battery voltage fluctuation, I wouldn't trust the afr reading... my guess on burning the piston with a rich afr is that its pretty difficult, burning the piston running lean is pretty easy. I know that you could be just experiencing a bad ground connection at the megasquirt, but I would give my power and ground connections a good check out. I know that I had some strange behavior with my unit and went back and redid all the power and ground connections and the strange behavior disappeared. GOOD LUCK

I know that I have to be careful when using ve analyze because my engine was having a rich misfire and it kept throwing more fuel at it because the afr goes lean when it misfires, it maxed out the ve value using the ve analyze, I finally manually removed most of the fuel and the misfire went away. I believe that the ve table is a direct link to the pw, if you increase your ve value, you will increase the pulse width proportionally.
This is the fuel calculation: PW = DT + (ReqFuel * MAP * Stoich/AFRtarget * VE[RPM,MAP] * AirDen * BaroCor * corrections)

[aw113sgte]

if the microsquirt is showing battery voltage fluctuation, I wouldn't trust the afr reading... my guess on burning the piston with a rich afr is that its pretty difficult, burning the piston running lean is pretty easy. I know that you could be just experiencing a bad ground connection at the megasquirt, but I would give my power and ground connections a good check out. I know that I had some strange behavior with my unit and went back and redid all the power and ground connections and the strange behavior disappeared. GOOD LUCK

I know that I have to be careful when using ve analyze because my engine was having a rich misfire and it kept throwing more fuel at it because the afr goes lean when it misfires, it maxed out the ve value using the ve analyze, I finally manually removed most of the fuel and the misfire went away. I believe that the ve table is a direct link to the pw, if you increase your ve value, you will increase the pulse width proportionally.
This is the fuel calculation: PW = DT + (ReqFuel * MAP * Stoich/AFRtarget * VE[RPM,MAP] * AirDen * BaroCor * corrections)

Good info, I got the bike back together today so I'll be looking into these details. After giving it some thought, I bet the voltage fluctuation is from the injector driver. It is on the same 12v line as Microsquirt and positioned inside the case. I wouldn't be surprised if those high amp firings of the coil cause the fluctuation in voltage. I'll try running it straight off the battery with a separate wire and see what happens. Probably should move the driver out of the case as well.

I'm not getting any misfires that I can feel. From what I've read, when detonating it's easy to burn components as the internals absorb more heat (exhaust temp drops as the heat isn't in the air, it's in the piston/head). Basically I could have been fine on AFR but the detonation caused the overheating.

[aw113sgte]

Well I relocated the ignition driver and ran a separate ground to it but it did nothing to change the voltage fluctuation. I am also running a lithium battery, perhaps that has an effect. I'll try a lead acid and also see if a large capacitor in parallel has any effect.