tip in spark retard aka Transient Ignition Retard

[muythaibxr]

Once we schedule a spark event, if the tooth it's scheduled from changes, we don't update that event... it happens then we schedule the next one, so if you could go from high timing (35-40 degrees) to a lower timing (20ish) in the time between spark events (very small, especially at revs higher than idle), you might be able to cause a problem. However, there's no real way to solve this one.

Ken

Ken,
Do you think the manifold can fill that quickly?

so far seems like non-issue to me.

On an ITB engine, there isn't much volume behind the throttle bodies, and that volume can fill very quickly. I don't however think it's going to fill enough between ignition events to cause us to be very far behind. You're certainly not going to go from cruise kPas (say 40ish) to 100 kPa in 1 step between ignition events.

I think worst case you might be 1 or 2 ignition events behind before everything catches up, which means your timing will barely be off from where you want it... it's not going to be a difference of actual timing of 40 degrees when you wanted 20, it'll be more like actual timing is 20.3 when you wanted 20. That's my initial take on it without doing the math anyway.

Ken

[muythaibxr]

So on a 4 cylinder at 800 rpm (idle for most 4 cylinders that I've seen) you end up with an ignition event every 37.5 ms. So that would be what I'd call worst case for a 4 cylinder.

How much can MAP change in 37.5 ms? 1500 rpm is 20 ms, 2000 rpm is 15 ms, etc... as you increase RPM to the areas you're likely to be driving hard in, the problem becomes drastically smaller if it was a problem to begin with.

Unless there is some other physics-based reason to do this, I don't think it's entirely necessary.

Ken

[techsalvager]

so far I've found that retarding spark in these low high rpm areas has helped with the issue.
4500rpms~
10-30% engine load
durning shift, then I slam open the throttle and engine load goes straight from there to 150%~ engine load
before my timing map had 38-39 in that area.
And knock was showing up after every shift and wot tip in event
throw in 20s all around that low load high rpm area
my knock has calmed down drasticly.
I need to do more testing though to make sure.

my knock control is though a j&s safeguard that has been hooked up to my megasquirt so I can datalog when knock events occur.

[dieselgeek]

so far I've found that retarding spark in these low high rpm areas has helped with the issue.
4500rpms~
10-30% engine load
durning shift, then I slam open the throttle and engine load goes straight from there to 150%~ engine load
before my timing map had 38-39 in that area.
And knock was showing up after every shift and wot tip in event
throw in 20s all around that low load high rpm area
my knock has calmed down drasticly.
I need to do more testing though to make sure.

my knock control is though a j&s safeguard that has been hooked up to my megasquirt so I can datalog when knock events occur.

Have you had this engine on a chassis or engine dyno? What I've found is, guys tuning to the edge of knock are *way* past their true MBT. Also, consider keeping things a bit rich even at low RPMs and high load, just like you would at higher RPMs. This is how I tune most engines I work on, and we don't see knock, and we get good transient throttle response etc. Consider trying these or other tuning strategies is my suggestion.

[techsalvager]

I've had it on the dyno once, next time I plan to tune in part throttle mbt, among other things for driveability.
whats a bit rich at high load low rpms to you for a turbo car.

at 4500 I'm not to worried as I won't be cruising in that area generally, but if it happens around 4000 to 3500 then I will have issues as I also need that area tuned in for curise for gas mileage.

[dieselgeek]

I've had it on the dyno once, next time I plan to tune in part throttle mbt, among other things for driveability.
whats a bit rich at high load low rpms to you for a turbo car.

NA or turbo, say at 100kpa at low RPMs like 1500, 12.5:1 isn't too rich.

If you get a chance, get on a mustang brand chassis dyno and steady state your engine at a couple low RPM / high load points, and play with the timing in realtime. Advance until the torque number stops going up. You will probably be surprised that it's a low number, safely away from detonation/knock/preignition.


We did this on a DTS dyno to map a 6.1 Hemi a few weeks back (NHRA Drag Pak stock eliminator engine), MAN does it snap with throttle once you have the low rpm/high load areas mapped this way.

[techsalvager]

no chance of using a mustang but I do have access to the same dynojet. Last time I only touched fuel, still learning, I am allways learning. But now I will do exactly that, part of the plan to tune mbt at part throttle and all around if possible, going for better responce as well as driveability.
the dynojet has a load cell on it.

I need to compare I have a smaller turbo on the car and I go into boost or near boost pretty easy on just daily driving.
Generally I leave it around stoich at 100kpa until I hit around 2psi before pulling fuel, but I really need to do some scientific way of testing to see if it will suck down gas more if I run it at around .85 lambda vs 1 lambda.
and with that I wish there was like a power enrichment type table to use so if I did go over a certain throttle percentage it would switch to a different commanded fuel
that or a closed loop commanded fuel table and a open loop commanded fuel table. and switch over to the open loop commanded fuel table after certain variables have been hit.

[Peter Florance]

no chance of using a mustang but I do have access to the same dynojet. Last time I only touched fuel, still learning, I am allways learning. But now I will do exactly that, part of the plan to tune mbt at part throttle and all around if possible, going for better responce as well as driveability.
the dynojet has a load cell on it.

I need to compare I have a smaller turbo on the car and I go into boost or near boost pretty easy on just daily driving.
Generally I leave it around stoich at 100kpa until I hit around 2psi before pulling fuel, but I really need to do some scientific way of testing to see if it will suck down gas more if I run it at around .85 lambda vs 1 lambda.
and with that I wish there was like a power enrichment type table to use so if I did go over a certain throttle percentage it would switch to a different commanded fuel
that or a closed loop commanded fuel table and a open loop commanded fuel table. and switch over to the open loop commanded fuel table after certain variables have been hit.

At 100kpa most modern engine I tune run 13.2 or richer. I do the same with turbo car and then even richer above that. This is always on a dyno and none of my cars ever detonate unless we have poor charge temp control

[muythaibxr]

In my experience most people tune the high load low rpm areas with too much advance as DG says.

I don't think this feature is necessary... I think it's a band-aid for a problem with the main ignition table.

Ken

[techsalvager]

Its not the high load low rpm I am seeing issues.
its low load, high rpm

transistion rom high rpm high load at say 20 degrees spark timing to low load high rpm of say 35-38 degree spark, back to high load high rpm going down to 20 degrees of spark

[jsmcortina]

Any chance of posting some SDcard datalogs of this at say 5ms interval including all relevant variables:
RPM, MAP, TP, SparkAdvance, TPSDOT, PW etc.

James

[muythaibxr]

Its not the high load low rpm I am seeing issues.
its low load, high rpm

transistion rom high rpm high load at say 20 degrees spark timing to low load high rpm of say 35-38 degree spark, back to high load high rpm going down to 20 degrees of spark

That *is* still at high load not low load... it's the transition from one to the other. I still say you probably have too much timing at the top part of your table, or between the low part and the high part if it's knocking when you stomp the throttle.

Ken

[muythaibxr]

Any chance of posting some SDcard datalogs of this at say 5ms interval including all relevant variables:
RPM, MAP, TP, SparkAdvance, TPSDOT, PW etc.

James

An msq would be good too... It would help us to know if you have things like 1st derivative prediction enabled (which you shouldn't need with a higher tooth-count wheel).

Ken

[richyvrlimited]

Generally I leave it around stoich at 100kpa until I hit around 2psi before pulling fuel,

Isn't stoich at 100kpa far too lean, no wonder you're lighting up the J&S like a christmas tree?

[techsalvager]

Generally I leave it around stoich at 100kpa until I hit around 2psi before pulling fuel,

Isn't stoich at 100kpa far too lean, no wonder you're lighting up the J&S like a christmas tree?

no I don't have any knock at stoich around 100kpa or into slight boost.

I will upload some logs and the tune plus ini is below

uploadtune.msq

[jsmcortina]

Just wondering what you included the ini file for?

James

[dieselgeek]

no I don't have any knock at stoich around 100kpa or into slight boost.

Caution!

Lately I've been doing a lot of individual cylinder AFR monitoring and tuning, and I can tell you with confidence that you are very likely to have 1 or 2 cylinders running way lean of stoich at 100kpa, if your average AFR is at stoich. In fact I haven't seen an engine yet that, up to 100kpa, doesn't have significant differences in AFRs with the exception of ONE that was running nearly all OEM components - intake, exhaust, cam, etc.

Be careful. The reason the OEMs run 12.5:1 AFRs in that same load range is not because peak power is made that way, bet because of cylinder-to-cylinder variances.


Meanwhile, you may find it interesting that the guys at top racing engine shops run their engines in the high 13s AFR once all 8 cylinders AFRs are matched.

[billr]

James, maybe I can help out about why the .ini In another thread I pointed out that I am not able to open all .msq's, using TS Lite. The error message is something like "missing .ini or .cfg file" (don't have exact wording right now) and I am prompted to browse for it. I abort at that point to avoid corrupting any of my files. Right after the post where I mentioned that problem a new post was made (maybe even this thread) that included the .ini with .msq I downloaded and saved the .ini, then opened the .msq in TS. Sure enough, the missing-.ini error popped up and I browsed to that .ini just saved and TS opened up the .msq just fine. Probably all related to TS vs TS Lite, right?

[jsmcortina]

Probably all related to TS vs TS Lite, right?

I don't think that full vs. Lite makes any difference. I have a project created for each Megasquirt variant and use the relevant project to open MSQs. BUT... as I'm a developer I've already got just about every ini file that's ever existed here already.

James

[billr]

When I open these posted .msq's I always open them in a "temp" project; if I start collecting .ini's in that temp folder will TS select the correct one automatically?