With my 2500 rpm converter recently added, wanted to see how I could take advantage of it a little more it in the tune... did some reading up on S_AUJP launch control and it didn't seem like the right way to go. The launch control mode once activated, lasts for ~3 second, but by then, I'm well into 2nd gear... I needed something that's over and done with as soon as the I get into the converter...

Figured out I could use the Power Enrich spark advance to get what I wanted...

Since my converter actually flashes to about 2300 rpm, I used the 400 and 1200 rpm rows in the power enrich spark advance. The 1200 rpm row gets me to 1199 rpm (next row is 2000 rpm). Given there's essentially no load on the converter below the flash speed, I put enough PE spark advance to get to 28 deg up to 2000 rpm, at which point the ECM gets back to just the main spark table and resumes at 28 deg. Everywhere else in the PE spark adder is 0 since I have the main spark table exclusively governing the spark timing.

The TPS% threshold for PE is 50% (I left it stock AUJP). So in reality I actually get some launch mode at part throttle.

The result is I can definitely feel an improvement in the "kick" off the line. With the Ostrich I can do back to back runs with and without the added timing and it's definitely noticeable.

 

I don't believe this is an issue with Launch Control (LC), but rather with how it's implemented.

Note there is a flag that will reduce the TOTAL time to 1.6 seconds from 3.2 seconds. When this bit is set, it also reduces the time each LC spark value is used from 0.2 seconds to 0.1 seconds. But regardless of the bit setting, the entire time does not have to be used. LC spark is terminated upon encountering the first 0 entry in the LC Spark Table, whereupon spark control is returned to Main Spark Table. This bit was added in v6 of S_AUJP.

For cases such as yours where full spark is needed immediately at launch:
--- Set bit: V4 OPT 1 (0x98E), b1, Launch Control Spark Usage Time (reduces TOTAL LC time to 1.6 seconds and each LC Spark Table element will be used for 0.1 seconds after launch).
--- Set LC Spark Table something like the examples below to exit LC immediately after launch and let the Main spark table take over. These examples assume 10 deg base advance, so an entry of 0.35 deg SA-REF (the minimum allowed) equates to 10 deg SA-TDC, 10 deg SA-REF =20 deg SA-TDC, etc. (here, assuming 30 DEG is the all-in spark at RPM "X" in the Main spark table).

The examples show the SA values above after entry in the XDF where they are converted to the decimal equivalent of the hex calibration value.

Example 1: 3 table elements
Will be at SA-TDC after launch: 10.35 deg for 0.1 sec, 19.84 degrees for 0.1 seconds, 30.04 degrees beginning 0.2 seconds after launch and thereafter assuming 30.04 degrees is the all-in main spark table value at RPM 'X'.


Example 2:
2 table elements-Will be at SA-TDC: 10.35 deg for 0.1 sec, 30.04 degrees for 0.1 seconds after launch and thereafter


Example 3
1 table element (no pic)-Change the first LC element to what you want it to be, 0 thereafter. (may be too much advance)

Any of these settings should provide the required "over and done".

If you try this, please provide feedback.

HTH, Elky

 

Elky,

That's interesting... didn't know you could reduce the duration to 1.6 sec. However, I think it still may be an issue trying to run this added timing based on elapsed time rather than based on rpm...

I went back and looked at some recent data from a WOT launch... I'm well past the converter flash speed after 1.6 sec (time it takes my foot to go from 0 to WOT, time for engine get up to the flash speed, tires break loose when the converter hits causing rpms to really jump, etc). Seems like if I try to time it such that the spark adder shuts off by say 2200 rpm, that's only about ~800 ms. If I put 0 for everything after 800 ms in the table, that drops my timing down to 10deg for the remaining 800 ms (coincidentally I'm running 10deg base timing)...

Unless I'm missing something? Seems like once the launch control activates, it has to run its course before yielding back to the main spark table?

 

No, no and no .

!
Perhaps I'm not explaining it properly. The 1.6 seconds and 3.2 seconds are the MAXIMUM durations of LC if there is not a zero entry in the Table.
The important factor I believe you're missing is that the use of the LC spark advance table ends, and LC ends, once a zero value is encountered in the table.
Take a look again at Examples 1 and 2 above. In Example 1, LC is actually not deactivated until after 0.3 seconds (0.1 secs per non-zero table entry), but because the third entry contains the desired advance when LC ends, LC is for all practical purposes over after 0.2 seconds
After launch occurs:
--- 0.0 to 0.1 secs = 0.35 REF = 10 TDC (This is also the advance provided while "RPM Staged" awaiting launch)
--- 0.1 to 0.2 secs = 10.0 REF = 20 TDC
--- 0.2 to 0.3 secs = 20.0 REF = 30 TDC
So at the end of 0.2 secs, you're at desired TDC advance even though LC will continue active providing 20 REF/30 TDC for another 0.1 sec.
Same in Example 2, except the effect of LC only lasts 0.1 seconds despite being active 0.2 seconds because you're again at main spark table advance.

Bottom line, multiply the number of non-zero table entries by 0.1 sec, and that's the duration of LC. Make sense now?

Regarding RPM and why it was not used, consider this. You launch and have unwanted tire spin, LC thinks RPM is up and applies more advance. Not what you want. Also remember, this is poor man's LC. There is no transmission control (for RPM) and no selective cylinder misfiring. Both way too difficult to implement in $8d!

 

Ahhhhh.... ok, now I see what you're saying. So when the table hits a zero entry, it automatically exits the table.

Hmmm, I'll have to try this.

 

Bingo! LC is done. . Please report results. Tks.

 

Looking at TP, the V4 OPT 1 b1 actually isn't labeled....

But here's some recent launch data below...

So if I set 0x998 (Launch Control, Activate/Continue if TPS% >=) to say 50%, then per the data, the LC spark adder countdown would actually start at time index between 446.004 and 446.147 (my guess I"m dealing with the sample rate of TP here)? By that time looks like I'm at probably ~800 rpm at around 200ms. I get to the ~2300 flash speed at about ~800 ms.

Which means the elapsed time between getting to 50% throttle and ~2300 rpm is about 600 ms. So then I'd want to put in 5-6 table entries of 18deg (to add to my base 10deg) in order to get 28 deg for this 600 ms.

 

This has to be because you're not using S_AUJP v6 and the associated XDF here: https://www.thirdgen.org/forums/diy-...ml#post6345030

The 1.6 second, 0.1 sec per table entry is ONLY available in v6. Time to upgrade .

Addition:
But consider even if not using v6 and you only have 0.2 secs per table entry. With a converter active at 2300, why would you not use the RPM Staged option (MPH=0, in gear and RPM >= the RPM Staged Threshold, say 2000-2100), and with a low advance in the first table element and greater value(s) in 1 or 2 elements thereafter. Now, with foot on brake or line-locked, your at 2000-2100 RPM at < 50% TPS. Then when you floor it to 100% TPS, LC is immediately activated and you don't have to wait for the RPM to advance to the stall RPM. Rather, bam, you're immediately launched and taking advantage of the 2300 RPM converter wake-up because RPM is already there.

A question about the posted data. It seems to be a long time until 100% TPS. Did you ease into throttle or quickly floor it? Looks like you were "easy".

 

Finally got around to trying this out. I ended up setting the flag to use 100 msec per division and ultimately went with about 800 msec of launch spark based on the data. With 10 on the initial, I put 15 into the first entry and 20 the rest.

Hard to say though if this works better than the "poor man's" approach I was using earlier. But, I'll keep it like this since this is the way you're supposed to do it... lol

I've raaaaarely ever stalled the converter while holding the brakes before a launch... the first time I did that (to just try out the new converter), the tires were just spinning wildly out of control. It's pretty much an un-useable scenario. I'm basically just going for flashing the converter as quickly as possible during a hard launch from idle (and not even necessarily going WOT at that)... get the converter to hit hard on a moderately hard launch if I want a jack-rabbit launch from a stoplight. So I have the TPS% threshold set to 50%. Figure if I'm going 50% or more throttle from a standstill, I'm wanting to launch a little hard...

 

Note that the launch spark table requires SA-REF and NOT SA-TDC (SA-REF = SA-TDC - Base). So not sure about the 15 and 20 entries. If base is 10, those entries seem awfully high.

 

I believe it takes the value put into the base timing and adds on to that, no? The last time I tried this and was monitoring in TP, that's what I was noticing. I was shooting for 25 BTDC and then 30 BTDC since there's no load on the engine until the converter hits. Then once the LC event is over, it reverts back to the timing in the main SA table.

I have 10 deg set in my base timing, thus my 15 and 20 deg entries. I haven't looked at this in TP data, so maybe I should to verify what I'm getting.

 

Nope!!!! You MUST use SA-REF. The idea is to reduce timing to reduce horsepower for an instant at launch (and maybe shortly thereafter), to avoid boiling the tires. So if you're base is 10 deg, an entry in the first table element of 0.35 (the minimal entry), would provide 10.35 deg SA-TDC (SA-REF + Base = SA-TDC). Currently with entries of 15 and 20, you're respectively providing 25 and 30 degrees SA-TDC. Might as well disable LC if that's the SA-TDC you need. .

Elky