Eprom Tuning - Big Update

[RX8SE3P]

UPDATE: 17 Jan 2014

I've deleted my previous ramblings as they're useless now that it's been figured out.

We've figured out the stock timing maps and converted them to actual timing figures. This is for both the DE and ZE ECU eproms.

Also looked at some of the chip makers eproms. Probinator chip: beware using it. See page two for pictures of the map. I'm surprised we have not blown up many of our engines from it. Timing is only changed at WOT and it's very very high. It's basically a DE ECU eprom clone with WOT timing changed. Take a look. You'd be absolutely 100% safer using a DE ECU with a KL02 VAF or a KL31 ECU.

The low octane map, which I am unsure as to when is used, obviously has lower timing values. My car is confirmed to run off the high octane map, perhaps this is decided by O2 sensors and or by CEL codes/knock sensor etc. I am code free and I suspect if you're code free and using decent octane fuel, you will use the high octane map. Probinator's low octane map reads 0 across the board at WOT. This would make your car run crap at WOT and also very rich. I suspect this was my issue when I dyno'd my car and got 146whp despite a rebuilt ZE with some bolt on mods.

We are now able to modify timing and know what the exact figures are. This will allow for some improvement over the KL31 ecu and should also help Megasquirt users for their own tuning as they can now see (on page two) what the actual timing figures are on the stock ECU eproms.

[crazycanadian]

The emulator and Eprome tuning would be really interesting to see... I played with that a little bit when I was running my old 90 cavalier..

14.7:1 isn't the right fuel ratio for power gains... it could, even in NA form cause harm to the motor.. Leaning out can help but there are a lot of factors that go into tuning...

My old cavaliers V6 liked leaner A/F ratios.. Factory was 12.1 - 12.3:1 when I dyno'd it... There was almost 20hp to gain leaning it out to 13.4 -13.6:1 and advancing the timing... Unfortunately I don't know the long term affects of the tune... I got stupid and ran a dry 65 shot nitrous kit on the car, forgetting to removed the tuned eprom chip first... Melted #5 piston...

[RX8SE3P]

I plan to mainly get it leaner at WOT as 10:1 on my last dyno is just far too rich. I probably won't touch fueling on the lower end of the load x rpm scale as it seems fairly fuel efficient and happy below 4k. I'd be happy with somewhere between 13-14 AFR at max load + rpms.

Not sure how many of us have gotten gotten good AFR with a chipped ECU?

If my friend lets me, I'll post up some fuel maps for comparison soon.

[marcdh]

It'd be great to see the factory timing advance!

[RX8SE3P]

Stock KL31 Eprom here. Remember these aren't actual numbers, it's just the way it appears so it's hard to know in the timing map what "213" is. I doubt it's 213 degrees advanced, there's hidden multipliers depending on the EPROM too. So this is why it's hard and not straight forward. The only way to figure it out a bit better will be to use a timing light and check the timing whilst messing with the tune on the fly.

[Thefez]

never mine lol...

[RX8SE3P]

never mine lol...

Hey?

Probinator

[RX8SE3P]

I've posted the low octane maps too now. About to post a KLDE stock eprom. I guess if the ECU detects knock it will use the low octane settings?

Most of the DE eproms look like this:

[RX8SE3P]

You can see in the timing map that even the stock KL31 seems to have better timing than the stock DE. Probinator timing seems to be identical to the DE yet even lower timing in the low octane map.

[RX8SE3P]

Huge Update - We figured out a couple of massive things today!

First of all, the timing graph load and rpm arrows are wrong. It's actually the same as the fuel map. Load is at maximum and RPM are at maximum on the bottom right corner.

We made sense of this by using a timing light that could measure advance by using a dial. For example, if you wanted to measure over 20 degrees you adjust a dial and it makes the reading visible. I forget what the type of timing light this is but it worked great.

We then reduce all timing figures on the high octane map (ecu seems to use high octane map if there's no issues) to a specific number and it keeps timing at that number. From there we went up 10 points on the map all the way to 200 measuring the timing. At a value of 40, the timing is almost -1. Every 10 points on the map increased timing by about 2-3 degrees consistently throughout the entire map.

We are now comparing to people's megasquirt maps to see the best timing values for power. The KL31 map is pretty good but in some spots compared to my friends MS it's off by over 10 degrees. There could be a huge gain in the mid range rpms from 2500-5K rpms. But of course we have to test further.

Map Value - Timing (approx)

42 - -1
52 - +1
62 - +3
72 - +5
82 - +7
92 - +9
102 - +11-12
112 - +14-15
122 - +17-18
132 - +20
142 - +23
152 - +25
162 - +28
172 - +30
182 - +32-33
192 - +35
202 - +37

So you can see the KL31 map at max rpms and load reads 167 which would be about 29 degrees advanced. The KLDE ECU reads 152 which is about 25 degrees. It's about 4 degrees off which would contribute to some power difference. The probinator is weird and is at over 200 which would be around 37 degrees advanced. I don't even think MS users go that much advance!

[RX8SE3P]

I'm just looking now at some MS maps and picked out one of Fowljesse's maps. I have read that his KL setup runs quite strongly and is naturally aspirated so it's a good start to compare timing with.

In comparison I can see that the KL31 map has much lower timing in the lower rpms.

Fowljesse's (sorry if this is an old map) has timing at around 26 degrees from as low as 1100rpms (top line of map).


Compared to the KL31 at max load (look at bottom row of table) you can see that the KL31 map at about 1100 has a value of 94 which is only around 9-10 degrees at best. At 3000rpm the MS map has 27 degrees of timing and continues to about 29-30 till the end. At 3000rpms the KL31 only has a value of 125 which is around 18 degrees. Still way off the mark. Only until 6000rpms does the KL31 map get to around 162 value which is 28 degrees!

We just have to figure out the fuel too and we should have a very good alternative to MS. Still will never be as good but hopefully we can close the power gap!

[Ryan]

Good job reverse engineering it. This will give people a good, reliable base to start their MS tunes. I'm interested to see what the stock maps look like with actual ignition advance values, it should give us new information on the performance of various hardware combos based on what packages came factory with what ECU.

[RX8SE3P]

Thanks, here's the thread on Probetalk that my friend put up http://forums.probetalk.com/showthread. ... 1924146229

He's the brains of the project - our numbers are a tiny bit different but only because I rounded them up a little different but I'd go with what Stoker100 has posted.

He's already made a conversion table so you can see the KL31 timing mad in degrees. He plans to run his MS to match the KL31 map and see how it runs. He's also sent me an Excel spreadhseet with the DE and ZE ECU timing compared (assuming the DE timing works the same, we may have to test).

[MrMazda92]

Here we go, yet more evidence that tuning bridges nearly all of the gap between the "ZE" and the other KL variants!

Well done, RX8SE3P, and as Ryan said you gave a very good starting point for those wanting to mess with MS as well! Myself included... I'll try to throw together 3 tunes, for KL01, KL31, and my own custom tune, so that I may dyno the differences next year when I finally take the plunge.

[RX8SE3P]

Stoker100 made a conversion spreadsheet in MS Excel. It's pretty accurate, we may be off by a degree here and there but he reckons it's within 3% tolerance. Just not 100% sure about the rpms and if it actually goes to 7980 or not. Doesn't matter much because timing tops off on the DE at around 6000rpms and never goes above 24 degrees. The ZE climbs to 27 degrees at around 6800 and most people would be making their peak power close to that with their ZE's.

Note that the table on the bottom is flipped. So the table on top reads 167 at the bottom right... which is 28.54 degrees of timing on the top right for the table below. Hope that makes sense!

KL31



KLDE