CX-7 tune with modified exhaust

[torquemaniac]

Anyone here done some tuning on a CX-7 with a custom exhaust (some help with the WG DC table would be really nice

I flashed the 93/98 octane load tune, but I have boost spike (and fuel cut) when I go WOT!!!

Its odds also that I reach a load of 2.23 with only 230 g/s...some guys on mazdaspeedeforums.org get to 260 g/s with a load of only 1.9 (manual transmission)...I have the automatic transmission...could the added stress (load) on the engine be coming from the automatic trans...

I dont think that the high load is due to the boost spike, because i can gradually buildup the boost with the accelerator pedal, and I still get to around 2.2 load

Thanks

[mituc]

For starters I'd say lower the BCS DC Base table values with 10% (even 15) between 2250 and 5000rpm. It would be nice to post a log (exported in CSV so that we can all open it) and see what's going on. A 10-15% decrease may bring you below the boost cut threshold and it's still within the -100 - +20% WGDC compensation so you will have good values to work on for dialing it perfectly.

The load is a direct result of air flow and RPM, among other things. If you're doing 230g/s at 5500rpm that's a load of, say, 1.6 (I don't know for sure because I didn't calculate it, but let's just say). If you're getting 230g/s at 4000rpm that's a load of 1.6*5500/4000=2.2.

For now, until you change the BCD DC table, stop going WOT, if you exceed load 1.9-1.95 and you don't have aftermarket fuel pump internals you risk blowing the engine.

[torquemaniac]

Thanks for the info on the BCS DC table adjustment (this is what I was looking for)

When my maf cal is done, I will post a log (I got fuel cut when I was logging the WOT part of the maf cal)

I already have fuel pump upgrade (see signature:)...what really concerned me is more a bent rod

One more question : when i do the maf cal logging, I can see that at idle im around -9.4ltft and it tapers to -8.4 at around 3 volts (maf voltage), all this with the desired and actual lambda pretty much matching all the way...when i get into open loop (3+ volts), the ltft goes to zero and the desired and actual lambda still match pretty good...is this normal...this seems weird to me...the trims goes from -8.4 to zero in seconds, but the desired and actual lambda are matching...shouldnt it be more gradual???

Thanks

[mituc]

The negative LTFT at idle is your MAF cal which right now it's too rich.
When in open loop (WOT) the LTFT's should go to 0, so that's normal too.

I just took a look at your mods too, i didn't do that before. You say there you have a MS6 turbo. Well, the CX7 AWD K04 turbo has a smaller compressor wheel as far as I remember, so that is the reason why with a BCS DC table designed for a CX7 turbo you have these issues.

[torquemaniac]

I can tell you that there is another major difference between the CX-7 and MS6 turbo : the exhaust intake on the CX-7 turbo gets very narrow (almost like the end of a trumpet), while the MS6 turbo as a more progressive taper and the throat is bigger

I guess the CX-7 turbo is like to increase the velocity of the exhaust gas (faster spool)...but its one hell of a restriction on the exhaust side!!!

What intrigue me about my fuel trims, is the abrupt change from closed loop to open loop...what I mean, is that in close loop the ECU makes corrections to keep the desired and actual lambda equal (-9.4 LTFT), but when i go into open loop the LTFT naturally goes to zero (no more corrections), yet my desired and actual lambda are still equal...so im running rich in close loop and then magicaly stop being rich in open loop???

Its like I have to make corrections in the MAF calibration table up to about 3 volts, then no corrections in the table from 3 volts to 5 volts (this makes for a big jump in flow between the close and open loop voltage)

Am I making some sense

[mituc]

If your MAF cal is fine in one area it doesn't necessarily mean that you're not miles off on some other area(s).
The ECU has some logic to learn a bit based on the Open Loop short fuel trims though. However, in open loop long therm fuel trims simply don't work.

Those almost -10 negative LTFT's in closed loop mean that in that MAF cal segment (delimited by two MAF cal break points) the values are off by an average of -9.4. Multiply those g/s values with, say, 0.95, smooth the courve a bit and see what happens. Also take a good look at the MAF cal tutorial on this forum.

[torquemaniac]

Maf cal is done and within ±2 (ltft)...

I looked at lowering the BCS DC after 2250rpm, but after looking at my log, I see that I never hit my load target in the 500 to 3000rpm (actual load is lower than desired load)

I lowered the BCS DC after 3000rpm, which kind of helped with the boost spike, but I find it odd that I never hit my load target in the lower rpm

When I look at a given rpm and APP in my log (500 to 2500rpm), the App requested load does not match with the actual load...even the DBW throttle and BCS DC logged values dont match with the table (at the given rpm of course)

Sassimac told me that the requested load are not really target, but rather maximum value...does this make sense...why command a value in the table if you never get to it?!?!

[mituc]

That's a target and a limit.
However, what is the target load and what's the load you actually get in that rev range?

Lowering the BCS DC values will not necessarily affect that unless the needed compensation percent is higher than about 20% (which is the limit for positive BCS DC compensation to hit a load target). But lowering it will help with the boost spikes, because instead of sudden and violent negative compensation the ECU will command higher and higher wastegate duty cycle to meet the load target.
One possible reason for not hitting the load targets is because you lowered the BCD DC values too much and you're doing the pulls in a lower gear (even 3rd may be considered a low gear if the necessary compensation is too big) and there's never enough time for the ECU to figure what's the needed BCS DC values in some rev ranges.

P.S.: ideally you should not go WOT as early as 2250rpm (or lower than like 3k rpm).

[torquemaniac]

Mituc,

I did not lower the BCS DC in the 500 to 200rpm range (kept as per the Versatuner OTS map : 500 to 1500 is at 100%, and 2000 is at 70%)

All other range (2250 to 6500) were lowered by 5% (I upped the load target and absolute load a bit also, so no more fuel cut)

I heard before of not going WOT before 2250rpm ...why is this so critical if the turbo doesnt reach full spool until 2500rpm...meaning there is no real power until the turbo is spooled!?

FYI, I am not judging your teaching...I just want some clarification

Talking about spoolup, I noticed that since I changed the turbo and downpipe (done at the same time), full spool happens at 3000 versus 2500rpm...I thought that the downpipe would compensate for the MS6 K04 slower spool, but this is not happening

I will post a log with the App load table so you can see the problem with load target in the low rpm range

Thanks for the great support Mituc

[mituc]

The MS3/MS6 factory K04 can make full boost at 2500rpm on stock intake/exhaust. This means that they will make some good boost even at 2200rpm (it can basically push air past atmospheric pressure even at 1800rpm), even if it's far from the maximum values. When you open up the exhaust that 2500rpm threshold can shift as low as 2200-2300rm depending on your mods, VVT advance, timing, and so on.

Try changing the BCS DC base tables and go to, say, 81% for the 2000rpm column.
I wonder if you could load the BCS DC base table and load tables in an excel sheet and post them here. Posting the tune may not help because we need the CX7 versatuner to read it and we don't have it...