Testing MAP sensor with vac gauge - due to rough idle

For non EFI mechanical discussion
Pete30nz
Posts: 246
Joined: Mon Aug 19, 2019 12:32 pm
cars: 1994 VR v6 berlina
1996 VS v6 executive
Location: Christchurch, NZ

Re: Testing MAP sensor with vac gauge - due to rough idle

Post by Pete30nz »

Ahh ok, so its more for real time tuning at a guess?



Heres my log taken during the drive to work this morning, attached
Attachments
drive to work.xdl
(517.27 KiB) Downloaded 81 times
User avatar
antus
Site Admin
Posts: 8237
Joined: Sat Feb 28, 2009 8:34 pm
cars: TX Gemini 2L Twincam
TX Gemini SR20 18psi
Datsun 1200 Ute
Subaru Blitzen '06 EZ30 4th gen, 3.0R Spec B
Contact:

Re: Testing MAP sensor with vac gauge - due to rough idle

Post by antus »

Pete30nz wrote:Ahh ok, so its more for real time tuning at a guess?
The idea is that if you know the AFRs the ECU was targeting, and you know the actual mixture from wideband, and you know the current VE, you can calculate how far out the VE is from that data. So the idea is that you capture it over a long drive, then playback and average the calculated VE in a histogram.

But in practice it doesnt work so well because

1) enrichment and enleanment are not taken in to account by the logic in the tunerpro ADX
2) lag in the measurement of the data is not known or taken in to account when calculating the new VE

so what happens is changes are applied to VE cells that are close but not completely right. This causes some cells to go further out of whack instead of closer, especially because of the delay. Your tables should in general be reasonably smooth but with this method the errors make the tables more and more spikey. A spikey looking table creates erratic fueling which when added to the errors above make things even worse. So you end up with a worse tune rather than a better one.

Whats better if using calcve is overlay target AFR, wideband AFR, calc VE in a monitor. Now play back and pause your wideband log on an area where things are pretty stable as far as load and rpm go, and where the wideband afr is furthest out from target afr. check that your not in decel enplanement (eg you want to see RPM going up, not down). Now turn on bubble tracing and look at the VE table and slide the playback marker in the monitor area around the part your looking at back and forth, and see which part of the VE your talking about. Then look at calc ve and see how far out you are in that area. Then select that area in the VE, and apply, say +4 to the area (for 4% greater VE/fuelling if thats your best guess as to how far out that area is) and then smooth the edges back in to the rest of the map. Now test again and see if you were right, or if you need more or less. Repeat and dial in the worst bits of the table until fueling is looking good.
Have you read the FAQ? For lots of information and links to significant threads see here: http://pcmhacking.net/forums/viewtopic.php?f=7&t=1396
User avatar
vlad01
Posts: 7780
Joined: Mon Oct 08, 2012 6:41 pm
cars: VP I S
VP I executive
VP II executive
VP II executive #2
VR II executive
Location: Kyneton, Vic

Re: Testing MAP sensor with vac gauge - due to rough idle

Post by vlad01 »

I had good success with VE learn but the other fundamental settings need to be right so the VE is consistent over a variety of conditions. I got very accurate high res VE data using it for the final trimming.

When other settings aren't right this function will go haywire and ruin the VE table.

My only complaint is that the NVRAM can get corrupt using it as it's possible to get a bad write during key off at the wrong time. I had it happen.


I had one case where I didn't know it was on and one a new modded engine it tuned itself over some 30 min pretty damn good. I was scratching my head to how a bucking and backfiring out of tune engine came good during the logging period when driving for the first time. Once I worked out what was going on, I downed loaded the cal again and most of the VE was already sorted lol. It does work well !
I'm the director of VSH (Vlad's Spec Holden), because HSV were doing it ass about.
User avatar
antus
Site Admin
Posts: 8237
Joined: Sat Feb 28, 2009 8:34 pm
cars: TX Gemini 2L Twincam
TX Gemini SR20 18psi
Datsun 1200 Ute
Subaru Blitzen '06 EZ30 4th gen, 3.0R Spec B
Contact:

Re: Testing MAP sensor with vac gauge - due to rough idle

Post by antus »

thats the first report ive heard of it working well, very interesting and good to hear.
Have you read the FAQ? For lots of information and links to significant threads see here: http://pcmhacking.net/forums/viewtopic.php?f=7&t=1396
User avatar
vlad01
Posts: 7780
Joined: Mon Oct 08, 2012 6:41 pm
cars: VP I S
VP I executive
VP II executive
VP II executive #2
VR II executive
Location: Kyneton, Vic

Re: Testing MAP sensor with vac gauge - due to rough idle

Post by vlad01 »

Interesting regarding the WB lag. I actually set the VE learn settings to be faster and more aggressive and it worked really good. I used the 14point7 WB so perhaps it was more than fast enough for the task.

I have the spartain2 kit now and that is quite a bit faster than the 1st gen I had.
I'm the director of VSH (Vlad's Spec Holden), because HSV were doing it ass about.
Post Reply