AFM voltage offset not reporting properly?

Nistune topics specific to the 6802 cpu

Moderator: Matt

Post Reply
RB30-POWER
 

Posts: 62
Joined: Fri Jul 20, 2007 4:43 pm

AFM voltage offset not reporting properly?

Post by RB30-POWER »

Hey Matt,

Well long story short, i've been road tuning the car for a few dyno days that have come and gone during the last fortnight.

I tuned the lower boost setting first around 9 psi, had perfectly flat AFR across the board.

Then when i was running higher boost (targeting 14-15psi) the car would go through this surge/hesitation around round 4000rpm, where the boost spiked, looking at the log showed lean afr, especially around the 4000rpm mark, ok i though, odd because it should calculate TP and keep AFR the same with increased air flow.

Anyway kept tuning around the hesisitation believing it was AFR related.

Got the AFR reasonbly flat, went to dyno day, done low boost run, 9psi, spun up 195rwkw, ok i thought this is good, i want over 300rwhp (225kw)

Dyno operator tried high boost and told me that it was pinging its arse off and backed off, resulting in 216rwkw at low rpm (4000rpm)

Dyno operater then tried middle boost setting around 12psi, same thing pinged its arse off.

He said minus some timing, so I expanded the maps because it was on the last load area during high boost and reduced the timing.

Took car out today, also thought it wise to drop the boost a little as well, just incase because it still seemed to have the hesitation/pinging to me.

Anyway put in on dyno, low boost run was 180rwkw, down 15kw on last dyno, went to middle boost done setting done 202rwkw, went high (but lower then last dyno) done 217rwkw, but i could still feel the hesitation there, just not as bad.

On the lower boost settings AFR was all over the shop, mainly rich coming onto boost.

Anyway came home, dissapointed again because the setup should crack 300hp at the treads but it didn't.

Long story short, I've been thinking, went into the logs and the AFM voltage was dead flat, which how can air flow ever be dead flat, so I worked out the AFM is maxed and i have been trying to tune around that, but i didn't think it was that because the software was showing only 4.6V (which i thought was low originally, but didn't think anymore of it)

So the software was/is not showing the correct voltage on the VQ map etc, it isn't adding the extra 1.5V to realtime readings, even though its in the adr file?

Is the adr file correct?

Anyways, ordered a Z32 AFM today, should be able to tune it properly.

But the stock AFM peaks out at approx 190-200rwkw(max)!!!!!!!!!!!!!!

Pics attached

this is the 15psi run, where he had to back off at 4000rpm due to pinging (because it was not calculating a higher tp and not going to higher ignition load point on map) notice afr are ok because i tuned around the afm, not realising at the time it was actually pinging.
Image


this is the middle boost setting run, 12psi done today, notice the afr dips, because this is the area where the afm is rich because its the highest load point the airflow meter can measure but not running 15psi today to use the fuel.

Image

a log, showing air flow voltage dead flat @ 4.6V, it can not be dead flat unless maxed, notice on VQ map its not adding the offset, also notice the area where the fuel map is heaps rich, this is as high as the afm would read, but it would get to that load point at only 9psi, that why after i done the high boost mapping not realising it was fucking my 9 psi map, because it was using the same cells.

Image

adr file showing offset should be added?
Image
PL
 

Posts: 957
Joined: Fri Feb 03, 2006 3:29 pm
Location: Adelaide, South Oz
Contact:

Post by PL »

Hi Michael,

As you are probably aware, Matt is away at present. Back Friday.

I can offer any insight about AFM offsets as I pretty much only tune SR's but I did notice something we have in common. The AFM voltage does seem to plateau after a certain point. Even though it's not topping out. I've seen this on almost every SR I've done - they nearly all use Z32 AFM so there's no way they're topping out.

Admittedly the AFM response is very non-linear so you'd expect it to plateau but I was surprised at just how flat it is. Pretty much gets to 4V at around 4500 and from there you only see 4.3V at 6500 (SR20DET + Z32 AFM):

http://www.plmsdevelopments.com/custome ... 1R_log.jpg

On engines running std AFM that top out I've noticed that you just get to a certain point and the mixtures just do a straight line upwards. It's quite distinctive.

I did an SR20DE+T awhile back and I've got logs of the AFM voltage both with std AFM (topping out) and after we fitted an RB20 AFM. The RB20 AFM didn't top out but as you can see, the two curves are remarkably similar. One's dead flat from 5900 onwards. The other is just CLOSE to dead flat! The difference to the mixtures was remarkable!

http://www.plmsdevelopments.com/custome ... fm_log.jpg

http://www.plmsdevelopments.com/custome ... fm_log.jpg

Pete L
Matt
Site Admin
 

Posts: 8961
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Post by Matt »

I'll have to get back to you on the offsets when doing log file recording/playback.

Double check it on the bench tomorrow night, doing a check against AFM input voltage. Let you know soon
RB30-POWER
 

Posts: 62
Joined: Fri Jul 20, 2007 4:43 pm

Post by RB30-POWER »

yeah, i could have always thrown a voltmeter on the actual output voltage i suppose, but it has to be maxed when you think about it, 200rwkw, honestly if it was really only half way on the vq map at that power, it means it would be good for 500rwkw, not likely.

either way, i have now thrown on a z32 air flow meter, done the hardware mods, changed vq maps, removed offset in the file etc.

haven't taken it out to adjust the k value yet, its raining...

when i do, i will let you know the approximate % of how much i had to change k value when changing the air flow meter for your software matt.

it defaulted changing my k from 372 binary to 936, that way to much correction my gut says, more like 500-600 should pull it up.

i'll start at 600 and go from there.

cheers.
Matt
Site Admin
 

Posts: 8961
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Post by Matt »

Raining over there? Not a cloud in the sky over here...

If you can get me that K value that would be fantastic. You now have a max reading for this AFM and we know the max reading for the Z32, so based on that can work out what the voltage difference compensation for these ECUs should be and make a BIG subtraction on the current calculated K value for these earlier JECS ECUs

FYI Z31 uses K value of 0x012F (303 in decimal) for bosch/ford cfi 440cc injectors/z32maf so I would use this as a starting point. So pull your K constant way back otherwise you will be running way too rich and foul your plugs before you even start

Let me know how you go with that

btw what values did you end up using for latency / latency change with your GTR injectors? Someone wanted a 'blind mail order tune' chip for GTR injectors on a R31 which I warned them against

However they persisted and I used 319 (0x13F) for K and 1540us (0x4D) for latency, but I left the latency change as 184 (0xB8) decimal. They reckon it runs fine with those settings but they havent been to a dyno yet to look at the mixtures
RB30-POWER
 

Posts: 62
Joined: Fri Jul 20, 2007 4:43 pm

Post by RB30-POWER »

The K value I was using with GTR 444cc @ standard fuel pressures (32psi @ 20" vac, 40psi @ 0" vac) was 372dec (174hex).

That was with standard Air Flow Meter.

I left the latency the same value as standard, from what I could tell, it was no different increasing that and using a lower K or alternatively leaving it stock value and adjust the K to suit. (rescaled the tp maps at same time of course)

This left the closed loop area with AFR around 13.5:1, which is far far better to have it trim fuel when O2 sensor activates rather then add it.

I found the map just after closed loop was a little lean was going leaner then 14.7:1, had to fatten all that area up to get 13.0:1 mixtures, top end on boost was all good really, until it maxed!

I found with standard latency value, when voltage dropped the AFR remained more stable and increased pulsewidth accurately, when latency value was increased, it was increased @ 14v, but below that voltage it went leaner then it should. (Not much, but enough to notice the effect it was having)

I don't agree with the latency value being used from the GTR ecu, as that ecu has resistors and saturated drivers, not peak hold style used in VL ecu.

The cold start maps had to be massively fattened up to get cold start back to factory conditions, (I still can't explain this one at all, got me farked, atleast you found the second map for it as well, why I was having trouble to begin with, just copied the same fattened map values into each map, perfect ever since)

Hope to get it out this weekend to adjust this Z32 AFM properly, hope its not double demerits because of XMAS, not worth the risk doing high speed runs when it double demerits.

Cheers
RB30-POWER
 

Posts: 62
Joined: Fri Jul 20, 2007 4:43 pm

Post by RB30-POWER »

wow, thats so strange.

changed the "k" value to 500 as a starting point, thinking higher to make up for lower reading z32 air flow meter.

started car, hmm running rich as fark...

eventually got the "k" value back to around 260 and the low end mapping was ok, but the top end way rich, so had to minus a fair wack of fuel in that area.

anyways, after a couple hours of tuning, end result is the k value is a little low, 02 disconnect, open loop, closed loop area is around 15.5:1 so the "k" should be a little higher.

i'm going to change back to the standard factory mapping in all fuel cells, then raise the "k" value a bit to fix the low end up properly.

but either way, the k" value will still have to be lower then the bosch air flow meter setting which was 372dec, i reckon after i tweak this one a bit, it will be around 280dec.

work that one out.

few other things i have noticed, but will start a new thread for that at later stage.
PL
 

Posts: 957
Joined: Fri Feb 03, 2006 3:29 pm
Location: Adelaide, South Oz
Contact:

Post by PL »

Sounds like an SR I did before Xmas. Your fuel pressure isn't too high is it???

ref : viewtopic.php?t=469

PL
RB30-POWER
 

Posts: 62
Joined: Fri Jul 20, 2007 4:43 pm

Post by RB30-POWER »

that would make sense, but i don't think that is the problem in this case.

i have a fuel pressure guage on the reg, so i can double check it easily to make sure anyway but.
PL
 

Posts: 957
Joined: Fri Feb 03, 2006 3:29 pm
Location: Adelaide, South Oz
Contact:

Post by PL »

Good idea. I'm starting to see the value in having a gauge fitted permanently!

PL
Post Reply