N54 HPFP VCV Tables

KClemente

Sergeant
Nov 26, 2019
284
147
0
Ride
E90 335i
I still have to read through the documentation, but I've had ip_cur_sp_prs_ctl_vcv defined for a while now and have yet to really test it. It seems to be a calibration for how much current = how much pressure.

The other table we all have is the current setpoint for flow setpoint. Not pressure like the one I listed above
 

carabuser

Lieutenant
Oct 2, 2019
870
1
765
0
UK
Ride
Z4 35i & 335i
Had a few more drive cycles this weekend and have had the idle rail pressure hunting happen at the end of the drive cycle 2 out of 8 drives. Based on what carabuser said above, I probably have no idea what I am doing.

That being said, I am going to keep tinkering with the 2 main tables. I am open to any guidance or to test new tables that have been discovered.
I'm out of the country at the moment but when I get some time back at my PC I can offer some additional logging channels to pinpoint your problem. You're on IJE0S right?
 

mj6234

Corporal
Nov 25, 2020
161
58
0
I'm out of the country at the moment but when I get some time back at my PC I can offer some additional logging channels to pinpoint your problem. You're on IJE0S right?
That is correct. IJE0S. If there is a document floating around where I can put in some time and help around identification, I am willing to put in time. I work in data/software engineering so have some foundation to build on. Not an assembly level programmer or anything like that, but not a total dumb dumb either.
 

carabuser

Lieutenant
Oct 2, 2019
870
1
765
0
UK
Ride
Z4 35i & 335i
That is correct. IJE0S. If there is a document floating around where I can put in some time and help around identification, I am willing to put in time. I work in data/software engineering so have some foundation to build on. Not an assembly level programmer or anything like that, but not a total dumb dumb either.
The most helpful part would just be a data log of the issue occurring showing the rail pressure.

Once I get home I can put a few key parameters into the custom log file like the VCV PWM, fuel pressure target and maybe some of the internal fuel mass targets.
 

mj6234

Corporal
Nov 25, 2020
161
58
0
The most helpful part would just be a data log of the issue occurring showing the rail pressure.

Once I get home I can put a few key parameters into the custom log file like the VCV PWM, fuel pressure target and maybe some of the internal fuel mass targets.

I will get a log of the idle behavior. I would also vote for the P value for the HPFP PID if possible. I think that might be the most telling one.
 

mj6234

Corporal
Nov 25, 2020
161
58
0
Sorry for the additional spam/thoughts/questions here, but trying to make sense of things here. Feel free to tell me I am a moron.

(1) Does the HPFP have an internal pressure release valve? I assume it does, but curious at what pressure it is set to?
(2) Looking back at all of the ads for the Helix and the POD, they show these flatlined charts pegged at 3200psi rail pressure. After thinking about that, there is no way that is indicative of a system operating properly with the PID active right? Is it just pegging at the relief valve pressure? If so, that seems not something that I'd feel good about.
(3) Does everyone just target 20MPa+ in the Fuel Pressure target table after installing the Helix? If so, why? For testing in the top row I am targeting 18MPa now (2600psi) to test.
(4) Are we SURE the 2 VCV tables have an impact on HPFP operation? I just want to see someone else say it before I spend more time tinkering with it. I believe I have seen changes in rail pressure behavior after modifying the tables, but don't want to convince myself I am seeing something I am not.
(5) If someone confirms 4 is true, to me I still have the flow to current table too high - especially in the 4500-5500 range. And I have adjusted the middle (20-195 L/H) cells down quite a bit in terms of current (and ultimately PWM).
(6) After the highway driving and long pull, I expected to have the racing rail pressure at idle, but did NOT have it. In fact after sitting in the garage, the rail pressure was very steady @ 715psi. I am going to go ahead and post up my current HFPF/VCV tables, but they are definitely A WORK IN PROGRESS AND USE AT YOUR OWN RISK.

Here is a log. I meant to get more ethanol content, but I guess the E85 pump is more like E65. Don't put too much stock in the spool shown in the logs, I went WOT at 2500rpm in 3rd gear. Frankly this is my first long WOT pull since I put in the Helix. I was half expecting my HPFP to get launched through the hood.

Mods (if it matters).
GC Lites
PFS Stage 2 LPFP
Helix (3x one)
5" ETS FMIC
N20 MAP
VSRF DPs
AFE Stage 2 Box Intake
VTT Inlets/Outlets
Flex Fuel
IJE0S Self Tune from stock bin with a lot of revisions (I know I could prob up WGDC in the upper RPMs. I havent started scaling boost up by ethanol content because I was capped on fuel at E40 @ 20psi. But will be working on that.


Current HPFP/VCV tables.

1650930623194.png
 
  • Like
Reactions: wheela

mj6234

Corporal
Nov 25, 2020
161
58
0
I hope this is in the spirit of this thread so others can see changes to tables and impact on logs, if people think it is not, I can take my testing offline.

Another log after some more VCV tinkering and still no rail pressure racing at idle since the changes before this (probably 8 drive cycles). I see I need to change some things up in the mid range regarding boost control. I am trying to not over-do it in terms of mid range boost, but ya I know that is doodoo tuning.

Wondering if the drop off at the end of the log is not enough HPFP DC, I will test some changes up there.

Sort of disappointed the LPFP seems to be close to out of juice already. I thought it could make a decent amount more than this :/.


Latest VCV settings.

1651017675791.png
 
  • Like
Reactions: Jeffman and wheela

mj6234

Corporal
Nov 25, 2020
161
58
0
New log. Might be a few days before another due to rain. HPFP held to target at 6800 in 3rd after the latest changes. To me, it still looks like too much HPFP DC in the mid-range. Target is still 18MPa (~2600psi) and was a few hundred psi over before it got all spiky.

I *think* these are reasonable guesses on the L/H consumption
Idle - 1.5 L/H
Highway cruise - 12 to 15 L/H
WOT at 6000 RPM in 3rd gear at 20psi - ~190 L/H

I am guessing the 3000 RPM is maybe 100 L/H.

Anyway, I will test some more to see if I can get that smoothed out or not.


Tables

1651165907027.png
 

mj6234

Corporal
Nov 25, 2020
161
58
0
OK, new thinking and first log with changes. The amp to DC table, I was thinking maybe it should be left stock. Reasoning, the pump is spinning faster, but the VCV operation itself hasnt changed - meaning the same amps should give the same duty cycle on the valve even after overdriving it.

So I reset the amp/DC table to stock and started messing with rescaling the flow to amp table. I can definitively say changes in this table matter. I reduced the 1 L/H and 2 L/H cells too far and the car would start, but the rail would bleed pressure until it stabilized at what the LFPP was outputting. I got a fuel pressure plausibility code on one of the attempts. Interesting, the first set of values that I got to that DIDNT do that were very close to stock. I guess that might make sense for several reasons.

Under that assumption, I took a stab at rescaling the table on a similar pattern, just with a flatter slope. This log is with that. Just from glancing at it, I may have gone too low, but I will be trying more values to see if I can get the curve smoother.

Last thought, I am not sure I am convinced changing the amps to DC table does anything. Is there a concept of 'reference tables' in the DME that arent used to dynamically control operation, but are there to show calibrations? I wouldnt think so due to the size of the bin and the need to store a ton of information, but it is a thought that crossed my mind.


1651191934247.png
 

carabuser

Lieutenant
Oct 2, 2019
870
1
765
0
UK
Ride
Z4 35i & 335i
You're on the right track with this. I've been doing the same and reducing the values in the "flow request to current request" table is helping to smooth things out. I logged all the parameters relating to the VCV and it's clear that with an upgraded HPFP setup you are overshooting the rail pressure setpoint which causes the VCV to back off abruptly.

I've put the 3 key params in here for you to log. The FUEL_MASS_REQ is what the DME is asking for. That uses the aforementioned table to generate a current value for the VCV (also in the log file) which is then converted into a PWM signal (also added to the logging).
 

Attachments

  • IJE0S_UserChannels.zip
    373 bytes · Views: 46
  • Like
Reactions: mj6234 and wheela

mj6234

Corporal
Nov 25, 2020
161
58
0
I will do more logging this weekend. What folder do I drop that in to pick up the new logging channels?
 

carabuser

Lieutenant
Oct 2, 2019
870
1
765
0
UK
Ride
Z4 35i & 335i
The MHD root folder. That varies on android version but it's the same folder the stock bin gets generated in
 
  • Like
Reactions: mj6234

mj6234

Corporal
Nov 25, 2020
161
58
0
@carabuser is the MAN!! Seriously, how can I give something back to you for the help? I can send you a few bucks, test something, or do whatever.

All logging channels are in place and haven't used them to update the tune. My latest guess produced this, which seems like it is in the ballpark. Just some fine tuning to do here.

Also have been working on boost curve. I learned how critical WGDC to WGDC Position is to spool. And now that is fixed, I realized my WGDC base is running a little hot!!! Please ignore the throttle closure in there, I will fix.

That being said, here is the log and the tables.


1651274149047.png
 

carabuser

Lieutenant
Oct 2, 2019
870
1
765
0
UK
Ride
Z4 35i & 335i
@carabuser is the MAN!! Seriously, how can I give something back to you for the help? I can send you a few bucks, test something, or do whatever.

All logging channels are in place and haven't used them to update the tune. My latest guess produced this, which seems like it is in the ballpark. Just some fine tuning to do here.

Also have been working on boost curve. I learned how critical WGDC to WGDC Position is to spool. And now that is fixed, I realized my WGDC base is running a little hot!!! Please ignore the throttle closure in there, I will fix.

That being said, here is the log and the tables.


View attachment 67384
It doesn't take long to get a few logging params so don't worry about it. If you ever need a tune then you now where I am.

Looking at that log it seems pretty slow to spool. Something up with the turbos or wastegates not closing.
 

mj6234

Corporal
Nov 25, 2020
161
58
0
It doesn't take long to get a few logging params so don't worry about it. If you ever need a tune then you now where I am.

Looking at that log it seems pretty slow to spool. Something up with the turbos or wastegates not closing.
I am definitely still learning on the tuning, so something could be off. I am on the original boost solenoids, maybe could be that. It looks like my WGDC is at the limit (75%) for almost all of the spool period. All my logs are 3rd gear and I have GC Lites. If you have a good spool log from hybrids I would definitely like to see where I can improve.
 

carabuser

Lieutenant
Oct 2, 2019
870
1
765
0
UK
Ride
Z4 35i & 335i
I am definitely still learning on the tuning, so something could be off. I am on the original boost solenoids, maybe could be that. It looks like my WGDC is at the limit (75%) for almost all of the spool period. All my logs are 3rd gear and I have GC Lites. If you have a good spool log from hybrids I would definitely like to see where I can improve.
If those are GC lites then you have a big boost leak, stuck wastegate or bad boost solenoid.
 

mj6234

Corporal
Nov 25, 2020
161
58
0
If those are GC lites then you have a big boost leak, stuck wastegate or bad boost solenoid.
Got it. 159k miles. Maybe time to change the solenoids. I am pretty sure no boost leak, I have had the turbos for a while - 75k miles, but don't think the WG is bad.
 

mj6234

Corporal
Nov 25, 2020
161
58
0
@carabuser - I got the racing idle captured with the new log parameters. Oddly the current request (and unsurprisingly HPFP DC request) didnt change BUT the fuel mass request tripled. I am pretty close to dialing in the VCV tables (see next post), so I really am not sure what might cause this?

Tapping the throttle dropped things back into normal, but the cycle will repeat.

 

mj6234

Corporal
Nov 25, 2020
161
58
0
@carabuser - I got the racing idle captured with the new log parameters. Oddly the current request (and unsurprisingly HPFP DC request) didnt change BUT the fuel mass request tripled. I am pretty close to dialing in the VCV tables (see next post), so I really am not sure what might cause this?

Tapping the throttle dropped things back into normal, but the cycle will repeat.


My log approximately 90sec after this behavior happened. Still working on a few things, I am a little puzzled by the drop off up top in rail pressure, but overall it seems like the line is less choppy and somewhat close to target for most of the pull (minus the end). The HPFP is literally close to brand new - bought from FCP to replace the dead one probably 8K miles ago. I am running E33 or something for this pull, so there is no way it should be out of capacity. LPFP pressure doesnt seem terrible, so not sure I can blame that?


1651448571638.png