VNT calibration.

Help and tips about fine tuning. Post your questions or how to in here.

VNT calibration.

Postby Relic » Mon Oct 15, 2012 1:22 pm

Currently on my 18th revision of the VNT calbration map for the boost table.
Learnt lots of does and donts on the way.
1. Only way is to lock torque limiters in each gear to a fixed IQ. Adjust the table column by column.
2. 4th gear appears to be most stable or the turbo intertia just screws with everything trying to get consistant logs.
3. You are never gonna get requested = actual in every gear. Better to underboost in 3rd than overboost/surge in 4th.

What experience tips and tricks do you use to tackle calibrating the turbo with the wastegate to make requested boost = actual boost ?
:)

Relic
Diamond
 
Posts: 410
Joined: Mon Aug 06, 2012 3:18 pm
Location: England

Re: VNT calibration.

Postby ecuedit » Mon Oct 15, 2012 1:57 pm

That's very sensitive map. If you want to make as finest setting as possible you have to test it.
The best way to get over boost is:

- almost raining day (temp below 20 degrees Celsius).
- little hill
- drive at 3000rpm in 4th gear and than leave the throttle to drop to 2100rpm and than suddenly press the throttle to 100%
if you do not get over boost you can go farther

you can also drive highway in 5th gear around 2300rpm and do the same on the hill - fully loaded (5people on the board :lol:)
very important moment is to leave rpm dropping and than push to 100%.

Let me explain.

If you raise vnt for xxx % on Peugeot 406 2.2 HDI it will work fine,
but if you do that to the heavy Citroen c5 it will over boost.

Do you want to get numbers, as well or just those "tricks" i posted now :)

User avatar
ecuedit
CARPRO
 
Posts: 2221
Joined: Fri Feb 24, 2012 10:10 pm
Location: European Union

Re: VNT calibration.

Postby Relic » Mon Oct 15, 2012 2:48 pm

Thanks matey,

no just general tips and tricks.... problems... resolutions.
Know what you mean about the lift off and floor it.
Turbo slows down too much and then takes ages to catch up when you floor it so overshoots.
With the setting I have done peak boost never goes over requested....just sits bang on the requested line even in 2nd.

I did alos find 3rd gear would sit on requested but overshoot maybe 100mbar in 4th gear where turbo intertia was not an issue.
This is unacceptable from a surg epoint of view.
Thats why I abandoned 3rd gear.
4th and 5th seems pretty similar with little if any difference.
So I chose 4th.


I realised a few weeks ago when I started tuning mine requested boost was way out from actual boost.
Most off the shelf maps just ignore the wastegate and increase the boost maps instead.
Bit with the flowmaps I have I wanted to take it as close to surge as possible so I needed a bit more accuracy.
I wanted to be able to set the requested boost at the surge line rather than higher and know what I asked for is what I got.

Problem 1.
There was a 2.25" inch > 2x 2" standard exhaust on there with 2x CATs and 2x silencers (middle and back).
This is now a 2.5" short single exit with 1x cat and 1x mid silencer.
Sounds sweet but considerably changed the back presure.

Problem 2.
Even before this exhaust the car used to overshot by 150mbar + at 2000rpm.
After the exhaust it would also over compensate and drop 150mbar afterwards.... cutting MAF and fueling :wtf:

I toyed with an increased MAP, but decided I needed to fix these issues.
So I remodelled the OEM requested boost so that it would not surge unless it was over 30c and less than 960mbar.
{This dropped boost upto 3000rpm (but the calibration was way out)...requested was way over actual anyway}.
Then I changed fueling for 17:1 @ 30c / 960mbar (worst case midsummer).
Instead of 66mg @ 1750/2000, I was now at 74mg @ 2800 to get full use of the boost...upper rpm was also increased.
Probably around 180bhp @ 4000 instead of the 140bhp OEM for the same boost.

The way my VNT map works starts at 60% and drops to 20% at high rpm.
So I think it works in the negative manner to VAG maps..probably 40 ~ 80% I would think on these.
But after much trial and error I did find its incredibly sensitive to subtle changes.
1% is roughly about 30mbar
Although this isnt hard and fast as it depends on what the setting was in the cell before.
Some areas I had to dump -5% at low and high rpm.
Mid range around 3k was not so different.

Anyway, after 12 attempts getting nowhere fast :eh: I started doing it properly one IQ column at a time with fixed TL IQ in 4th.
60mg I am happy with.
55mg should be done tonight.
Then I will work on 65,70,75.

I did intend to do EGR off as I am overshooting with that switched off.
Might take a week or two to recover before I do that though :roll: :lol:

Relic
Diamond
 
Posts: 410
Joined: Mon Aug 06, 2012 3:18 pm
Location: England

Re: VNT calibration.

Postby Relic » Mon Oct 15, 2012 3:18 pm

Typical 3rd gear WOT before I started VNT work.
You do not have permissions to view the files yet. You have to be registered and you have to make at least 3 quality / unique posts.

Relic
Diamond
 
Posts: 410
Joined: Mon Aug 06, 2012 3:18 pm
Location: England

Re: VNT calibration.

Postby ecuedit » Mon Oct 15, 2012 3:25 pm

:) I remember when i was playing with N75 map - great times :wtf: .
Multiply your number of attempts at least by 10 :lol:

Every hardware change require some more testing...
In general if you do not do any hardware change than it is quite easy.

But with time and experience you get nice feeling.

I suggest when you test focus on one map, in your case calibrate N75 map until you reach perfection,
but if you change other maps too than you will destroy your calibration and you will be at the beginning again.

Make an order how you will change maps at the end play with VNT,
you set fueling, requested boost and so on, than at the end you see how faar you are with achieving the desired boost...

If you put more fuel, you get more gasses and and if you are facing overboost problems you will hardly
get to the point to catch correct values.

In short text:
Set all desired values and limiters, after that look in diag how far you are with readings to the requested value,
than you regulate valves to achieve desired values.

But you probably aready know all that.

I hope I explained well.

Hope that helps a bit.

User avatar
ecuedit
CARPRO
 
Posts: 2221
Joined: Fri Feb 24, 2012 10:10 pm
Location: European Union

Re: VNT calibration.

Postby Relic » Mon Oct 15, 2012 5:41 pm

Only another 100 attemtps to go eh ? :cry:

I was thinking last night when I change the boost and fueling I bet I will be back to square one :yawn:

Oh well! Its good practice I guess. And you are right. It is getting easier and quicker to guess what changes you need to make, how much and where.
:thumbup:

Relic
Diamond
 
Posts: 410
Joined: Mon Aug 06, 2012 3:18 pm
Location: England

Re: VNT calibration.

Postby ecuedit » Mon Oct 15, 2012 9:51 pm

I am 100% you will get it. :thumbup:

User avatar
ecuedit
CARPRO
 
Posts: 2221
Joined: Fri Feb 24, 2012 10:10 pm
Location: European Union

Re: VNT calibration.

Postby Relic » Tue Oct 16, 2012 1:53 pm

Getting there.
55mg.... need another +1.5% @ 4000
5 minute gap between logs a,b,c with no adustment.
These are the 1st 3 WOT runs after map change.

Bit better than above at least :D

55mg-4th-a.png

55mg-4th-b.png

55mg-4th-c.png
You do not have permissions to view the files yet. You have to be registered and you have to make at least 3 quality / unique posts.

Relic
Diamond
 
Posts: 410
Joined: Mon Aug 06, 2012 3:18 pm
Location: England

Re: VNT calibration.

Postby nexus665 » Wed Oct 17, 2012 11:03 pm

Hi,

great info - will need to do the same soon once my car's running again.

Blew its second turbo recently - I'm guessing the N75 map wasn't adjusted at all or not enough - since the injectors were swapped for quite a bit larger ones (.184->.216) and the 10mm pump piston for an 11mm one, they'd have to be to not make the turbo overspool a lot due to the massively increased exhaust volume caused by lots more fuel being injected.

Everything is decalibrated, injecting ~58mg/stroke max or thereabouts but diagnostics say 48,8mg max. Limp mode from overboost seems inactive, as well - would have at least given me a chance to save the turbo.

VCDS logs say boost overshoot is okay, but VCDS is very slow (1-2Hz) and cannot even catch these spikes IMO. Didn't have a manual boost gauge installed, but that's already ordered, as well. Stupid of me not to fit one right away after the first turbo blew.

So still a lot to learn, will be able to investigate better once the car's running again with the hybrid turbo and I've managed to read out the ECU (seed key changed or read routine patched, needs bootmode, no OBD read possible) and done some logs. I'm pretty sure that the N75 map will match the stock one very closely...will report.

Regards

nexus665
Silver
 
Posts: 71
Joined: Wed Oct 17, 2012 6:04 pm
Location: Austria

Re: VNT calibration.

Postby Relic » Thu Oct 18, 2012 1:55 pm

ecuedit wrote:If you put more fuel, you get more gasses and and if you are facing overboost problems you will hardly
get to the point to catch correct values.


Think a have hit one of your problems.
Never managed to test upto 4500rpm+ as I was running it 4th gear.
I have since dropped to 3rd to do the high rpm.

From 4250rpm onward I cant seem to calibrate the boost.
Doesnt matter how much I open the wastegate.
I am guessing either;
1. I am way off the cam. (Peak Oem power is at 4000rpm)
2. I am running too much fuel. (Too much exhaust to bypass)
3. My wastegate is not opening beyond 80%. (Although Duty % says that it is.)

This is torque limited to 55mg
You do not have permissions to view the files yet. You have to be registered and you have to make at least 3 quality / unique posts.

Relic
Diamond
 
Posts: 410
Joined: Mon Aug 06, 2012 3:18 pm
Location: England

Next

Return to Fine Tuning, Testing, Diagnostics and Analysis

Who is online

Users browsing this forum: No registered users and 0 guests