Tune for 3 bar map sensor

Hi,
if in car sits stock 2.5 bar map and putting in 3 bar map sensor, how to calculate or make boost map to see more than those stock 2500mbars?

Pull the tuning info from a PD130/150 with a 3bar sensor on it

Do you mean just copy boost maps from PD130 file?

map linearization should be changed from 2600 to 3120 for correct reading. leave the 200 value the same
do this on all 3 maps (manual,auto,etc)
use vagedcsuite program to do this

for boost map and boost limiter, this depends on what turbo you have.

Awesome, thanks

What he said!

If you look at PD150 file..map linearization goes to 3000 but that value's axis number is a bit lower. Recalculate it for ours, it becomes 3120.
And you can verify....w 3000 key on/engine off... atm sensor says around 1000 map said 990. with 3120, map now matches atmospheric

nowere havent seen what to do for 4bar map sensor?
how to make calculation,will my tuner man know what to do and what i have to say to him?

Your tuner should know...
But some of them just play with numbers, not knowing why.

Its simple: if the old said 2600
then its 4160
(3 bar would be 3120)

If 2500, then its 4000. (some ecu calib map at lower max v)
(3 bar would be 3000)
200 stays 200.

thanks!
i ve done more search and find a litle bit more info,but yours is more precise
I would like to learn more of this,ecu maps,where can a get cable and software for tuning(for my car)i have search e-bay found nothing,maybe i didnt search wright,....
does anyone have maps for my car,original,tuned,tuned for bigger injectors,turbo,egr delete,...just asking....

get mpps clone from ebay
sign up on ecuconnections....read and learn about edc15v
make your own map...by testing and adjusting

robnitro said:

Your tuner should know...
But some of them just play with numbers, not knowing why.

Its simple: if the old said 2600
then its 4160
(3 bar would be 3120)

If 2500, then its 4000. (some ecu calib map at lower max v)
(3 bar would be 3000)
200 stays 200.

Click to expand...

You assume exactly 5.00000 volts for the reference volts supplied by the ecu to the sensor at all times, all conditions, hot, cold, 10.8 battery volts ot 16 volts at the battery. In real life this is not the case. The tolerance for the reference volts varies in the real world and the Bosch engineers knew this. According to the Bosch data sheets, depending on which sensor you choose, the design Vref tolerance varies from something like 4.75 volts to 5.25 volts (or so). In some models it's 4.88 volts Vmin or something else for other combination of ecu/sensor. These Vmin are reflected in the linearization "map" (a stretch to call a 2-point set of dat a a "map" , I know, I know) in the ecu. So you see, the upper measured volts is not arbitrary but intentional based on the sensor and ecu Vref tolerance designed into the ecu's physical circuitry.

If you assume 5.00000 volts Vref at all times in your remap, then you run the risk of saturating the sensor at the upper end of its measuring range, or even having the ECU always measure lower/higher pressure than what's actually there (introducing inaccuracies IOW). Some tuners -you may be surprised to learn who- do this intentionally - for example to avoid limp mode from overboost when the boost valve can't quite keep up (not here though) or else to fool VCDS during data logging or who knows why exactly- but the bottom line, this practice is wrong (IMO).

There are very good reasons some of these linearizations are like they are in the various ecus and sensor combinations and more consideration goes into these things than you may see on the surface.

I suppose if you can find a way to guarantee 5.250000 volts at the sensor at all times under all conditions then go for it!

I didn't wamt to say this but I was tuned in the past and always with key off my actual was reading 20-30 mbar low compared to atm. They put 3000 even though stock for my car is 2600 (for 2.5 bar). I was told it was my map sensor.

Simple, I checked the bosch map tech specs. Saw why it was 2600 and now 3120 (for 3 bar). Now actual matches atm.

Thats why I said to check what the tune has now. Some TDIs say 2500...but the axis for %volts goes to a lower number.

So knowing that and going by what the ecu HAD for the previous sensor, where does the feed voltage come into play? Also it would be stupid to run a 22 psi (1.5bar) tune with a 2.5 bar map, but I was told by a tuner that it works (probably what you mean w skipping limp, no leeway for higher readings to trigger overboost limp).

All bosch map sensors have the same curve and go to the same %voltage (4.6v?) on the spec sheet.
There's 2 sections, one is a different type, but in the ones we normally use, they share the same curves, just different range.

So thats why there is no issue with those calculations.
Also battery voltage doesn't change the voltage output of a 5v output. That would mean by running a 14.5v alt compared to a 13.8 bosch, readings would be off...which is not the case. The voltage regulator circuit can keep 5v output until much lower even. Same tech in a dc to dc converter, like your cellphone charger...put 6v to it and you still get 5v usb power out...with a bit less current.

I only want you to consider that MAYBE there's a solid reason or two why the stock linearizations only go to 4.6 or 4.8 volts or whatever, and it really isn't arbitrary and "wrong"

I don't agree with extending the linearization to match a full 5v, because (among other reasons) in a proper tune you shouldn't be using that upper part of the sensor range anyway.

But when it comes right down to it, it's really not all that important, so do what you think is best, and most importantly have fun.

I never said it was arbitrary, I didn't even explain how to change the axis, so I am confused.

I was saying how I was using the EXISTING table , to change values to different maps, like what the OP and other reply asked for. That's how they do it, unless they want to mess with axis.

I'm not changing the voltage at which it reads to. In fact, going by what is already there, the y axis shows 82 (.082%=0.41v)
and 989 (4.9v) ON STOCK ALH/ASV/ATD ecu.

It's not extending it to 5v. It's changing what is already there from 2.5 to 3 bar. From 2600 stock to 3120. I also mentioned how some maps have 2500, that should be set to 3000.

On ASZ map for example, goes to .951, which is 4.76v.

So you are saying that we should change the axis to .951, for 3000, instead of using .989 for 3120.
I understand that, but if the 2.5 and 3 bar maps on the bosch spec sheet have the same specs and same curve, where is the problem to leave it what the ALH/ASZ/ATD maps use?

I do agree that tuning should be well within the max of the sensor. A 3 bar map should not be tuned to use 3 bar.
I was annoyed when I was told by another tuner in the past that a 2.5 bar map can be used for a 22 psi tune.. Without that safety margin, like you mentioned, there is very little control and possibly never go into limp.

Edit: LOL another funny thing in the past regarding limits.... I was getting WFC (qa adjuster exceeded) limp mode at high rpm after I swapped to a 11mm pump. Setting adaptation group 01 to 32710 and below fixed it.
I was told it must be my pump so I dealt with it. As soon as I got my tuning cable, I saw why. The top of the pump voltage map was calling for 5 volts!?! And I thought my pump was going bad... sheesh. Why even go that high???

I think you are reading too much into my remarks maybe.

Also I kind of took things a bit out of context, my bad. (but not too much)

fact is, sensors , gages, etc are *never* spec-ed at 0-100% of full scale, because it's simply not possible to be linear over 0-100% of full scale. But that's kind of besides the point.

My point is:

Bosch said (when they calibrated all the ecus that use a 051C map sensor) that 3000 kPa is valid at 95.1% of full scale, and this is the setting in every factory ecu I've seen that uses what is commonly known as "the" 3-bar map sensor.

Bosch likewise said that "the" 2.5 bar sensor can go top 98.9% of full scale and still be accurate at 2600 kPa. This is what you see in ALH, etc.

Strangely enough, these values from the ecus exactly match what's specified in the Bosch data sheets as the upper bounds of the measuring range for the respective sensors. Amazing concidence, eh?

Extrapolating on paper from 95.1% to the 98.9% may work on paper but I have no data to assure me that this will actually remain valid in practice.

Therefore, to my mind, the right way to use a 3bar sensor in an ecu that originally came with 2.5 bar, is to just change that single byte in the axis from 989 to 951, and set the upper scale to 3000 instead of 2600, because this is what Bosch does. It's not really hard to change one more byte.

All things considered it's not really that important either way. Best practices won't use the full available range of the sensor anyway, but lord knows we've all challenged this a tad from time to time for convenience sake.

Now, back to debating how many angels fit on the head of a pin...

Thanks for the clarification.

From the bosch sheets I have seen,
3 bar 200-3000 with p2 at 4.65v
2.5 bar 200-2500 with p2 at 4.65v.
Both with characteristic curve "1"

So, there was no indication that they would have differences, besides the clue of the TDI ecus that use 3bar at .951 aka 4.74v...confusing as bosch uses 4.65v for p2.
Do you know why arent they using a lower voltage?

Angels don't obey space/ time, so the answer is undefined

So I have a question now.. (My ALH with stock VNT15...)

I'm self tuned to 17psi sustained (I'm sustaining 2170mbar absolute), and getting some brief spikes in 5th gear (lower ratio) at 2500-3000rpm on WOT to 2570mbar, TOP LIMIT of my 2.5bar MAP Sensor.

The brief spikes (roughly 21-22psi) come down to sustained 17psi... however, every once in a while, I get sent to limp mode for MAP Sensor overvoltage, meaning I hit the peak of the sensor...


Should I?:
1. Change to 3bar MAP Sensor + edit MAP Linearization at address 0749F0 to:
951 3000
82 200

or

2. Further raise my N75 duty cycle at the higher fuel IQ and 2500-3000 RPM? To prevent such spikes?

Simplest way is to lengthen your vnt actuator..
Or find the n75 max table and decrease it. Not at home so I can't show you where it is.

The n75 table you speak of doesnt handle transient spikes but is a guide for steady state load.

robnitro said:

Simplest way is to lengthen your vnt actuator..
Or find the n75 max table and decrease it. Not at home so I can't show you where it is.

The n75 table you speak of doesnt handle transient spikes but is a guide for steady state load.

Click to expand...

Hmm very interesting. Thanks for letting me know. I'll try researching some more about it. When you do have time, I'd love to learn about those other n75 maps.

Possibly until then I will get a 3bar MAP to prevent limp mode from that? Any real downsides? I realize that the 2.5bar MAP is enough for my sustained pressures.

Theres also a setting for time above pressure when limp hits. what is your svbl set to btw?

try the rod adjustment.. if u can check where it cracks n hits the stop.. should be 4-6 crack, 18-22in hg stop. the second number is more critical

robnitro said:

Theres also a setting for time above pressure when limp hits. what is your svbl set to btw?

try the rod adjustment.. if u can check where it cracks n hits the stop.. should be 4-6 crack, 18-22in hg stop. the second number is more critical

Click to expand...

SVBL is at 2250mbar, target boost at high fueling (50mg) is set at 2182mbar (gives me actual measured 2172mbar sustained)

Ahh yeah, the 3 bar map definately wouldn't help.
Your map can do up to 2500 mbar, and you aren't even close to that!

you might want to set the svbl a bit higher...
2172 is 17.23 psi (at sea level)

2250 is a bit too close... 18.38 psi.

In most stock ecu's and my previous tune I saw around 150 mbar higher svbl than max, which gives room for the boost control (and not hit limp quickly).
If you use 150mbar, use 2332 mbar, which is 19.58 psi...
Or conservative 100mbar would be 2282 mbar, which is around 18.85 psi.

2 types of limp, one is exceeding svbl, and the other which is less strict, exceeding request boost for x time. You might be tripping svbl too easily.

But the actuator adjustment is really critical. Think of lengthening the rod as shifting ALL the n75 controls to be less "extreme".

For the n75 max vane adjustment by PID (control loop that handles spool up), search inside your maps using vagedcsuite a value of 7500 or 5500 (I see in 03 GN ecu address 5645c- but yours may be different). It's a 2d table, axis rpm, and values are the raw value of percent n75 (eg 7500-> 75%, which is actually 25% in vagcom/vcds)

You can tweak it a bit, in your case of spikes at 2500-3000, lower those values to 7000 lets say, and also 3250 and 3500 to 7500, as you want it to smoothly taper down.

It's how I handled some areas of spikes on my tune (after double checking my actuator was set ok!)... I learned that the normal n75 map didn't do much at all.

robnitro said:

Ahh yeah, the 3 bar map definately wouldn't help.
Your map can do up to 2500 mbar, and you aren't even close to that!

you might want to set the svbl a bit higher...
2172 is 17.23 psi (at sea level)

2250 is a bit too close... 18.38 psi.

In most stock ecu's and my previous tune I saw around 150 mbar higher svbl than max, which gives room for the boost control (and not hit limp quickly).
If you use 150mbar, use 2332 mbar, which is 19.58 psi...
Or conservative 100mbar would be 2282 mbar, which is around 18.85 psi.

2 types of limp, one is exceeding svbl, and the other which is less strict, exceeding request boost for x time. You might be tripping svbl too easily.

But the actuator adjustment is really critical. Think of lengthening the rod as shifting ALL the n75 controls to be less "extreme".

For the n75 max vane adjustment by PID (control loop that handles spool up), search inside your maps using vagedcsuite a value of 7500 or 5500 (I see in 03 GN ecu address 5645c- but yours may be different). It's a 2d table, axis rpm, and values are the raw value of percent n75 (eg 7500-> 75%, which is actually 25% in vagcom/vcds)

You can tweak it a bit, in your case of spikes at 2500-3000, lower those values to 7000 lets say, and also 3250 and 3500 to 7500, as you want it to smoothly taper down.

It's how I handled some areas of spikes on my tune (after double checking my actuator was set ok!)... I learned that the normal n75 map didn't do much at all.

Click to expand...

This is gold. Thank you!

Also forgot to mention that I have catalytic delete, so my turbo theoretically will spool up faster; I will go to 7250 at 2250rpm, and 7000 from 2500rpm to 3500rpm. I'll set SVBL at 2300, and I'll do testing. Will check actuator as well, to the psi recommendations above.

I use VagEDC15 Suite with WinOLS Demo for in depth (mine is a 03 GN as well). I've tweaked and logged for about a month, learned a ton, and had a couple of overseas tuners check the bin for recommendations.

Can I PM you if I have more questions, compare my file, etc.?

Please please do check the actuator. Being with less restriction I would aim for hitting the stop at around 22 in Hg, which is what I use on VNT17, with a high flow cat.

Check my other posts here, a bunch of us put up how to do a few interesting things. Also on ecuconnections.com with the same name. I forget the tdiclub thread name right now.

You can PM me.

robnitro said:

Please please do check the actuator. Being with less restriction I would aim for hitting the stop at around 22 in Hg, which is what I use on VNT17, with a high flow cat.

Check my other posts here, a bunch of us put up how to do a few interesting things. Also on ecuconnections.com with the same name. I forget the tdiclub thread name right now.

You can PM me.

Click to expand...

Did a quick flash with 7000 as the data, also 2300mbar SVBL.

Car "feels" more sluggish in terms of immediate response, might be in my head.

However I'm still reaching 2570mbar actual manifold pressure in 4th gear spikes, then drops down (N75 charge pressure-duty cycle in VCDS goes to 93% to do so).
Oddly enough the sustained pressure isn't as well controlled now. For instance when I would before hold a steady 2172mbar, now it's usually higher somewhere in the 2200's, or lower in 2130mbar.
No limp mode though. It may all be from the higher SVBL.

I'll flash it with stock 7500 on the N75 map, and do more logging, bringing it down by 100 (1%) at a time.


Still didn't have time for actuator modding as I'm preparing another car of mine for Autocross this weekend. Although, I think can replicate a rod lengthening by moving down the whole N75 duty cycle map values, no? The actuator behaves as it should.

Adjusting the rod is better because you give more spring force which is used to open vanes.

2200 is not dangerous for vnt15... most tuners use 2250 or so for vnt15 request (18 psi).

The vnt 15 compressor map allows 2.3 bar in mid range btw... taper down boost past 4k... I posted my calculations somewhere in the "help me understand my ecu" thread.

robnitro said:

Adjusting the rod is better because you give more spring force which is used to open vanes.

2200 is not dangerous for vnt15... most tuners use 2250 or so for vnt15 request (18 psi).

The vnt 15 compressor map allows 2.3 bar in mid range btw... taper down boost past 4k... I posted my calculations somewhere in the "help me understand my ecu" thread.

Click to expand...

Ok, wonderful. Today I did a lot of logging, after flashing again.

I think I've come up onto something.

Sometimes the boost requested and actual boost follow each other PERFECTLY. I mean it doesn't even overboost very much. Tiny spike (40mbar), then drops down and raises to hold 2172mbar right next to requested. Maybe TOO well controlled.

AND SOMETIMES (at random) it spikes hard and holds actual boost higher than requested (ex: requested 2182, actual 2280-2300, after large spike) and doesn't keep to the line very well on the graph. The reported N75 duty cycle goes up to 93% to try to correct it, but never manages to bring it down.

Seems like something is acting up sometimes.

I'm going to check my N75 valve first. Is there a way to clean it? Does it get dirty?

Enabled said:

Ok, wonderful. Today I did a lot of logging, after flashing again.

I think I've come up onto something.

Sometimes the boost requested and actual boost follow each other PERFECTLY. I mean it doesn't even overboost very much. Tiny spike (40mbar), then drops down and raises to hold 2172mbar right next to requested. Maybe TOO well controlled.

AND SOMETIMES (at random) it spikes hard and holds actual boost higher than requested (ex: requested 2182, actual 2280-2300, after large spike) and doesn't keep to the line very well on the graph. The reported N75 duty cycle goes up to 93% to try to correct it, but never manages to bring it down.

Seems like something is acting up sometimes.

I'm going to check my N75 valve first. Is there a way to clean it? Does it get dirty?

Click to expand...

Not to say that you don't have an issue with the N75/actuator/vanes, but this ECU only has one N75 map per coding section. Later cars sometimes had two, or were even gear dependent (5+ N75 maps). If you want the same boost pressure in varying load conditions (as in different gears) the duty cycle may need to be different. So, there is always going to be some variation.

vwmikel said:

Not to say that you don't have an issue with the N75/actuator/vanes, but this ECU only has one N75 map per coding section. Later cars sometimes had two, or were even gear dependent (5+ N75 maps). If you want the same boost pressure in varying load conditions (as in different gears) the duty cycle may need to be different. So, there is always going to be some variation.

Click to expand...

Yes, I understand what you mean. But that variance was happening at the same RPM, and same gear, just different times. Either way, my vanes have been cleaned, actuator works well, and now I just cleaned my N75 valve and lubricated its insides, moves very smoothly and it works. Will log tomorrow.