View Full Version : E46 320d 150hp - new turbo, decat & other mods, tune needs adjustment?

10th October, 2016, 10:05 PM
Hello, clever people of Digital Kaos! :)

I have messed with this car for about 2 years now and it is almost perfect, finally.

It's an e46 320d 150hp 5-speed manual. EGR completely removed, both cats empty, new Garrett gt1749v turbo, swirl flaps deleted etc. + a remap.
After fitting the new turbo, removing swirl flaps and decatting, the boost comes in a bit later (~2100rpm instead of 1800rpm) and it has a small spike which is overcorrected by the N75.
I have the necessary equipment to run tests and make turbo logs etc., and I have made a few runs today - here's one of them:

3rd gear 1500rpm, full throttle:

As you can see the N75 is over-reactive (there's a boost spike up to 2,65bar, then a loss of boost, then another spike) and boost maxes out a bit too late for a small turbo like this...

So the question is, is it possible to correct these issues with tuning? e.g. have the N75 hold the geometry shut a little more until 1800rpm and then let it open a bit faster, but not too fast like in the graph? Or is this a problem with the hardware (turbo etc)?

I will attach two logs in .csv format, both in 3rd gear. If you need any more logs etc. to help me, I'll gladly go make them. Maybe together we can make this car perfect :)
Thanks a ton!

11th October, 2016, 10:50 AM
Here's the graph in a more readable state - didn't know the forum reduced them so much:


11th October, 2016, 11:27 AM
Correction will be always possible, but N75 configuration is last thing.
First needed is electro mechanical parts, there will be always little delay and spikes. But your turbo overboost to much and remapping will not help, only can decrease and change PID's but that's no solution..
N75 is new, and original now, vacuum is correct ?

11th October, 2016, 11:42 AM
N75 is new and also tried changing with the one on EGR (they are identical on the e46). All vacuum lines, vac pump and reservoir freshly changed. Also the piping and gaskets on the boost piping are new so it's completely leak free.
The turbo is basically original, with a CHRA from Melett and the housings from a crashed car, built to spec.

Without the remap, it was a bit better and didn't really overboost, the spike occurs either when I shorten the actuator or put on the remap.

Here's a graph without tune, notice it spikes very minimally but is also a bit laggy:
http://i442.photobucket.com/albums/qq143/martinlink007/BMW/160mbar%20amp%209%20turns%20III%20gear_zpsbsojuhtf .png

Thanks for the input :)

11th October, 2016, 06:33 PM
How much is IQ when the spike occurs?
I mean, maybe is too much... and the exhaust gases are too much for original position of the vanes...
You must calibrate the N75 map in sectors where the spike occurs...
Don't play with actuator settings, leave it at original position.
That's my opinion.

11th October, 2016, 08:41 PM
The actuator is almost exactly at stock position, maybe 0,5 turns longer as it was overboosting to 2730mbar earlier. I can set it back to where it was as soon as the reason for the overboost is found and fixed.
What's bothering me a bit more is the boost lag - I'd love to have full boost at at least 1900rpm, if not earlier.

7th April, 2017, 08:39 PM
So a lot of time has passed and I've had very little time to deal with the problem. Car drives well apart from a very brief flat spot where the boost (and fueling) is reduced at 2000 rpm.
The overboost issue after 2500 rpm was fixed only by N75 % lowering about 200 rpm before it happened and as you can see, boost control is better than perfect :)


What's still not right is the loss of boost and fueling, though it's better than it was.
The N75 is getting freaked out by something (not overboost since it barely crosses desired!) and also fuel is reduced. Looks like PID kicking in too hard and without any real reason.
What would be the next step to get rid of the dip, and maybe get it to boost even earlier/faster?

I was thinking about shortening the actuator until it spikes+overboosts and stays above desired constantly, and then reduce N75 before the spike. Does this sound like it's accomplishable? Or anyone have any other ideas how to get it to boost faster and smoother?

I've attached the tuned and modified file along with the .csv log which now includes fuel request and actual.

Thanks for all the help!