Author Topic: Can VEMS control a variable geometry turbo?  (Read 18679 times)

Offline BimmerDread

  • Jr. Member
  • **
  • Posts: 17
  • BHP: 2
Can VEMS control a variable geometry turbo?
« on: December 05, 2010, 04:30:00 pm »
Has anyone messed about with using VEMS output to control a variable geometry turbo such as the Holset 351VGT? These vary the turbine housing volume from 3mm2 to 25mm2.

I am looking at using this set up to achieve fast spool using TPS and RPM to activate it on an BMW s38b36 engine. My idea is that for:

TPS<50% all RPM range to set the housing to 25mm2 which I have calculated to to spool 1 bar from ~4500rpm.

TPS >50, RPM 1000 to 2500 to set the housing to say 10mm2 for spool at 1800 RPM,

TPS > 50%, RPM 2501 to 4500 housing 18mm2

TPS > 50%, RPM 4501 to Redline (Set to 7000) housing at 25mm2

The concept is to control spool. The built in boost control will monitor boost. 

People have been messing about with controlling this turbo. See here http://www.homemadeturbo.com/f7/holset-vgt-he351ve-controller-117261/
It is a long read but with good info.

Offline GintsK

  • Hero Member
  • *****
  • Posts: 1257
  • BHP: 50
Re: Can VEMS control a variable geometry turbo?
« Reply #1 on: December 05, 2010, 05:51:49 pm »
PWM - yes
CAN - no.

Offline gunni

  • Hero Member
  • *****
  • Posts: 1492
  • BHP: 37
Re: Can VEMS control a variable geometry turbo?
« Reply #2 on: December 05, 2010, 06:11:29 pm »
I thought about this 2 years ago and at that time was going to use a mechanical linkage to the turbo to control the movement.
But with a pressure actuated one a boost solenoid would work the same.

You can use the alternative pwm table to do this based on manifold pressure against rpm.

Which should provide you with the control ability . If needed then you should be able to get a TPS/RPM pwm table from the developers.

Offline BimmerDread

  • Jr. Member
  • **
  • Posts: 17
  • BHP: 2
Re: Can VEMS control a variable geometry turbo?
« Reply #3 on: December 05, 2010, 07:38:45 pm »
Its CAN :(

So I think I have 3 options:
1) A PWM solenoid and lever and not the stock actuator.
2) A on/off solenoid activated on TPS and RPM
3) The mechanical route (Boost activated).

Thanks guys.

Offline GintsK

  • Hero Member
  • *****
  • Posts: 1257
  • BHP: 50
Re: Can VEMS control a variable geometry turbo?
« Reply #4 on: December 05, 2010, 08:14:23 pm »
Some drives for VNT was just PWM. May be it is possible to find some..

Offline gunni

  • Hero Member
  • *****
  • Posts: 1492
  • BHP: 37
Re: Can VEMS control a variable geometry turbo?
« Reply #5 on: December 05, 2010, 10:33:59 pm »
Its CAN :(

So I think I have 3 options:
1) A PWM solenoid and lever and not the stock actuator.
2) A on/off solenoid activated on TPS and RPM
3) The mechanical route (Boost activated).

Thanks guys.

What is the actuation process ? is it pressure based or mechanical linkage within the adjustment unit?

Offline BimmerDread

  • Jr. Member
  • **
  • Posts: 17
  • BHP: 2
Re: Can VEMS control a variable geometry turbo?
« Reply #6 on: December 06, 2010, 01:53:27 am »
The turbo uses a mechanical linkage operated by a motor.
« Last Edit: December 06, 2010, 01:57:00 am by BimmerDread »

Offline gunni

  • Hero Member
  • *****
  • Posts: 1492
  • BHP: 37
Re: Can VEMS control a variable geometry turbo?
« Reply #7 on: December 06, 2010, 02:27:09 am »
then just replace the motor?

my idea was to use a servo motor to handle the mechanical movement.

Offline GintsK

  • Hero Member
  • *****
  • Posts: 1257
  • BHP: 50
Re: Can VEMS control a variable geometry turbo?
« Reply #8 on: December 06, 2010, 11:17:02 am »
Servomotor from what?
There is harsh environment!

Offline andreNL

  • Full Member
  • ***
  • Posts: 122
  • BHP: 8
    • My Wiki membersPage
Re: Can VEMS control a variable geometry turbo?
« Reply #9 on: December 06, 2010, 11:35:46 am »
Maybe the steppermotor driver can be used

Offline gunni

  • Hero Member
  • *****
  • Posts: 1492
  • BHP: 37
Re: Can VEMS control a variable geometry turbo?
« Reply #10 on: December 06, 2010, 12:27:59 pm »
Servomotor from what?
There is harsh environment!

the servo doesn´t have to be directly close to the turbo, it can be mounted elsewhere and rotate a arm.

Offline GintsK

  • Hero Member
  • *****
  • Posts: 1257
  • BHP: 50
Re: Can VEMS control a variable geometry turbo?
« Reply #11 on: December 06, 2010, 12:56:04 pm »
Ok. What servo?
===
Stepper isn't best idea. It require firmware modificatons. Servo usually use pulse width. So PWM with right frequency could be used. VEMS can control VNT + control W-gate using two independent PWMs.
SI engine require w-gate too.

Offline gunni

  • Hero Member
  • *****
  • Posts: 1492
  • BHP: 37
Re: Can VEMS control a variable geometry turbo?
« Reply #12 on: December 06, 2010, 03:27:58 pm »
I can´t remember which, but it has to have a bit of torque available

Offline BimmerDread

  • Jr. Member
  • **
  • Posts: 17
  • BHP: 2
Re: Can VEMS control a variable geometry turbo?
« Reply #13 on: December 08, 2010, 12:36:57 am »
So... after some thought I decided to modify my parameters. This is what I came up with:
At 12mm2 spool comes on at 2000 rpm and back pressure effect begins at 4100 rpm.
At 23mm2 spool comes on at 4000 rpm and back pressure effect begins at 8000 rpm.

I want to have a two state (on/off) solenoid that when off keeps the volume at 23mm2 and at
12mm2 when on.

A function table would look like this:

TPS            RPM           State      Volume
<50%         Varies         Off          23mm2   
>50%         0 - 4k         On          12mm2
>50%         4k - 8k        Off          23mm2
 
The waste gate would manage any boost ripples on transition.

I needed a mechanical system that would have pressure at all rpms so I am opting for an oil pressure based system (like vanos ;))

I want to feed the oil to the waste gate port that would operate the system.

What do you think?

Offline gunni

  • Hero Member
  • *****
  • Posts: 1492
  • BHP: 37
Re: Can VEMS control a variable geometry turbo?
« Reply #14 on: December 08, 2010, 01:53:01 pm »
Most wastegates would leak.