VNT15 use experience ?
Turbo Mini Forums

What are your experiences in use of VNT turbos on our petrol engines ?

Most of VNT are for diesels.
I've used VNT15 on my 998+ CR 9.3:1. The turbo comes from a Golf/Passat 110hp 1.9TD (ref AFM) engine. My conclusion is that it requires an electronic control, not just a boost valve on wastegate.

When VNT nozzles are closed the turbo spins from 1500 RPM (despite of a 2 times smaller engine) and at around 2500 RPM is in surge (very audible). When adjusted the boost comes later (at 2500-3000) but it's still hard to limit at 5000 RPM, it's quickly in surge again because a simple pneumatic wastegate control does not open nozzles enough.


PS: to help people searching for a quick answer:
I managed to control the turbo using an RC servo. However 1.9L diesel VNT15 is not properly sized for a 1L petrol engine. The turbine is too small, it makes spin a 1.9L sized compressor too fast, so the surge is easily reached.

Edited by alpa on 27th May, 2008.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

alpa , could you create a linkage to make the movement of the wastegate magnified to a ratio that would work ,perhaps non linear ? this is what i was thinking of doing .

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

The law is unpredictable because it depends on RPM, pressure and the engine. The VNT15 came with a wastegate actuated by DEPRESSURE, which means they use the diesel brake assistance pump to control the wastegate.

Edited by alpa on 5th May, 2008.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

I'm still fighting against the boost LOL.
I've modified a standard wastegate into an "adjustable", I can change the spring inside. No way, still blowing too much air. I can only choose between a very low boost (105kpa) everywhere or a higher boost which grows as RPM goes up.

I think the turbine must be controlled dynamically. Close nozzles to start spinning, open them as soon as the boost is here.
On most of diesel engines there is no throttle, they can use the MAP to know the true compressor pressure. So they can electronically control the true boost level.

I wonder how to use an RC servo.

Edited by alpa on 13th May, 2008.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

Hi,
On the VWs they use a 3 port solenoid valve to bleed off the signal. They also have a MAP/MAT sensor in the pipework after the turbo. An RC servo would do, they need a PWM signal on the white wire. 1ms pulses and 1520/1600 us neutral point.
Cheers,
Gavin :)

VEMs Authorised Installer / Re-seller. K head kits now available!

WB/EGT gauges. Click here for customers write-up

Visit www.doyouneedabrain.co.uk

My Mini build diary

The 3 port solenoid (valve) creates a leak, so it makes boost go up. The problem with the very low boost is to get an even lower boost and then to make it grow.

With VNT I have to control the turbo even when the engine is throttled, otherwise the boost is here even at 70kpa @ 3000 (I can here the surge mode) and it's obviously not enough to activate the wastegate.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

Alpa,

What actuator are you using?

does it have pressure and vacum connectionslike this..

http://www.thedodgegarage.com/turbo_vnt.html

Very interesting !
I've already ordered a high torque Hitec servo but will try a dual connection system.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

So I modified the wastegate to have a vacuum connection and installed the boost 3 way valve on the vaccum side. This actually eliminates light load surge. I did not have time to map boost control, for the moment as soon as boost comes the vacuum side is open to the ambient pressure.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

I'm definitely unable to control it with a wastegate. The 2 connection wastegate is a very clever solution but not enough for me.
This is certainly due to a badly sized turbine for a petrol application. I'll see if a servo will do better once I'd received it.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

Have you tried disconecting the vanes an positioning them wide open?

try this and see if it limits boost, if not then IMO the turbine is to small.

I have a gt25v on mine and have added a conventional wategate also incase boost does not limit with the vanes.

I want boost everywhere. Right now boost comes at 1500 RPM and reaches the requested 110kpa level at 2500. Surge appears when the throttle is suddenly open, the boost goes at 115-120 before the wastegate reacts and turbo slows down.

If I statically open nozzles then boost comes at 3000 or latter and keeps on raising until compressor enters surge at some RPM, not what I want. But of course this works, this is what I did in the very beginning.

An independent electric control should allow me setting nozzle at required position before the boost comes, which would prevent these boost explosions.

The goal is to run at 120kpa. When N/A with Maniflow this engine was running 95kpa with ambient at 91.

gt25 is much bigger than gt15. And I'm not looking for high power, torquey 60HP would be enough.

If I had enough place I would added an external wastegate. Right now the engine looks like ordinary N/A MPi, except the air hoses strangily going under the intake before entering the throttle body.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

I want boost everywhere too!!

so, with the vanes competely open the turbo still surges?

it sound to me like the actuator is just not reacting to boost quick enough when the throttle is opened wide.

I did see a setup on youtube which also had a connection to the throttle cable that appeared to partially open the vanes, this might help prevent the boost spikes.

I did not try to completely open nozzle, with a half-way opening the boost comes at around 3000. It's still a 110HP 1.9L turbo, not a ridiculous thing.

I actually weakened the wastegate spring at max, loser spring would not fight against gaz pressure.

Cable system is probably more complicated to do and harder to tune than a servo.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

I installed the servo, wrote the control code (40 minutes). And spent hours trying to understand why the servo had convulsions with running engine.
I checked for bugs, dirty power line or control signal.
Actually the servo does not like at all being near a running engine (probably ignition, mine is static). Even without control signal, just powered, it's moving all the time unless I put it 1.5m away the engine. Going to shield and we'll see.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

Thats very interesting!

interferance is somthing I had not considered with a stepper,

I assume the convulsions stop without the engine running.

do factory fitted idle steppers have screened wire? i didnt think they did.

The problem isn't in the stepper but in the electronics. Servo contains a circuit. Even if I power servo from an external supply and don't connect the input (idle mode) it moves as soon as it's near the running engine.
I tested with a shielded servo and it's the same. Which proves the noise comes from the wires and not directly into the circuit.
There are 2 solutions:
- try to wire the control signal and power in a shield and far from ignition, filter the power (preferred solution I'll try now)
- install the servo electronics in the car (put apart the servo electronics).

Edited by alpa on 22nd May, 2008.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

Hi,
I've heard of this happening on RC cars too, never had a problem with it myself, the servos are only 150mm away from the plug lead.
Cheers,
Gavin :)

VEMs Authorised Installer / Re-seller. K head kits now available!

WB/EGT gauges. Click here for customers write-up

Visit www.doyouneedabrain.co.uk

My Mini build diary

what about openeing the servo and loadign up the input signal with as low a wattage ressitor as your drive circuit can cope with and a small suitable capacitor to sink any spikes to ground. sheild the unit too and i cant imagine any issues.

a well regualted power supply will probably also be necesary! either that or look into stepper controll and soem kind of potentiometer. bit more involved, but less succeptable to interference. very interesting stuff your up to :)

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

Oh ok that makes sense, when I do this i will be using (or tying to use) a 4wire idle stepper motor so should nit have the same issues hopfully.

Edited by alpa on 23rd May, 2008.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

This may be of interest-

http://www.phenostream.com/Products/OSM_01.aspx

Thanks. Interesting site. I knew OpenServo.

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm

stainless helical ht leads google 'msd' ?

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

std 998 A+, g295, MD266, RHF4, 109hp @0.8bar/5400rpm