Wanted - Clever Folk

[SiWhite]

Evening all.....

I'm almost finished installing my variable nozzle turbo onto my 200Tdi.

For those who don't know much about them, they vary the angle that exhaust gasses strike the impeller, meaning at low exhaust flow (low revs) is forced through small apertures to spin the impeller quickly. Then, at higher gas flow (revs) the vanes open, reducing back pressure and spinning the impeller faster.

One area of concern in my installation is actuating the variable nozzle. It's currently set to alter with boost pressure, which was easy with the old turbo's diaphram. This seemed nice at first - at low boost the vanes will be close together, spinning the impeller quickly and generating good low end torque. The problem this causes is at fast cruising speeds with lowish boost, the engine will be very 'boosty' and fast to react to throttle input. I could overcome this with a 'cruise lever', forcing the vanes to stay partially open at speed, but it's a pain.

I'd like to have the vanes actuated dependant on engine speed. This would give all the low end torque benifets I wanted when fitting the turbo, bit without the spiky nature of the boost dependant system.

The question is - how do I convert the changing engine speed to about an inch of movement? I had some initial thoughts about using a scrap tacho I have in conjunction with a radio controlled car type servo, but would struggle with interfacing them. I'd also thought about having a speedo cable driving a steam-engine type govenor (two spinning balls affair) driving a cable, but I'd not really want the moving parts nature of this. I'd also thought about using engine oil pressure driving a small ram, but this would alter with engine temperature on start-up.

My question is - what would you do? Am I over worried about the boost-dependant system (bearing in mind it's not running yet) or should I work on an engine speed system? And if the boost's no good, how should I do the engine spped system for best reliability?

[Gromit]

The question is - how do I convert the changing engine speed to about an inch of movement? I had some initial thoughts about using a scrap tacho I have in conjunction with a radio controlled car type servo, but would struggle with interfacing them.

This was the first thing that came to mind.

Servos are driven by a pulse stream from the receiver, but I reckon it would be possible to make up a controller that took input from a potentiometer and drove your servo to a corresponding position. You could then drive the pot from the tacho.

There's lots of info on the web on controlling servos, from the robotics folks as well as R/C.

I've a bunch of old servos here, maybe even receiver. Free to you if you want them.

[Gromit]

here's a servo driver. 18 bucks.

[FridgeFreezer]

I would use the 2nd injector channel on my ECU to control a PWM actuator allowing me to create a boost map dependent on RPM and load. Either that or use the inbuilt boost control.

But since you have a silly diseasal engine such technology is too good for you

I would use a 555 timer set up as a monostable and connected to a filtered tacho feed (EG alternator W terminal) so that as RPM increases the duty cycle of the output increases. Again, a PWM actuator such as butchering a PWM idle valve would be one way of doing it.

You could do it any number of ways really, depends what you want to use as an actuator.

[RRX]

I would stick with the normal VNT actuator as at least the turbo is controlling itself, no silly electricery to get wet like petrols have, as for having it engine speed based surely this removes the point of a VNT with the basic idea being that they are able to spool up alot quicker at low revs and when you put your foot down and the vanes need to be able to move to control the boost levels which means sometimes opening them fully to slow the exhaust turbine down which would take a very complex controller

[Turbocharger]

I'd be inclined to leave it boost-controlled, otherwise there could be low-speed high-boost situations where the turbo surges - it's too early in the experiment to blow your turbo up!

I imagine the 'correct' answer is to create something which uses both variables (Megasquirt and a map?) but it seems a lot of effort for something that already works reasonably well.

[FridgeFreezer]

Ideal application for a MicroSquirt methinks.

[dollythelw]

Simple of simple answers - longer travel accelerator pedal a throttle damper (Pug 405 has a nice one), is there a linkage on the actuator Si?

Technofest answer; a whole heap of electronics to give a map driving a servo but thats going to need engine load to be taken into consideration hmm

[will_warne]

Si, FYI mine always ran reasonably high boost while crusing (0.8 - 1.0 bar) so I'd stick with what you've got and maybe try a damper as Jez suggested.

If you want to try something else I'd link the boost to throttle rather than to revs. Think about when you want the boost to come on - high throttle regardless of rpm.

Pop round and stick your head under the bonnet if you want to see what the factory did.

[Aragorn]

i cannot see why having fast throttle response at cruise is a bad thing?

surely knowing that if u open the thottle it goes immediately is better than planting your foot to the floor and getting nothing, Limiting boost by RPM would only make it worse to drive as RPM doesnt relate to road speed so you'd effectively neuter its performance in lower gears when u do actually want it...

whats the turbo from originally? you may find if the new engines significantly larger you'll need to adjust the VNT mechanism to stop boost spikes at wide throttle low RPM states

if the boost pressure isnt spiking and is staying within your chosen range i really dont see the problem

[SiWhite]

There's an off chance I've been trying to solve a problem that might not exist...

I initially chose boost to control the vanes as it was easy! I've been thinking about what would be the best parameter to actuate the vanes, and it's a fairly complex problem. I'm trying to achieve a situation where the engine has loads more low end torque with better throttle response - pulling out of junctions and pulling away into a hill climb when off road, for instance. I really don't mind what it does to performance above about 20mph.

My concerns with using boost to control the vanes are that I can imagine a situation where the engine revs and boost are both high (and therefore there's lots of exhaust going through the turbo) like when accelerating hard. Then, when coming off the throttle suddenly, there's still a lot of exhaust going through the turbo but boost drops quickly, closing the vanes and forcing the exhaust through smaller apertures. But will the VNT keep the boost high and the vanes open, preventing this further rise in boost?

I know this is what a VNT is designed to do, but I'm concerned about damaging the engine with excessive boost, especially as the VNT has no wastegate and thus no way of getting rid of excess boost.

Current thinking will have the VNT set fairly 'tamely', with a limit to the ammount the vanes can close. I'll plan to increase this as testing progresses, and I hope I'll find a setting where it doesn't go above about 18psi.

Does the collective genius of the forum techies think that a modified dump valve, set to open above 18psi, might do anything?

The turbo is a Garrett GT22V, a very common turbo taken from a Merc M Class Diesel. Merc acutate it using vaccuum, which is high at idle but drops as revs rise.

Jez, I did think about linking the vanes to the throttle, but there are plenty of situations (like towing up long hills) where the engine is working hard and planting the throttle would open the vanes, dropping boost. Or did you imagine the vanes being open at no throttle input and closing with pedal travel?

[will_warne]

There's an off chance I've been trying to solve a problem that might not exist...

I initially chose boost to control the vanes as it was easy! I've been thinking about what would be the best parameter to actuate the vanes, and it's a fairly complex problem. I'm trying to achieve a situation where the engine has loads more low end torque with better throttle response - pulling out of junctions and pulling away into a hill climb when off road, for instance. I really don't mind what it does to performance above about 20mph.

My concerns with using boost to control the vanes are that I can imagine a situation where the engine revs and boost are both high (and therefore there's lots of exhaust going through the turbo) like when accelerating hard. Then, when coming off the throttle suddenly, there's still a lot of exhaust going through the turbo but boost drops quickly, closing the vanes and forcing the exhaust through smaller apertures. But will the VNT keep the boost high and the vanes open, preventing this further rise in boost?

I know this is what a VNT is designed to do, but I'm concerned about damaging the engine with excessive boost, especially as the VNT has no wastegate and thus no way of getting rid of excess boost.

Current thinking will have the VNT set fairly 'tamely', with a limit to the ammount the vanes can close. I'll plan to increase this as testing progresses, and I hope I'll find a setting where it doesn't go above about 18psi.

Does the collective genius of the forum techies think that a modified dump valve, set to open above 18psi, might do anything?

The turbo is a Garrett GT22V, a very common turbo taken from a Merc M Class Diesel. Merc acutate it using vaccuum, which is high at idle but drops as revs rise.

Jez, I did think about linking the vanes to the throttle, but there are plenty of situations (like towing up long hills) where the engine is working hard and planting the throttle would open the vanes, dropping boost. Or did you imagine the vanes being open at no throttle input and closing with pedal travel?

Si, I've got to head out now but you're thinking was sound when I spoke to you a couple of weeks ago - dump valve pre actuator should stop boost going through the roof.

Best solution would be to get it running and play around. If it starts doing anything silly cut the fuel. As long as you don't have issues with surge then I wouldn't worry about boost getting too high in testing as the kill switch will stop you damaging anything.

[dollythelw]

I was thinking maybe an alternative method of dampening the spikey boost by altering the travel of the vane actuating crank slightly, (increasing ratio between travel and boost).

where are you "T"'ing into to get the boost control from?

[simonr]

In answer to the original question - I would make something like the governor on a lawn mower - an engine driven fan blows on a flap which is repelled against a spring. Flap position then proportional to engine speed.

I think a better solution might be to put a damper on the VGT actuator which would take out the spikes.

One way to achieve this is by putting a constriction in the tube that feeds the actuator.

OK it slows the throttle response - but it just means you have to get the orifice right so it smoothes out the spikes but still gives you decent response.

Si

[SiWhite]

Si, I like that idea A LOT

I got some little 6mm inline 'flow restrictors' from a kind forumeer some time ago.

How about I fit a little one-way valve to let pressure into the diaphram, and then a 'flow restrictor' on a tee piece after the one-way valve to slowly exhaust the pressure from the diaphram? That would ensure fast response to the throttle and a slow release (snigger...) when the throttle is released.

Are 5mm one way valves available?

Jez - I was planning to take boost from the inlet manifold, tee'd into the line running from there to the fuel pump. Sound sensible, or would taking it from the turbo be better?

[RedLineMike]

anychance o a piccy of it bolted to the engine?

did you have to retap the exhuast manifold to take the turbo?

[eliott]

How did this story ended? I've just fitted a turbo from AUDI 2.5tdi V6 , fit directly to 200tdi manifold, but need to solve the boost control.

[Aragorn]

You really want something like this:

http://dmn.kuulalaakeri.org/vnt-lda/