I didn't mean to say that there are not situations where the vanes wouldn't be all the way open and you are most likely correct in the condition you just used as an example. The majority of your fuel economy gain/loss however is going to come from transient throttle conditions in which case there is rarely vacuum and fully open vanes will not be the most efficient state for the turbine housing. After reading my post again I probably should have gone a little further with my explanation your on the right track.

Your comment about the EGT's is right on for these turbo's as well the older models were known for having issues at sustained high egt's however there are many newer models (Holset HE) for example that would be fine for 90% of gasoline applications. Diesel engines actually see much higher EGT's the difference is they are basically cold air pumps when they are off the throttle so the components are cooled more often so the heat build up is less of an issue. Your use will make a big difference as to whether EGT's will be an issue or not. Drag racing or normal street driving would probably be a non issue, road course or maybe even autocross could pose a problem if you aren't using the right turbo for your application and controlling it well (and watching your EGT's)

 

So who's gonna do it!?! We have some crazy people twin charging s2000's when this seems much more explorable in terms of efficiency.

 

twin charging unless its sequential (which i haven't seen yet) is nothing but a publicity stunt. VGT has real benefits. I definitely plan on doing it, but it will most likely be on my street car until I see if there is any top end power to be had (and how much) because the technology comes at a significant weight penalty compared to a standard turbo.

 

Weight? How much more could it possibly weigh?

 

I run a 4088R which weighs right at 30lbs, if I wanted a VGT turbo that made about the same power I would be looking at about 60lbs... so almost double on, a street car probably no big deal, on my race car where I have replaced bolts with shorter ones and shaved off studs to save less than an ounce at a time I have a problem adding 30lbs if I'm not going to get some serious performance gains. Keep in mind that spool is a total non issue on a race car where your lowest shift point leaves you over 6000rpms unless your in the wrong gear. There would have to be major top end gains with significantly higher efficiency for it to be worth it.

 

The he351ve weights 57lbs with all the trim intact. I have one, just checked. Most of the weight is from the HUGE turbine housing.

Weight is an indeed an issue, but not so much for me. The exhange of having to add 30lbs under the hood, for the potential benefits of VGT is well worth it IMO.


As for the issue of this diesel items reliability on a gasoline engine, people have been running this turbo w/o problem on higher HP gasoline engines. Just like with most things, if you use it correctly it will work. If you keep the vanes shut for too long, it will break. The simplicity of the vane system on the Holset as opposed to the garrett system for example is one of its strong points. The DSM community in particular has had a lot of people using this model while using the VGT feature w/ a mechanical actuator, and so far they seem to break motors more often they turbos.

As to opening up more power in the top end, doubt it. If you pick a turbo w/ turbine housing for power from 7k-9k rpm then thats what you get. You sacrafice low and mid, but no real loss since that is unimportant for a race car anyways. For my street car, im looking for boost from as soon as possible, all the way up to 8.5k. Ill add 1.5% more weight to my car if it means 50% more torque @ 3k rpms.

I'm so eager to get this going i think for now I might just start my tuning using a mechanical actuator to control vane depth. The electronic for this will take some time

 

I don't know if your car is already running on the haltech or what kind of investment you have in your current setup but a motec M800 would control it without issue not to mention the other benefits of the system of the haltech, it even has direct servo motor control.

I wasn't trying to dispute the advantages of a VGT system so don't take my post the wrong way I'm not trying to discourage you. I think what your doing is awesome and I'd be happy to help in any way. Like I said I plan on installing one on my street car but the gains just aren't there for a race car.

I would like to make a kit that included one but as of now the best option seems to be an M800 for control which no one will pay up for because they don't understand the advantages. If I can figure out how to do a PNP AEM or Haltech kit with some external pwm to servo motor circuitry (which should be pretty simple) for a reasonable cost then I will most likely turn it into a kit.

 

I have a brand new Garrett AVNT turbo, not sure of the exact model, but it is for a ford power stroke diesel I believe. Got it brand new in the box for less than $100 on ebay ;-) I did some research, and it supports up to around 500hp, so should be perfect for my needs. I did a little research on this as well, but never really ended up getting anywhere. Has anyone been able to successfully use on of the turbo's and control the vein geometry with a standalone aftermarket ecu??

 

Yes, like I said an M800 will control it to beyond OEM capabilities, personally I wouldn't do it with anything less on my own car but I could easily see it working with something less like an AEM or Haltech with some work and some compromises.

 

Industrial control system vendors (such as AGM Electronics and Phoenix Contact) have signal conditioners which can convert almost any input signal to any other signal you can imagine. The problem signals you mentioned are linear, so it's easy as pie for those guys.

You can certainly purchase one to convert the frequency signal representing turbine speed into a signal that could be read by any standalone ECU. You would need to specify the range of both input and output of the signal conditioner. In this case the output range is probably 1-5V ( to match the standard analog inputs on most standalones). The conditioner's input range is 0-?? Hz. Calculate this ?? by the formula:

Max Hz = max turbine RPM * # teeth on encoder wheel / 60
Round up to the nearest decent number, ie 18.435kHz becomes 20.0kHz.
Enter this rounded up number as the upper range of the analog input scale value in the ECU.

You could also get a signal conditioner to convert an ECU analog output into something that would drive the linear actuator solenoid which moves the guide vanes. You would need to know the solenoid's specs - signal range for full travel. This would be the output range for the signal conditioner. For the conditioner's input range, it would depend on which ECU output you want to use. You could try to use the ECU's standard wastegate driver output. It may be advantageous to use this if the ECU's native wastegate controller/maps/etc can be utilized. In this case the signal conditioner's input is 0-100% PWM duty cycle (you would need to find out the voltage level). If the ECU's native wastegate controller turns out to be incapable of handling all the modifications needed, then you might need to use another analog output. These typically have a standard signal range of 1-5V. So that would be the input range of the signal conditioner in this case.

But just getting the inputs and outputs working is the easy part of this job. EMAP and turbine speed bring two more variables into the control scheme (in addition to TPS, RPM, and MAP). You have 5 independent variables and two dependent variables. You have a lot of data (5-D tables), you may need multiple PID loops (controllers), and the controllers may need to interact in complicated ways (hi or low-limit clamping, split-ranging, cascading, remote setpoint, feedforward/biasing,etc,). You really need to find out how the OEMs control theirs.

If the above paragraph sounds like Greek to you, then you have a LOT of researching to do. I have nearly 20 years experience in industrial control system design, implementation, startup, and optimization. But I have never worked on any system this complicated. I would not undertake this project myself without a good understanding of how the OEMs have sucessfully controlled this bad boy. Using their model as a go-by, you could probably get the "tuning" numbers in the maps to work.