Electric wastegate control by turbo shaft speed?

[chevyburnout1]

Work is slow today and so I've been pondering a few projects I wouldn't mind tackling. In my ever pursuit of making my truck a "do-everything" type of truck I have stumbled upon a few questions on something I would like to try. I was curious if anyone has setup an electric boost controller that would tie into the signal output of the Garret speed sensor? My small turbo is a GT3788 with a .89a/r housing. It is very easy to overspool. I have a typical mechanical regulator controlling the gate that I have set based off of the turbine shaft speed gauge. Works very well for normal conditions as I can keep the small charger in it's perfect efficiency island based off of speed and pressure ratio. However if I change elevations obviously the pressure ratios change and the wastegate needs to be reset. A few weekends ago I hauled my camper over Trail Ridge Road here in Colorado which is like 12,500ft elevation. On an 800us pulse I could easily push the small turbo passed it's operating rpm, so it got me thinking into this project.

This isn't really a problem I'm experiencing, more so a challenge that I would like to wrap some thought into. Curious if anyone else has experience with electronic boost controllers, diesel related or not.

 

[duratothemax]

Shouldnt be hard to make the actual device...have a module that monitors shaft speed, atmospheric pressure, boost pressure...and then the output is the waste gate control...

It would probably take some fiddling to calibrate though. But once dialed in, it would be a perfect way to control things.

EDIT: wouldnt even need that second pressure sensor. Just have the module read the MAP sensor at key-on-engine-off (or below XXX turbine shaft speed), and then assume that as the current baro/atmospheric pressure.

Ben

 

[JoshH]

Shouldnt be hard to make the actual device...have a module that monitors shaft speed, atmospheric pressure, boost pressure...and then the output is the waste gate control...

It would probably take some fiddling to calibrate though. But once dialed in, it would be a perfect way to control things.

EDIT: wouldnt even need that second pressure sensor. Just have the module read the MAP sensor at key-on-engine-off (or below XXX turbine shaft speed), and then assume that as the current baro/atmospheric pressure.

Ben

Being used in a compound setup, I would think it would be better to reference the boost between stages and total boost rather than atmospheric.

 

[lts1ow]

A device that would take the 0-5v square wave and say once you hit a certain RPM you want, then opens a solenoid run in parallel with your MBC should dump WG pretty easy. That would be pretty simple but you basically hit the wall, dump WG and until your shaft speed eclipses your lower range its gonna stay open.

 

[chevyburnout1]

I was just thinking something simple that once it hits the 110k rpm limit it would start pwm a dump solenoid running the gate. But now you really have me thinking about factoring in the pressure ratios for being even more specific/ an extra safety barrier. I like it! How exactly do normal electric boost controllers work? I'm assuming it had a boost reference into it but that's the most of my knowledge.

 

[duratothemax]

is this wastegate variable and have to be finely controlled? Or just "open/closed"

 

[chevyburnout1]

I was assuming variable controlled... However I guess if the controller is just controlling the timing event when the boost reference hits the wastegate then technically it could just be an on/off control. This setup with the small Garrett is still relatively new so I haven't found the lower and upper limits of the gate yet.

 

[chevyburnout1]

I guess to sum up a bit, I originally am looking to eliminate the mechanical pressure regulator for the gate, and instead utilize just a pwm solenoid to control pressure to the gate for full control.

 

[DAVe3283]

...How exactly do normal electric boost controllers work? I'm assuming it had a boost reference into it but that's the most of my knowledge.

On modern gas cars with wastegates (like my 2013 Chevy Cruze Eco), the computer varies the amount of boost provided to the wastegate using something like the PPE Boost Fooler, but using a solenoid to PWM the bleed port very quickly. This allows full analog control of the wastegate.

This whole process is controlled by the ECU, which already has all the sensors it needs. Looking through the tune, it appears it is doing some math to estimate shaft speed, and factors that into the turbo operation somehow, so having an actual shaft speed sensor would be gravy.

is this wastegate variable and have to be finely controlled? Or just "open/closed"

As I mentioned above, by PWM-ing the bleed solenoid, you can control the wastegate very precisely.

A good deal of math (and/or trial and error tuning) would be needed to get an optimal control strategy, but that would certainly let the setup operate at maximum efficiency at all altitudes.

Being used in a compound setup, I would think it would be better to reference the boost between stages and total boost rather than atmospheric.

Yep, exactly. It is a shame boost sensors are so expensive, but once you install them, you should be able to T the output to the wastegate controller and to the 0-5V inputs on the FlashScan V2, and log both stages of boost while you are at it, which would be a neat bonus.

 

[chevyburnout1]

On modern gas cars with wastegates (like my 2013 Chevy Cruze Eco), the computer varies the amount of boost provided to the wastegate using something like the PPE Boost Fooler, but using a solenoid to PWM the bleed port very quickly. This allows full analog control of the wastegate.

This whole process is controlled by the ECU, which already has all the sensors it needs. Looking through the tune, it appears it is doing some math to estimate shaft speed, and factors that into the turbo operation somehow, so having an actual shaft speed sensor would be gravy.

As I mentioned above, by PWM-ing the bleed solenoid, you can control the wastegate very precisely.

A good deal of math (and/or trial and error tuning) would be needed to get an optimal control strategy, but that would certainly let the setup operate at maximum efficiency at all altitudes.

Yep, exactly. It is a shame boost sensors are so expensive, but once you install them, you should be able to T the output to the wastegate controller and to the 0-5V inputs on the FlashScan V2, and log both stages of boost while you are at it, which would be a neat bonus.

This is relatively what I was thinking. Like you said how most new cars regulate their mechanical gated turbos. So in theory I would have to setup the mechanical gate to open with no electrical help at 12,000ft, then the pwm solenoid could bleed off reference pressure to the gate to keep it closed the lower in elevation I go. So sea level technically would be the max pulse width for the bleed solenoid, if not completely open. Would take some tuning for sure which I'm not against. The electronic converter that takes the turbo shaft speed signal and converts it to that pwm is where I am a newb. I am definitely willing to try though. I'm one of those people that enjoy building the Legos more than actually playing with them once they're built lol.

 

[DAVe3283]

This is relatively what I was thinking. Like you said how most new cars regulate their mechanical gated turbos. So in theory I would have to setup the mechanical gate to open with no electrical help at 12,000ft, then the pwm solenoid could bleed off reference pressure to the gate to keep it closed the lower in elevation I go. So sea level technically would be the max pulse width for the bleed solenoid, if not completely open. Would take some tuning for sure which I'm not against. The electronic converter that takes the turbo shaft speed signal and converts it to that pwm is where I am a newb. I am definitely willing to try though. I'm one of those people that enjoy building the Legos more than actually playing with them once they're built lol.

Yeah, that's the gist of it. I am one of those people (like Ben & Fingers) who can make the electronic module to run that, but sadly, I rarely have the time to make cool things like that for other people. At the moment, my free time is spent on an electronic project for my Mom, when I am not working on personal projects.

I'm not sure how to read the shaft speed sensor, but I wouldn't think it would be very difficult. PWM control of the solenoid would be easy for sure. The algorithm & tables needed would probably be the hardest/longest part, and would be very setup dependent.

If Ben or Fingers don't make a module by the time I get my compound turbo setup (ETA: 5 years ), I'll probably make one of my own. If it works out well, I'd be willing to make you one for cost.

 

[chevyburnout1]

I have always envied people like you that have the knowledge and abilities to create these modules and setups. I'm infatuated with Ben's modules he creates and have been saving up for a mass purchase of them here hopefully soon haha. I specialize in the crude, simple setups like tying in big ugly relays and a bunch of wiring, just to make something like my DRL's on my Caddy shut off when the turn signals engage haha. I have a cheap pwm driver at home that I may rig up to something like a gasoline EVAP purge valve I have laying around. Actually see how hard the mechanics will be to setup. If it produces results then just figuring out how to convert the speed sensor signal will be my only downfall. It looks like it's either a 5v or 12v square wave signal depending on if you utilize the gauge as well.

 

[duratothemax]

Yeah, its probably a 5v square wave hall-effect sensor. Are there two wires or three wires going to it? What does it read off of on the turbine wheel?

I'd love to explore this, but unfortunately Im spread pretty thin with time and money to devote to R&D of low-volume-selling products.

I really wish I just had unlimited $$ so I could spend all day messing around with new ideas and inventing new things...the mundane day-to-day stuff gets boring, its the inventing/designing and problem solving that I really like.

 

[Hot COCOAL]

Damn you, smart people....
:hail:
:book:
:food:
:yourock:
:teknology:

I love threads like this...to you guys its so simple, and me I'm like WTF just happened:rofl:

 

[chevyburnout1]

Yeah, its probably a 5v square wave hall-effect sensor. Are there two wires or three wires going to it? What does it read off of on the turbine wheel?

I'd love to explore this, but unfortunately Im spread pretty thin with time and money to devote to R&D of low-volume-selling products.

I really wish I just had unlimited $$ so I could spend all day messing around with new ideas and inventing new things...the mundane day-to-day stuff gets boring, its the inventing/designing and problem solving that I really like.

I completely know what you're saying there. I would love if I could strap my truck to the dyno all day and just keep making minuscule changes to tuning over and over again, seeing what each changes create. Trying different setups, etc. But yeah it is a 2 wire output from the sensor. It get's it's reading from the actual compressor wheel vane passing by the tip. On the back of the gauge you have to set the dipswitch that correlates with how many vanes equals one revolution. The instructions on Garrett's website list the equation to calculate the speed.

 

[lts1ow]

A simple EBC would more than likely work.

Downside is, none of them can take a target boost level and hold. All of them will require coming up on target boost and then dialing in your levels. IE, can't tell EBC I want 10psi and always getting 10psi.

I have a basic eboost street in the car and after some fininicky adjusting its decent now. You will need to either find a road to hit max shaft speed or get it on a dyno but once you get that upper limit set its easy. In my case I only run WG for now, so I can adjust have fast/slow it opens/closes gate and I have an overboost setting that will in theory dump WG open. But, once you set what PSI level you want, it should always adjust based on elevation level IMO.

You can get a cheapo pod mound EBC for a few hundred.

OR, simplest way to do it, would be a hobbs switch set to your boost level at which your shaft speed is getting out of comfort zone, that then can control a solenoid that dumps WG. Thats stupid simple, but again, its an On/Off vs slow ramp up or down. Good safety setup though I suppose.

 

[chevyburnout1]

I understand what you're getting at with the cheap boost controller, but my thought on that is it only references a static boost pressure. My assumption is at high elevation the max boost pressure that charger can output vs shaft speed will be lower, in comparison to a low elevation where the boost pressure can be higher with the same shaft speed. This is all assumption though. Unfortunately I wasn't watching boost numbers when on top of Trail Ridge. The difference between sea level and 12,000ft may only change this setup a few psi.

After all that said we still have to take into account what the large charger is doing. At max shaft speed the large charger has already started taking off. I've noticed boost pressure can have a somewhat wide range at the same shaft speed, in regards to the tune and fuel.

I have the gate set at 100,000rpms here at 5000ft. Max for this charger is 111,000rpms. I saw a little over 120,000 on the pass.

Luckily I have access to a Mustang loaded dyno here at work which can hold the truck back at most speeds, so I look forward to creating something to try out at least! I believe I can get the mechanical/pwm side figured out where I can have a wide range of wastegate/shaft speed control. If it works great then I'll do some more research on the computer aspects of it.

 

[chevyburnout1]

Found a boost controller that can utilize turbo shaft speed! Only $1400! :roflmao::barf:

http://www.ams1000.com/index.php?option=com_content&view=article&id=50&Itemid=59

 

[juddski88]

It would be nice to have a bleed valve controlled directly by a pwm solenoid much like the one on our vvt turbos. On actual turbines we have a butterfly valve with an electric over hydraulic actuator that has been utilized by companies such as Solar for 30yrs. The hydraulic actuator acts as another deadband control much like it does on our turbos, and there is also deadbands programmed into the logic based on position so the valve moves consistently. It really is a simple system and is very reliable at high temps and pressures.

 

[lts1ow]

I believe any MS module will be able to configure a 0-5v square wave input and trigger boost control if you program it as such. A basic MS is pretty damn cheap these days and since you wouldn't need an engine harness, its not terribly pricey.