Garrett 712120 VNT Hella 6NW 008 412 actuator.
Garrett 712120 VNT Hella 6NW 008 412 actuator.
(11-03-2009, 04:01 PM)300SD81 I think this is the basic idea of how the code is gonna work:
1. try to maintain X psi of boost
if boost > X, open vanes unless fully opened
if boost < X, close unless fully closed
2. if exhaust pressure too high, open vanes even if boost < X
3. if egt too high, fully open vanes, sound alarm
(11-03-2009, 04:01 PM)300SD81 I think this is the basic idea of how the code is gonna work:
1. try to maintain X psi of boost
if boost > X, open vanes unless fully opened
if boost < X, close unless fully closed
2. if exhaust pressure too high, open vanes even if boost < X
3. if egt too high, fully open vanes, sound alarm
(11-05-2009, 10:32 AM)Tymbrymi(11-03-2009, 04:01 PM)300SD81 I think this is the basic idea of how the code is gonna work:
1. try to maintain X psi of boost
if boost > X, open vanes unless fully opened
if boost < X, close unless fully closed
2. if exhaust pressure too high, open vanes even if boost < X
3. if egt too high, fully open vanes, sound alarm
For optimal fuel economy, and if you have gobs of free time, the ideal thing to do would be to calculate engine load and only generate enough boost needed for that load. That would require LOTS of tuning though.
Sounding the alarm is a good idea, but if you reduce boost, the EGTs will go up since there is less air mass to cool off the injected fuel. The ALDA will cut back the fuel some when the boost goes away, but if the pump is that modified, it will still probably have EGTs that are too high.
I'm really really lucky... at work we have these AWESOME scopes:
http://www.tek.com/products/oscilloscopes/tds5000b/ They've spoiled me senseless.
(11-05-2009, 10:32 AM)Tymbrymi(11-03-2009, 04:01 PM)300SD81 I think this is the basic idea of how the code is gonna work:
1. try to maintain X psi of boost
if boost > X, open vanes unless fully opened
if boost < X, close unless fully closed
2. if exhaust pressure too high, open vanes even if boost < X
3. if egt too high, fully open vanes, sound alarm
For optimal fuel economy, and if you have gobs of free time, the ideal thing to do would be to calculate engine load and only generate enough boost needed for that load. That would require LOTS of tuning though.
Sounding the alarm is a good idea, but if you reduce boost, the EGTs will go up since there is less air mass to cool off the injected fuel. The ALDA will cut back the fuel some when the boost goes away, but if the pump is that modified, it will still probably have EGTs that are too high.
I'm really really lucky... at work we have these AWESOME scopes:
http://www.tek.com/products/oscilloscopes/tds5000b/ They've spoiled me senseless.
EDIT: Apparently made a post while I was blacked out drunk last night... I got it working Actuator sweeps from 0 to 100%, Video coming soon. I lost my cell phone at the bar the other night so I gotta borrow a camera..
VIDEO: http://www.youtube.com/watch?v=MkGsYd9mttI
used the webcam on my laptop, so the qualitys bretty bad...
UPDATE: Found out I used an undersized 3.3V voltage regulator and fried the ethernet controller when it failed short-circuit.. Was also having issues with the 3.3-5v level shiftter at >16MHz, so I decided to do a complete redesign for v2.0.. New board will be fully 3.3V and will use a microcontroller with integrated ethernet. Reusing left over PCB, so new board is approx 3in x 1in.
Also discovered that the actuator responds fine to both 12V and 5V PWM signals, and erratically to 3.3V PWM.
(11-08-2009, 03:40 AM)300SD81 Also discovered that the actuator responds fine to both 12V and 5V PWM signals, and erratically to 3.3V PWM.
(11-08-2009, 03:40 AM)300SD81 Also discovered that the actuator responds fine to both 12V and 5V PWM signals, and erratically to 3.3V PWM.
V2 of the controller is complete, and COMPUTER CONTROL IS WORKING I can move the actuator to any position using 3-5 byte UDP packets sent over the network. Now if I ever get a turbo installed into the car, I can control it with any computer with a network port
this is awesome work guys!
after the holiday season I'll definitely be looking to put an electronically actuated VNT into the GM Benz. If your project hasn't been completed by then, I'd be happy to help out (I have a good bit atmel and arduino experience).
the next big project I'd like to work on is a fly-by-wire accelerator with a force-feedback pedal. I'd like to make my IP rack actuated by a servo, and replace the function of the ALDA with force-feedback. The pedal would offer no resistance until you reach a point where the microcontroller decides you're about to start making black smoke, at which point the pedal pushes back against you. The key here is that the force-feedback would be weak enough that if you mashed the pedal, you could overcome it. This would allow you to inject additional fuel in emergency situations, while remaining environmentally friendly in day-to-day driving.
so of course, having an arduino-controlled VNT would complement that system nicely!
(11-19-2009, 03:35 PM)cell this is awesome work guys!
after the holiday season I'll definitely be looking to put an electronically actuated VNT into the GM Benz. If your project hasn't been completed by then, I'd be happy to help out (I have a good bit atmel and arduino experience).
the next big project I'd like to work on is a fly-by-wire accelerator with a force-feedback pedal. I'd like to make my IP rack actuated by a servo, and replace the function of the ALDA with force-feedback. The pedal would offer no resistance until you reach a point where the microcontroller decides you're about to start making black smoke, at which point the pedal pushes back against you. The key here is that the force-feedback would be weak enough that if you mashed the pedal, you could overcome it. This would allow you to inject additional fuel in emergency situations, while remaining environmentally friendly in day-to-day driving.
so of course, having an arduino-controlled VNT would complement that system nicely!
(11-19-2009, 03:35 PM)cell this is awesome work guys!
after the holiday season I'll definitely be looking to put an electronically actuated VNT into the GM Benz. If your project hasn't been completed by then, I'd be happy to help out (I have a good bit atmel and arduino experience).
the next big project I'd like to work on is a fly-by-wire accelerator with a force-feedback pedal. I'd like to make my IP rack actuated by a servo, and replace the function of the ALDA with force-feedback. The pedal would offer no resistance until you reach a point where the microcontroller decides you're about to start making black smoke, at which point the pedal pushes back against you. The key here is that the force-feedback would be weak enough that if you mashed the pedal, you could overcome it. This would allow you to inject additional fuel in emergency situations, while remaining environmentally friendly in day-to-day driving.
so of course, having an arduino-controlled VNT would complement that system nicely!
(11-21-2009, 03:32 AM)300SD81 (80MHz, 128K RAM, Ethernet.. Should be more than enough, I hope..)
(11-21-2009, 03:32 AM)300SD81 (80MHz, 128K RAM, Ethernet.. Should be more than enough, I hope..)
(11-21-2009, 05:01 AM)ForcedInduction(11-21-2009, 03:32 AM)300SD81 (80MHz, 128K RAM, Ethernet.. Should be more than enough, I hope..)
I guess nobody really needs more than 640k of ram after all.
(11-21-2009, 05:01 AM)ForcedInduction(11-21-2009, 03:32 AM)300SD81 (80MHz, 128K RAM, Ethernet.. Should be more than enough, I hope..)
I guess nobody really needs more than 640k of ram after all.
I got my panel scope this weekend (yes 300SD81, the other scope I have makes the one you pictured look like a state of the art piece). Good news is that the arduino is outputting PWM at 7.1ms and 5v. The bad news is that its not making the actuator actuate. Pretty dismayed at this point. Am going to redo the whole setup again and verify I don't have any bad wires.....
My version of the controller itself is functional, however it merely takes commands from a laptop which does the actual calculations and sensor reads. Final version with a microchip controlling it won't be complete until march 2010, when microchip starts shipping the 795F512H chip i want to use....
(12-08-2009, 05:25 PM)lars How close are you guys to have a complete working hella controller? approximately in weeks/months?
If so, will you sell a unit? Cuz I am closing up to buy a 2559v
(12-08-2009, 05:25 PM)lars How close are you guys to have a complete working hella controller? approximately in weeks/months?
If so, will you sell a unit? Cuz I am closing up to buy a 2559v
(12-08-2009, 07:19 PM)300SD81 Final version with a microchip controlling it won't be complete until march 2010, when microchip starts shipping the 795F512H chip i want to use....
(12-07-2009, 02:08 PM)winmutt (yes 300SD81, the other scope I have makes the one you pictured look like a state of the art piece).
(12-07-2009, 02:08 PM)winmutt Good news is that the arduino is outputting PWM at 7.1ms and 5v. The bad news is that its not making the actuator actuate. Pretty dismayed at this point. Am going to redo the whole setup again and verify I don't have any bad wires.....
(12-08-2009, 07:19 PM)300SD81 Final version with a microchip controlling it won't be complete until march 2010, when microchip starts shipping the 795F512H chip i want to use....
(12-07-2009, 02:08 PM)winmutt (yes 300SD81, the other scope I have makes the one you pictured look like a state of the art piece).
(12-07-2009, 02:08 PM)winmutt Good news is that the arduino is outputting PWM at 7.1ms and 5v. The bad news is that its not making the actuator actuate. Pretty dismayed at this point. Am going to redo the whole setup again and verify I don't have any bad wires.....
(11-06-2009, 04:22 PM)300SD81 Wouldn't the huge amount of excessive fuel cause it to cool down? I thought that was why the Finnish superturbos smoked so much.. to use the excess fuel as cooling.No. Excessive fuel is to help spooling up big turbos and it produces much heat. Lean AFR (as it is in basically every stock diesel engine) cools the engine and is good for fuel economy.
(11-06-2009, 04:22 PM)300SD81 Wouldn't the huge amount of excessive fuel cause it to cool down? I thought that was why the Finnish superturbos smoked so much.. to use the excess fuel as cooling.No. Excessive fuel is to help spooling up big turbos and it produces much heat. Lean AFR (as it is in basically every stock diesel engine) cools the engine and is good for fuel economy.
(01-12-2010, 10:58 AM)muuris With smaller VNT turbos that kind of smoking will get the vanes stuck pretty quickly -- but as they spool up nicely even without smoke, why even bother.
(01-12-2010, 10:58 AM)muuris With smaller VNT turbos that kind of smoking will get the vanes stuck pretty quickly -- but as they spool up nicely even without smoke, why even bother.
I totally forgot, 300SD81 came over and showed off his controller. We hooked up mine and spent some time dickering around with it. I am going to switch over to bit banging when I have time, neither of us could identify an issue with the current setup
(01-12-2010, 04:07 PM)300SD81 How quickly are you talking? My software cycles the vanes on every startup. Will that be enough to keep it from sticking?Cycling helps, as well as some extra movement during normal driving. But heavy smoking will get them stuck eventually, so with such a small VNT just run the engine with minimal smoke and you're fine.
(01-12-2010, 04:07 PM)300SD81 How quickly are you talking? My software cycles the vanes on every startup. Will that be enough to keep it from sticking?Cycling helps, as well as some extra movement during normal driving. But heavy smoking will get them stuck eventually, so with such a small VNT just run the engine with minimal smoke and you're fine.
DUDES!!!
I got my hella actuator to work!
video: http://www.youtube.com/watch?v=ENSbcuCAXt0
code, etc: http://jason.pepas.com/wiki/index.php?title=HellaSweep
I hacked around with the above code and experimentally found some limits on the frequency which it will respond to:
I couldn't get the controller to respond to 5v PWM. I had to use an N-channel mosfet and drive it with 9v PWM.
this is so badass!
(01-23-2010, 08:23 PM)Kiwibacon Cool, an excellent job you've done there.
So does a change from 9-12v change the working frequency?
(01-23-2010, 08:23 PM)Kiwibacon Cool, an excellent job you've done there.
So does a change from 9-12v change the working frequency?
Cell, I had mine all running off the same power supply, at first a 12v battery and then a PC p/s. No go. I do see you are using a different register value than me. My biggest problem is lack of time.
I thought this might aid in understanding VGT control logic. Its a data log from a 25 minute city and highway test drive of a 2006 Gillig Low Floor. Its powered by a Cummins ISM that has a Holset HE431V and 5-speed ZF automatic.
Interesting note; the actuator doesn't go above 70% closed or below 30% open.
This might help as well.
http://www.sae.org/technical/papers/900889
Another update.
Both the actuator and sensor boards have been working pretty good. I have the basic algorithm for turbo control written in C running on my laptop and running using simulated input. Couldn't route a decent one-layer interface between the ethernet PHY and the PIC32MX795, so I'm having a PCB manufactured for all 3 boards in one piece since the minimum board size was too big for just the control board. Heres the mostly finished design I made during class today.. Still needs a little work since I want to use the USB interface on the PIC32 as well.
Minimum order is 3 identical PCBs, so if they work correctly then I will have 2 complete systems available to whoever wants them. ($40 each for PCBs only + part list, at cost for me; or $250 fully soldered.)
Well... I lost everything on my hard drive. I'm reworking the boards now, and decided to just go 4 layer...routing these things with 4 layer board is trivial! Makes me want to make a bunch of stuff I thought I could never do by hand. I think I'm going to send them for manufacturing in a week or so, so if anyone is interested, I'll have the 2 extra PCBs for sale with a full part list and software support as soon as I receive them.
(03-30-2010, 01:19 PM)300SD81 Well... I lost everything on my hard drive. I'm reworking the boards now, and decided to just go 4 layer...routing these things with 4 layer board is trivial! Makes me want to make a bunch of stuff I thought I could never do by hand. I think I'm going to send them for manufacturing in a week or so, so if anyone is interested, I'll have the 2 extra PCBs for sale with a full part list and software support as soon as I receive them.
(03-30-2010, 01:19 PM)300SD81 Well... I lost everything on my hard drive. I'm reworking the boards now, and decided to just go 4 layer...routing these things with 4 layer board is trivial! Makes me want to make a bunch of stuff I thought I could never do by hand. I think I'm going to send them for manufacturing in a week or so, so if anyone is interested, I'll have the 2 extra PCBs for sale with a full part list and software support as soon as I receive them.
How are you controlling the PWM? The freq is listed in here I believe 8bit at 141hz. If you know how the CAN bus signals the turbo we would certainly like that information. You are using this on a Ford?
(11-05-2009, 10:32 AM)Tymbrymi(11-03-2009, 04:01 PM)300SD81 I think this is the basic idea of how the code is gonna work:
1. try to maintain X psi of boost
if boost > X, open vanes unless fully opened
if boost < X, close unless fully closed
2. if exhaust pressure too high, open vanes even if boost < X
3. if egt too high, fully open vanes, sound alarm
For optimal fuel economy, and if you have gobs of free time, the ideal thing to do would be to calculate engine load and only generate enough boost needed for that load. That would require LOTS of tuning though.
(11-05-2009, 10:32 AM)Tymbrymi(11-03-2009, 04:01 PM)300SD81 I think this is the basic idea of how the code is gonna work:
1. try to maintain X psi of boost
if boost > X, open vanes unless fully opened
if boost < X, close unless fully closed
2. if exhaust pressure too high, open vanes even if boost < X
3. if egt too high, fully open vanes, sound alarm
For optimal fuel economy, and if you have gobs of free time, the ideal thing to do would be to calculate engine load and only generate enough boost needed for that load. That would require LOTS of tuning though.
(12-01-2010, 10:39 AM)winmutt How are you controlling the PWM? The freq is listed in here I believe 8bit at 141hz.
(12-01-2010, 10:39 AM)winmutt If you know how the CAN bus signals the turbo we would certainly like that information.
(12-01-2010, 10:39 AM)winmutt You are using this on a Ford?
Hi Thanks for your reply.
(12-01-2010, 10:39 AM)winmutt How are you controlling the PWM? The freq is listed in here I believe 8bit at 141hz.
(12-01-2010, 10:39 AM)winmutt If you know how the CAN bus signals the turbo we would certainly like that information.
(12-01-2010, 10:39 AM)winmutt You are using this on a Ford?
(12-02-2010, 02:58 AM)Darqs By the way... actuator which I can control via PWM doesn't send CAN messages at all.
(12-02-2010, 02:58 AM)Darqs By the way... actuator which I can control via PWM doesn't send CAN messages at all.
(12-02-2010, 07:48 AM)Tymbrymi Here's your answer! It appears the controllers have different software. They might have the electronic components for both CAN and PWM on the circuit board, but they apparently don't want both PWM and CAN to be enabled at the same time. Kinda odd in my opinion!
(12-02-2010, 07:48 AM)Tymbrymi Here's your answer! It appears the controllers have different software. They might have the electronic components for both CAN and PWM on the circuit board, but they apparently don't want both PWM and CAN to be enabled at the same time. Kinda odd in my opinion!
(12-02-2010, 07:48 AM)Tymbrymi(12-02-2010, 02:58 AM)Darqs By the way... actuator which I can control via PWM doesn't send CAN messages at all.
Here's your answer! It appears the controllers have different software. They might have the electronic components for both CAN and PWM on the circuit board, but they apparently don't want both PWM and CAN to be enabled at the same time. Kinda odd in my opinion!
(12-02-2010, 07:48 AM)Tymbrymi(12-02-2010, 02:58 AM)Darqs By the way... actuator which I can control via PWM doesn't send CAN messages at all.
Here's your answer! It appears the controllers have different software. They might have the electronic components for both CAN and PWM on the circuit board, but they apparently don't want both PWM and CAN to be enabled at the same time. Kinda odd in my opinion!
(12-02-2010, 10:03 AM)winmutt Is that a wag or you know this for a fact. That would indicate more than likely there is an eeprom somewhere.....
(12-02-2010, 10:03 AM)winmutt Is that a wag or you know this for a fact. That would indicate more than likely there is an eeprom somewhere.....
Just ordered up a GT2359V, should have it in about a week. I'll be watching this thread with even more interest now!
Express PCB is a great company
(12-01-2010, 10:39 AM)winmutt If you know how the CAN bus signals the turbo we would certainly like that information. You are using this on a Ford?
(12-01-2010, 10:39 AM)winmutt If you know how the CAN bus signals the turbo we would certainly like that information. You are using this on a Ford?
This is the message from the Hella to the ECU or ECU to Hella? I would imagine it is just a status packet if its from the Hella.
(12-08-2010, 10:25 AM)winmutt This is the message from the Hella to the ECU or ECU to Hella?
(12-08-2010, 10:25 AM)winmutt I would imagine it is just a status packet if its from the Hella.
(12-08-2010, 10:25 AM)winmutt This is the message from the Hella to the ECU or ECU to Hella?
(12-08-2010, 10:25 AM)winmutt I would imagine it is just a status packet if its from the Hella.
I know very little about CAN. What I would expect is that the Hella actuator gives a status, ready for work, current vane positions etc.
What is needed is to find someone willing to read the codes during normal operation. I believe this actuator is used with CAN in europe in Fords. Mondeo? maybe? I would look on ford forums and see if anyone has pulled up these details.
(12-09-2010, 11:32 AM)winmutt I know very little about CAN. What I would expect is that the Hella actuator gives a status, ready for work, current vane positions etc.
(12-09-2010, 11:32 AM)winmutt What is needed is to find someone willing to read the codes during normal operation.1. Car with actuator via CAN control.
(12-09-2010, 11:32 AM)winmutt I believe this actuator is used with CAN in europe in Fords. Mondeo? maybe? I would look on ford forums and see if anyone has pulled up these details.I don't know if only in Fords. It's strange that actuators controled via PWM have the same part number like actuators controled via CAN! What is the difference?
Hi
(12-09-2010, 11:32 AM)winmutt I know very little about CAN. What I would expect is that the Hella actuator gives a status, ready for work, current vane positions etc.
(12-09-2010, 11:32 AM)winmutt What is needed is to find someone willing to read the codes during normal operation.1. Car with actuator via CAN control.
(12-09-2010, 11:32 AM)winmutt I believe this actuator is used with CAN in europe in Fords. Mondeo? maybe? I would look on ford forums and see if anyone has pulled up these details.I don't know if only in Fords. It's strange that actuators controled via PWM have the same part number like actuators controled via CAN! What is the difference?
Well "position" is easy enough I would assume its 0X00 0X00
"command" on the other hand is a different story, I would imagine you will just have to cycle through 256 numbers to find out
Its amazing how much data that thing is providing, stuff you would never know
Where did you get this image?
What was the can baudrate? Are you sure the PWM is not working? I have a hard time the same part number has different interfaces.
I guess temperature is ambient...