Holset HE351Ve VNT electronic control
Holset HE351Ve VNT electronic control
05-12-2009, 05:14 PM
#1
what I plan to use:
Arduino microcontroller w/AT328
http://arduino.cc/en/Main/ArduinoBoardDuemilanove
with motor control shield :
http://www.adafruit.com/index.php?main_p...ucts_id=81
to control this stepper motor:
http://www.robotshop.us/rbsoy14-soyo-uni...motor.html
I have the hardware pretty much sorted, I plan to use the IP rack position sensor
for "throttle" position, an EGT input, and possibly RPM input and maybe others.
I am currently working on the software design, and would like some feedback
on the operation..
I plan to post some pseudo code later on tonight with more details, but basically here are a few things to think about..
static relationship between rack position and amount of boost delivered
so lets say max boost is 20psi,
at idle, no boost
off idle to say 1/8 throttle 5 psi
1/8 throttle to 1/4 throttle 8 psi
and so on with a linear relationship..
or a non-linear relationship with a curve matching boost with rack position
based on rack position to amount of boost required to burn that much fuel efficently
this will probably have to be arrived at with a trial and error process, but the great thing
is that it will only require software changes, and can be done rather quickly.
what about cruising on highway with a small throttle input ?
other mechanisms i plan to implement:
feedback loop that monitors EGT and when EGT reaches pre-set levels
reduces boost, and if EGT reaches a pre-set max level reduces boost even more
or opens nozzle to max
(maybe with a time control.. like 1150deg OK but not for more then 1 min or so)
feedback loop that monitors engine temperature and limits boost
until engine warmed or reduces or eliminates boost when getting too hot.
2 or 3 mode boost maps controlled by switch for econ/normal/sport type mode
cycle nozzle on startup and or shutdown to keep mechanism clean
possibly a static boost control knob that moves nozzle ignoring throttle position signal
for testing/tuning/data collection etc..
full throttle triggers that could :
disengage A/C compressor
engage max cooling fan for A/W intercooler
other ?
any other ideas / feedback ?
I have abandoned the possibility of using the BLDC motor and CANbus controller already on the factory Holset controller.. I will however re-use the housing and reduction gears.
Im sure control via CANbus is possible, but for me.. not now.
I think that will be a project for an OM606 car
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
oel_brenner
05-12-2009, 05:14 PM #1I have assembled most of the parts needed to build the control system for the Holset HE351VE on my OM603.
what I plan to use:
Arduino microcontroller w/AT328
http://arduino.cc/en/Main/ArduinoBoardDuemilanove
with motor control shield :
http://www.adafruit.com/index.php?main_p...ucts_id=81
to control this stepper motor:
http://www.robotshop.us/rbsoy14-soyo-uni...motor.html
I have the hardware pretty much sorted, I plan to use the IP rack position sensor
for "throttle" position, an EGT input, and possibly RPM input and maybe others.
I am currently working on the software design, and would like some feedback
on the operation..
I plan to post some pseudo code later on tonight with more details, but basically here are a few things to think about..
static relationship between rack position and amount of boost delivered
so lets say max boost is 20psi,
at idle, no boost
off idle to say 1/8 throttle 5 psi
1/8 throttle to 1/4 throttle 8 psi
and so on with a linear relationship..
or a non-linear relationship with a curve matching boost with rack position
based on rack position to amount of boost required to burn that much fuel efficently
this will probably have to be arrived at with a trial and error process, but the great thing
is that it will only require software changes, and can be done rather quickly.
what about cruising on highway with a small throttle input ?
other mechanisms i plan to implement:
feedback loop that monitors EGT and when EGT reaches pre-set levels
reduces boost, and if EGT reaches a pre-set max level reduces boost even more
or opens nozzle to max
(maybe with a time control.. like 1150deg OK but not for more then 1 min or so)
feedback loop that monitors engine temperature and limits boost
until engine warmed or reduces or eliminates boost when getting too hot.
2 or 3 mode boost maps controlled by switch for econ/normal/sport type mode
cycle nozzle on startup and or shutdown to keep mechanism clean
possibly a static boost control knob that moves nozzle ignoring throttle position signal
for testing/tuning/data collection etc..
full throttle triggers that could :
disengage A/C compressor
engage max cooling fan for A/W intercooler
other ?
any other ideas / feedback ?
I have abandoned the possibility of using the BLDC motor and CANbus controller already on the factory Holset controller.. I will however re-use the housing and reduction gears.
Im sure control via CANbus is possible, but for me.. not now.
I think that will be a project for an OM606 car
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
05-12-2009, 10:53 PM
#2
(05-12-2009, 05:14 PM)oel_brenner static relationship between rack position and amount of boost delivered
so lets say max boost is 20psi,
at idle, no boost
off idle to say 1/8 throttle 5 psi
1/8 throttle to 1/4 throttle 8 psi
and so on with a linear relationship..
or a non-linear relationship with a curve matching boost with rack position
based on rack position to amount of boost required to burn that much fuel efficently
this will probably have to be arrived at with a trial and error process, but the great thing
is that it will only require software changes, and can be done rather quickly.
what about cruising on highway with a small throttle input ?
I have noticed on the 2003 Powerjoke at work that boost pressure is controlled in steps (0-5-10-15-20-25psi) loosely correlated to the throttle input. It usually goes to 15-20psi on normal acceleration and 25 at WOT with tow/haul mode on. Highway is usually 0 or 5.
ForcedInduction
05-12-2009, 10:53 PM
#2
(05-12-2009, 05:14 PM)oel_brenner static relationship between rack position and amount of boost delivered
so lets say max boost is 20psi,
at idle, no boost
off idle to say 1/8 throttle 5 psi
1/8 throttle to 1/4 throttle 8 psi
and so on with a linear relationship..
or a non-linear relationship with a curve matching boost with rack position
based on rack position to amount of boost required to burn that much fuel efficently
this will probably have to be arrived at with a trial and error process, but the great thing
is that it will only require software changes, and can be done rather quickly.
what about cruising on highway with a small throttle input ?
I have noticed on the 2003 Powerjoke at work that boost pressure is controlled in steps (0-5-10-15-20-25psi) loosely correlated to the throttle input. It usually goes to 15-20psi on normal acceleration and 25 at WOT with tow/haul mode on. Highway is usually 0 or 5.
05-16-2009, 08:25 AM
#3
One question for now: Why do you want to reduce boost at x °F? Fueling is based on boost/RPM/load/etc - but it's the fueling that is causing the higher EGT's - boost cools the EGT's.
Cheers,
Matt
'07 W211 OM642
'95 W124.131/722.435, 211k - daily driver/Superturbo project - OM606.962 with "M" pump...under construction! (build thread here)
'99 W210.025 - gone, but not forgotten
'94 Dodge/Cummins - gone, but not forgotten
HoleshotHolset
05-16-2009, 08:25 AM #3Very cool project!
One question for now: Why do you want to reduce boost at x °F? Fueling is based on boost/RPM/load/etc - but it's the fueling that is causing the higher EGT's - boost cools the EGT's.
Cheers,
Matt
'07 W211 OM642
'95 W124.131/722.435, 211k - daily driver/Superturbo project - OM606.962 with "M" pump...under construction! (build thread here)
'99 W210.025 - gone, but not forgotten
'94 Dodge/Cummins - gone, but not forgotten
JohnyWard
Unregistered
05-19-2009, 06:03 PM
#4
So the plan is to do a ground up retrofit of the stepper motor? You've got at least 3 analog inputs: egt, boost, "tps" ; Have you considered using the factory compressor speed sensor? I don't know what output you would get from the sensor, but it would make a great over speed protection code line.
On the CAN system, ForcedInduction posted a holset input schematic which includes a High and Low input into the turbo. Could the electronics package within the turbo be a CAN node itself? If it was could a microprocessor/CAN node outside with a series of analog to ditigal outputs sucessfully communicate to the turbo?
Thanks for the information to date, I can't wait to see the progress on the stand alone controller!
JohnyWard
05-19-2009, 06:03 PM #4Hello, I'm new to this forum, and have joined a couple of others the try and figure out the holset VNT interface.
So the plan is to do a ground up retrofit of the stepper motor? You've got at least 3 analog inputs: egt, boost, "tps" ; Have you considered using the factory compressor speed sensor? I don't know what output you would get from the sensor, but it would make a great over speed protection code line.
On the CAN system, ForcedInduction posted a holset input schematic which includes a High and Low input into the turbo. Could the electronics package within the turbo be a CAN node itself? If it was could a microprocessor/CAN node outside with a series of analog to ditigal outputs sucessfully communicate to the turbo?
Thanks for the information to date, I can't wait to see the progress on the stand alone controller!
05-20-2009, 10:42 AM
#5
(05-16-2009, 08:25 AM)HoleshotHolset Very cool project!
it's the fueling that is causing the higher EGT's - boost cools the EGT's.
Agreed, VERY cool project! Can't wait to hear about progress.
Also, more boost (to a point within the compressor's efficiency map), will cool the fire. The only way you are going to be able to have software regulation of EGTs is to control the fueling with software. That is what I want to do eventually (remove the governor, replace it with software), as well as computer control of the turbo. Effectively a complete engine management system for the 617. Thanks for the various links you posted oel_brenner. I don't know anything about electronics yet and it will be an education watching your progress, hopefully in time I will be able to contribute something!
GREASY_BEAST
05-20-2009, 10:42 AM #5(05-16-2009, 08:25 AM)HoleshotHolset Very cool project!
it's the fueling that is causing the higher EGT's - boost cools the EGT's.
Agreed, VERY cool project! Can't wait to hear about progress.
Also, more boost (to a point within the compressor's efficiency map), will cool the fire. The only way you are going to be able to have software regulation of EGTs is to control the fueling with software. That is what I want to do eventually (remove the governor, replace it with software), as well as computer control of the turbo. Effectively a complete engine management system for the 617. Thanks for the various links you posted oel_brenner. I don't know anything about electronics yet and it will be an education watching your progress, hopefully in time I will be able to contribute something!
05-20-2009, 12:44 PM
#6
(05-20-2009, 10:42 AM)GREASY_BEAST That is what I want to do eventually (remove the governor, replace it with software), as well as computer control of the turbo. Effectively a complete engine management system for the 617.
That's interesting - because I have a software controlled governor on my car and I'd rather have it fully mechanical.
IMHO, the only way I'd go for electronic is if it was common-rail. Then it's really worth it to be able to manipulate a lot more settings than just a SW controlled governor.So far, the IP's on the OM606.962's seem like they're pretty bulletproof...I'm not worried about it.
Cheers,
Matt
'07 W211 OM642
'95 W124.131/722.435, 211k - daily driver/Superturbo project - OM606.962 with "M" pump...under construction! (build thread here)
'99 W210.025 - gone, but not forgotten
'94 Dodge/Cummins - gone, but not forgotten
HoleshotHolset
05-20-2009, 12:44 PM #6(05-20-2009, 10:42 AM)GREASY_BEAST That is what I want to do eventually (remove the governor, replace it with software), as well as computer control of the turbo. Effectively a complete engine management system for the 617.
That's interesting - because I have a software controlled governor on my car and I'd rather have it fully mechanical.
IMHO, the only way I'd go for electronic is if it was common-rail. Then it's really worth it to be able to manipulate a lot more settings than just a SW controlled governor.So far, the IP's on the OM606.962's seem like they're pretty bulletproof...I'm not worried about it.
Cheers,
Matt
'07 W211 OM642
'95 W124.131/722.435, 211k - daily driver/Superturbo project - OM606.962 with "M" pump...under construction! (build thread here)
'99 W210.025 - gone, but not forgotten
'94 Dodge/Cummins - gone, but not forgotten
05-20-2009, 10:02 PM
#7
(05-19-2009, 06:03 PM)JohnyWard Have you considered using the factory compressor speed sensor? I don't know what output you would get from the sensor, but it would make a great over speed protection code line.
On the CAN system, ForcedInduction posted a holset input schematic which includes a High and Low input into the turbo. Could the electronics package within the turbo be a CAN node itself? If it was could a microprocessor/CAN node outside with a series of analog to ditigal outputs sucessfully communicate to the turbo?
yes, I plan to use the turbo RPM signal to protect against overspeed... that feature will probably be in version 2.0
The CAN system is way more complex then just sending PWM signals on the data wires.. its a whole system with a master controller
and each "node" ( the Holset controller is a node ) communicates back to the main system. The master system does not just send a signal to open
and close the VNT it sends full message packets in CANbus protocol. Just too much work for me right now.. someday when I own a OM606
and have a LOT more time I might figure out how to retrofit a native CAN turbo controller.
(05-16-2009, 08:25 AM)HoleshotHolset Very cool project!
One question for now: Why do you want to reduce boost at x °F? Fueling is based on boost/RPM/load/etc - but it's the fueling that is causing the higher EGT's - boost cools the EGT's.
(05-20-2009, 10:42 AM)GREASY_BEAST Also, more boost (to a point within the compressor's efficiency map), will cool the fire. The only way you are going to be able to have software regulation of EGTs is to control the fueling with software.
I agree, boost without fuel will cool EGT, but as I have no way to control fueling electronically (for now)
so my thinking was, if I am leaning into the GO pedal and not paying attention to the EGT gauge AND I ignore the
yellow warning EGT temp LED if I pull boost to nothing via VNT, power drops off immediately , black smoke cometh
and hopefully thats enough to communicate to driver to get right foot out of it...
Thanks for the well wishes everyone.. I hope to be moving the VNT via software
over the long weekend coming up..
so UPDATE:
I have the controller built, and the stepper motor controller built..
(have not done that much soldering in a long time)
I have the motor turning under software control
I have removed the original BLDC motor and PCB from the Holset housing
FYI its pressed in, and I had to destroy it to remove it..
it was not made to be a serviceable at all.
I have some machine work to do to the motor housing in order to
mount my new stepper motor.
working on the software now, but I still need to figure why I cant get any kind of signal
from the rack position sensor off the IP ( see other thread )
also, on control of fueling, I seem to remember once the IP is past a certain point
either position or RPM, pushing the ALDA plunger down has no effect ?
I removed my ALDA long long ago, but I was thinking a servo to re-enable it
but use the mechanism to reduce fueling triggered by a high EGT event.
just a thought..
This post was last modified: 05-20-2009, 10:02 PM by oel_brenner.
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
oel_brenner
05-20-2009, 10:02 PM #7(05-19-2009, 06:03 PM)JohnyWard Have you considered using the factory compressor speed sensor? I don't know what output you would get from the sensor, but it would make a great over speed protection code line.
On the CAN system, ForcedInduction posted a holset input schematic which includes a High and Low input into the turbo. Could the electronics package within the turbo be a CAN node itself? If it was could a microprocessor/CAN node outside with a series of analog to ditigal outputs sucessfully communicate to the turbo?
yes, I plan to use the turbo RPM signal to protect against overspeed... that feature will probably be in version 2.0
The CAN system is way more complex then just sending PWM signals on the data wires.. its a whole system with a master controller
and each "node" ( the Holset controller is a node ) communicates back to the main system. The master system does not just send a signal to open
and close the VNT it sends full message packets in CANbus protocol. Just too much work for me right now.. someday when I own a OM606
and have a LOT more time I might figure out how to retrofit a native CAN turbo controller.
(05-16-2009, 08:25 AM)HoleshotHolset Very cool project!
One question for now: Why do you want to reduce boost at x °F? Fueling is based on boost/RPM/load/etc - but it's the fueling that is causing the higher EGT's - boost cools the EGT's.
(05-20-2009, 10:42 AM)GREASY_BEAST Also, more boost (to a point within the compressor's efficiency map), will cool the fire. The only way you are going to be able to have software regulation of EGTs is to control the fueling with software.
I agree, boost without fuel will cool EGT, but as I have no way to control fueling electronically (for now)
so my thinking was, if I am leaning into the GO pedal and not paying attention to the EGT gauge AND I ignore the
yellow warning EGT temp LED if I pull boost to nothing via VNT, power drops off immediately , black smoke cometh
and hopefully thats enough to communicate to driver to get right foot out of it...
Thanks for the well wishes everyone.. I hope to be moving the VNT via software
over the long weekend coming up..
so UPDATE:
I have the controller built, and the stepper motor controller built..
(have not done that much soldering in a long time)
I have the motor turning under software control
I have removed the original BLDC motor and PCB from the Holset housing
FYI its pressed in, and I had to destroy it to remove it..
it was not made to be a serviceable at all.
I have some machine work to do to the motor housing in order to
mount my new stepper motor.
working on the software now, but I still need to figure why I cant get any kind of signal
from the rack position sensor off the IP ( see other thread )
also, on control of fueling, I seem to remember once the IP is past a certain point
either position or RPM, pushing the ALDA plunger down has no effect ?
I removed my ALDA long long ago, but I was thinking a servo to re-enable it
but use the mechanism to reduce fueling triggered by a high EGT event.
just a thought..
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
05-21-2009, 08:55 AM
#8
(05-20-2009, 10:02 PM)oel_brenner also, on control of fueling, I seem to remember once the IP is past a certain point
either position or RPM, pushing the ALDA plunger down has no effect ?
I removed my ALDA long long ago, but I was thinking a servo to re-enable it
but use the mechanism to reduce fueling triggered by a high EGT event.
just a thought..
I'm not sure whether the ALDA works that way or not. I thought when you remove the boost line to the ALDA that limited fueling across the board, but I've never tested from this standpoint.. I was always trying to get more fuel not less
I guess you could re-attach it or wedge something in there to depress the plunger and go for a drive... a really slow, boring one.
I really like the idea of putting some sort of engine protection in the system. No sense to modernizing it if its not going to be user-friendly.
GREASY_BEAST
05-21-2009, 08:55 AM #8(05-20-2009, 10:02 PM)oel_brenner also, on control of fueling, I seem to remember once the IP is past a certain point
either position or RPM, pushing the ALDA plunger down has no effect ?
I removed my ALDA long long ago, but I was thinking a servo to re-enable it
but use the mechanism to reduce fueling triggered by a high EGT event.
just a thought..
I'm not sure whether the ALDA works that way or not. I thought when you remove the boost line to the ALDA that limited fueling across the board, but I've never tested from this standpoint.. I was always trying to get more fuel not less
I guess you could re-attach it or wedge something in there to depress the plunger and go for a drive... a really slow, boring one.
I really like the idea of putting some sort of engine protection in the system. No sense to modernizing it if its not going to be user-friendly.
05-22-2009, 10:55 PM
#9
stepper motor
holset actuator gear section
modified gear case cover (I machined the top off)
and the shaft/ drive gear out of the old BLDC motor
![[Image: VNT-actparts.jpg]](http://www.coastalboatsales.com/jfraley/VNT/VNT-actparts.jpg)
just have to mod the shaft a bit, make a shaft coupling
to attach the drive gear to the stepper, mount the stepper
on the gear case, and a few other hundred things..
![[Image: VNT-actpart2.jpg]](http://www.coastalboatsales.com/jfraley/VNT/VNT-actpart2.jpg)
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
oel_brenner
05-22-2009, 10:55 PM #9micro-controller and stepper control board
stepper motor
holset actuator gear section
modified gear case cover (I machined the top off)
and the shaft/ drive gear out of the old BLDC motor
just have to mod the shaft a bit, make a shaft coupling
to attach the drive gear to the stepper, mount the stepper
on the gear case, and a few other hundred things..
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
09-07-2009, 06:43 AM
#10
deerefanatic
09-07-2009, 06:43 AM #10So, what's happening with this project? I'm wanting to put this same turbo on a 12valve cummins and want something better than a wastegate actuator to control it.... I'm hopelessly lost though when it comes to programming.......
09-11-2009, 06:29 PM
#11
Section106
09-11-2009, 06:29 PM #11If you get this up and working can I send you my HE351Ve to modify? I'm planning on putting it on a 617.
09-11-2009, 07:19 PM
#12
I am working on a turbo flange and connector pipe to mount the turbo up, then fab up an exhaust, and drive around with the manual control and get some base-line numbers for RPM / throttle position / load etc... fine tune the calibration and hook up the rest of the sensors.
once the code is close I will post it here as well as on github.
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
oel_brenner
09-11-2009, 07:19 PM #12I have the micro-controller moving the stepper back and forth based loosely on what I plan to use for a throttle position sensor, still have to machine a base to mount the stepper to the modified holset housing.
I am working on a turbo flange and connector pipe to mount the turbo up, then fab up an exhaust, and drive around with the manual control and get some base-line numbers for RPM / throttle position / load etc... fine tune the calibration and hook up the rest of the sensors.
once the code is close I will post it here as well as on github.
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
09-15-2009, 05:42 AM
#13
But I'd like to see about your controller to retrofit into an earlier model dodge with the cummins engine. Could the TPS signal be adjusted to various voltages? The factory VP44'ed trucks have an APPS (TPS) on them and use a value of around 5 volts (I'll have to verify the range).
You have gained my attention.
But I'd like to see about your controller to retrofit into an earlier model dodge with the cummins engine. Could the TPS signal be adjusted to various voltages? The factory VP44'ed trucks have an APPS (TPS) on them and use a value of around 5 volts (I'll have to verify the range).
09-16-2009, 10:13 AM
#14
What are you using for boost and throttle sensors? The boost sensor on the manifold is a ground switch at 16psi, the switch on the 85 linkage opens at about half WOT and is adjustable. I have ordered a samples boost sensor (2 are supposed to come in the order) as well as an amp ic for the EGT signal. I also plan on using the alternator to generate a tach signal for the 84 and 85 tachs (I believe the earlier ones are fairly easy as well), as I forgot to swap the bellhousing when I swapped my engine.
This post was last modified: 09-16-2009, 10:24 AM by winmutt.
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
Oel we should collaborate on the code.
What are you using for boost and throttle sensors? The boost sensor on the manifold is a ground switch at 16psi, the switch on the 85 linkage opens at about half WOT and is adjustable. I have ordered a samples boost sensor (2 are supposed to come in the order) as well as an amp ic for the EGT signal. I also plan on using the alternator to generate a tach signal for the 84 and 85 tachs (I believe the earlier ones are fairly easy as well), as I forgot to swap the bellhousing when I swapped my engine.
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
09-18-2009, 05:33 PM
#15
(09-16-2009, 10:13 AM)winmutt Oel we should collaborate on the code.
What are you using for boost and throttle sensors? The boost sensor on the manifold is a ground switch at 16psi, the switch on the 85 linkage opens at about half WOT and is adjustable. I have ordered a samples boost sensor (2 are supposed to come in the order) as well as an amp ic for the EGT signal. I also plan on using the alternator to generate a tach signal for the 84 and 85 tachs (I believe the earlier ones are fairly easy as well), as I forgot to swap the bellhousing when I swapped my engine.
I will post the code on github
( I am oel-brenner on there )
I am using this var resistor as a stand-in for a TPS
http://www.robotshop.us/inex-variable-re...board.html
not sure what I am going to use for production, but for initial testing
I plan to connect above slider to the accel pedal to start to capture some data and get baseline information
have not searched for a boost sensor yet, which ones are you looking at ?
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
oel_brenner
09-18-2009, 05:33 PM #15(09-16-2009, 10:13 AM)winmutt Oel we should collaborate on the code.
What are you using for boost and throttle sensors? The boost sensor on the manifold is a ground switch at 16psi, the switch on the 85 linkage opens at about half WOT and is adjustable. I have ordered a samples boost sensor (2 are supposed to come in the order) as well as an amp ic for the EGT signal. I also plan on using the alternator to generate a tach signal for the 84 and 85 tachs (I believe the earlier ones are fairly easy as well), as I forgot to swap the bellhousing when I swapped my engine.
I will post the code on github
( I am oel-brenner on there )
I am using this var resistor as a stand-in for a TPS
http://www.robotshop.us/inex-variable-re...board.html
not sure what I am going to use for production, but for initial testing
I plan to connect above slider to the accel pedal to start to capture some data and get baseline information
have not searched for a boost sensor yet, which ones are you looking at ?
cars
1991 W126 350SDL turbodiesel
1987 W124 E300D turbodiesel
1987 W126 300SDL turbodiesel
1984 W107 SL300D turbodiesel
1974 W115 /8 300D diesel
trucks
2001 Dodge RAM 3500 4x4 Cummins turbodiesel
boats
1974 Uniflite "Salty Dog" powered by
2x OM617.951 Mercedes Benz 5Cyl turbodiesels
09-18-2009, 09:46 PM
#16
(09-18-2009, 05:33 PM)oel_brenner(09-16-2009, 10:13 AM)winmutt Oel we should collaborate on the code.
What are you using for boost and throttle sensors? The boost sensor on the manifold is a ground switch at 16psi, the switch on the 85 linkage opens at about half WOT and is adjustable. I have ordered a samples boost sensor (2 are supposed to come in the order) as well as an amp ic for the EGT signal. I also plan on using the alternator to generate a tach signal for the 84 and 85 tachs (I believe the earlier ones are fairly easy as well), as I forgot to swap the bellhousing when I swapped my engine.
I will post the code on github
( I am oel-brenner on there )
I am using this var resistor as a stand-in for a TPS
http://www.robotshop.us/inex-variable-re...board.html
not sure what I am going to use for production, but for initial testing
I plan to connect above slider to the accel pedal to start to capture some data and get baseline information
have not searched for a boost sensor yet, which ones are you looking at ?
MPX5500DP http://courses.essex.ac.uk/CE/CE134/rest...PX5500.pdf
Should be good for 50-55 psi. Funnily enough Hella has throttle position sensors but I don' think they are cheap. The picture in their catalog is much like (exact?) the position sensor on the garrett vnt actuator.
I went to github and couldnt find your username. I was thinking to good basic set of code includes for different. This is my first time for this kind of stuff, I do business apps. Have always wanted to do embedded work.
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
(09-18-2009, 05:33 PM)oel_brenner(09-16-2009, 10:13 AM)winmutt Oel we should collaborate on the code.
What are you using for boost and throttle sensors? The boost sensor on the manifold is a ground switch at 16psi, the switch on the 85 linkage opens at about half WOT and is adjustable. I have ordered a samples boost sensor (2 are supposed to come in the order) as well as an amp ic for the EGT signal. I also plan on using the alternator to generate a tach signal for the 84 and 85 tachs (I believe the earlier ones are fairly easy as well), as I forgot to swap the bellhousing when I swapped my engine.
I will post the code on github
( I am oel-brenner on there )
I am using this var resistor as a stand-in for a TPS
http://www.robotshop.us/inex-variable-re...board.html
not sure what I am going to use for production, but for initial testing
I plan to connect above slider to the accel pedal to start to capture some data and get baseline information
have not searched for a boost sensor yet, which ones are you looking at ?
MPX5500DP http://courses.essex.ac.uk/CE/CE134/rest...PX5500.pdf
Should be good for 50-55 psi. Funnily enough Hella has throttle position sensors but I don' think they are cheap. The picture in their catalog is much like (exact?) the position sensor on the garrett vnt actuator.
I went to github and couldnt find your username. I was thinking to good basic set of code includes for different. This is my first time for this kind of stuff, I do business apps. Have always wanted to do embedded work.
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
09-19-2009, 06:40 AM
#18
This post was last modified: 09-19-2009, 06:44 AM by winmutt.
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
VCV is good I suppose because it will work across all models pretty much, where as my plans to use the 85 linkage sensor is only suitable for 85 and up and only show WOT.
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
09-20-2009, 08:02 AM
#20
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
Why do you need a range? Seems to me the only time that throttle position matters is for passing when coming out of a cruise mode.
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
09-20-2009, 02:52 PM
#21
The set point for boost is just as high as the driver wants, not always max. boost when touching the pedal. This is how it works for the TDIs and will work for my SD. I believe this is necessary for harmonic boost development and driveability.
Tom
each pedal position has its own set boost. Starting from very low boost at low pedal position, ending in 0.9 set boost for max. pedal position.
The set point for boost is just as high as the driver wants, not always max. boost when touching the pedal. This is how it works for the TDIs and will work for my SD. I believe this is necessary for harmonic boost development and driveability.
Tom
11-09-2009, 12:54 AM
#22
im doing the same thing but with gas motors in mind, as the holset is the only vgt/vnt that can take it. i to am using the arduino but instead of a stepper im using the DC brushless motor driven by one of these:
http://www.jrkerr.com/picsrvsc.pdf
and at $30.00 in single qty. i built the whole works for $100.00 i still have some things to work out (like the fact that i've built a whole control system and still dont own one of these turbos) encoder feedback is the biggest problem right now, i've come around to the idea of laser etching quadrature lines on the back of one of the gears to make an encoder. i would love to use comutation logic to decode position, but that dosent seem possible at the moment.
for control (when ever i get one of these things on a engine) i plan on using the shaft speed sensor on the turbo and the arduino PID library (i love arduino) to control turbocharger RPM. since turbo RPM=boost (see Turbochargers by hugh macinnes) you would just set the required psi (RPM) and there is your boost controler. as a side benefit the PID in the arduino will close the vgt only enough to maintain the set turboRPM. so it will create only enough backpressure to maintain the set boostPSI regardless of engine RPM(way cool). ofcorse all this will need a throtal position sensor to tell everything what to do. thats the basic layout. another interesting thing im going to try is running a without a blow off valve, having everything hooked to the throtal position sensor may make that possible.
keep us all updated, some day somebody will make one of these turbos work, its just to cool and to cheap not to. besides having a 2.4l that hits 25psi at 3000rpm and has intake pressure that is higher than the exhaust pressure all the time, is what its all about.
sweet, some one else is doing this to...
im doing the same thing but with gas motors in mind, as the holset is the only vgt/vnt that can take it. i to am using the arduino but instead of a stepper im using the DC brushless motor driven by one of these:
http://www.jrkerr.com/picsrvsc.pdf
and at $30.00 in single qty. i built the whole works for $100.00 i still have some things to work out (like the fact that i've built a whole control system and still dont own one of these turbos) encoder feedback is the biggest problem right now, i've come around to the idea of laser etching quadrature lines on the back of one of the gears to make an encoder. i would love to use comutation logic to decode position, but that dosent seem possible at the moment.
for control (when ever i get one of these things on a engine) i plan on using the shaft speed sensor on the turbo and the arduino PID library (i love arduino) to control turbocharger RPM. since turbo RPM=boost (see Turbochargers by hugh macinnes) you would just set the required psi (RPM) and there is your boost controler. as a side benefit the PID in the arduino will close the vgt only enough to maintain the set turboRPM. so it will create only enough backpressure to maintain the set boostPSI regardless of engine RPM(way cool). ofcorse all this will need a throtal position sensor to tell everything what to do. thats the basic layout. another interesting thing im going to try is running a without a blow off valve, having everything hooked to the throtal position sensor may make that possible.
keep us all updated, some day somebody will make one of these turbos work, its just to cool and to cheap not to. besides having a 2.4l that hits 25psi at 3000rpm and has intake pressure that is higher than the exhaust pressure all the time, is what its all about.
11-16-2009, 12:40 PM
#23
I have OM606 in my daily driver. Last summer I had GT3788V (from a 2005 Duramax), but I wasn't satisfied with it. Didn't reach full boost until ~3500rpm. IP is mechanical from OM603, modified by a diesel shop and partly myself, puts out 115mm^3 and lets the engine rev up to 6500rpm. The turbine wheel/housing on the GT37V were way too big even for OM606, so the vanes could not be opened more than 50%, or the boost/EMP would dive immediately, even on full boost. So I had to have vanes in a not-so-flowing position, and that caused considerably high EMPs on high revs. With max boost set to 1,5bar there was ~2,0bar EMP, and with 1,8bar of boost EMP was 2,9bars. Dynoed with 1,8bar setting, got 388hp and 580Nm on the engine. Not that bad, but should get a 20hp and 100Nm more with lower EMPs and the same amount of fueling.
This winter I will be converting it to HE351V. It should have much better response, as both the turbine and compressor wheels are way smaller. I would have liked to try out a smaller turbine housing on GT37v, but didn't find any for sale. In Europe it's very hard to find these big VNT/VGT turbos second hand.
I have oopic controlling my turbo, a throttle position sensor on the throttle linkage, two MAPs (4bar abs, boost and exhaust gas pressure), engine oil pressure and temperature sensor etc. Didn't have time/motivation to make the RPM input circuits, that will also be done this winter. I also wrote a data logging program to laptop, so I can watch boost/EMP/etc data realtime when driving, and data is also saved so after test drives it's easier to see how the program should be altered.
I drove 8000km with the above setup last summer and tried out a couple of ways to control the system. Now the program lets the vanes be in ~30% opened position on idle and cruising (gas pedal pressed less than 10%). When pedal is pressed more than that, the position of the vanes is primarily controlled by the boost (also when pedal is depressed quickly, the vanes go into 15% opened position for a while). There is a limit for EMP, when it gets above a certain limit, it opens vanes more to lessen it. There is also a separate wastegate in the exhaust manifold, driven only by EMP (set to 2,8bar), which effectively reduces EMP peaks and acts as a safety/backup device if the electrical controls would fail.
With HE351V I'll simplify the control system. A pneumatic cylinder instead of the stepper motor, which let's the vane be in around 50% position normally. When accelerator is depressed, vacuum is directed to the cylinder, forcing the vane to close (against a mechanical adjustable limit). When boost reaches about 0,3bar or EMP goes over 1bar, vacuum is removed and EMP controls the cylinder, pushing vane open when EMP goes high. Thus it doesn't control boost directly, but tries to maintain a certain EMP level. I'm assume the vane can't hold EMP down enough, so the wastegate will do the rest. Boost will hopefully be around the wanted level.
Just some thoughts of mine, hope they give inspiration to finish your project.
This post was last modified: 11-16-2009, 12:51 PM by muuris.
This is very interesting thread.
I have OM606 in my daily driver. Last summer I had GT3788V (from a 2005 Duramax), but I wasn't satisfied with it. Didn't reach full boost until ~3500rpm. IP is mechanical from OM603, modified by a diesel shop and partly myself, puts out 115mm^3 and lets the engine rev up to 6500rpm. The turbine wheel/housing on the GT37V were way too big even for OM606, so the vanes could not be opened more than 50%, or the boost/EMP would dive immediately, even on full boost. So I had to have vanes in a not-so-flowing position, and that caused considerably high EMPs on high revs. With max boost set to 1,5bar there was ~2,0bar EMP, and with 1,8bar of boost EMP was 2,9bars. Dynoed with 1,8bar setting, got 388hp and 580Nm on the engine. Not that bad, but should get a 20hp and 100Nm more with lower EMPs and the same amount of fueling.
This winter I will be converting it to HE351V. It should have much better response, as both the turbine and compressor wheels are way smaller. I would have liked to try out a smaller turbine housing on GT37v, but didn't find any for sale. In Europe it's very hard to find these big VNT/VGT turbos second hand.
I have oopic controlling my turbo, a throttle position sensor on the throttle linkage, two MAPs (4bar abs, boost and exhaust gas pressure), engine oil pressure and temperature sensor etc. Didn't have time/motivation to make the RPM input circuits, that will also be done this winter. I also wrote a data logging program to laptop, so I can watch boost/EMP/etc data realtime when driving, and data is also saved so after test drives it's easier to see how the program should be altered.
I drove 8000km with the above setup last summer and tried out a couple of ways to control the system. Now the program lets the vanes be in ~30% opened position on idle and cruising (gas pedal pressed less than 10%). When pedal is pressed more than that, the position of the vanes is primarily controlled by the boost (also when pedal is depressed quickly, the vanes go into 15% opened position for a while). There is a limit for EMP, when it gets above a certain limit, it opens vanes more to lessen it. There is also a separate wastegate in the exhaust manifold, driven only by EMP (set to 2,8bar), which effectively reduces EMP peaks and acts as a safety/backup device if the electrical controls would fail.
With HE351V I'll simplify the control system. A pneumatic cylinder instead of the stepper motor, which let's the vane be in around 50% position normally. When accelerator is depressed, vacuum is directed to the cylinder, forcing the vane to close (against a mechanical adjustable limit). When boost reaches about 0,3bar or EMP goes over 1bar, vacuum is removed and EMP controls the cylinder, pushing vane open when EMP goes high. Thus it doesn't control boost directly, but tries to maintain a certain EMP level. I'm assume the vane can't hold EMP down enough, so the wastegate will do the rest. Boost will hopefully be around the wanted level.
Just some thoughts of mine, hope they give inspiration to finish your project.
11-16-2009, 02:20 PM
#25
(11-16-2009, 01:25 PM)ForcedInduction Very nice setup! What chassis is that?
Thanks! It's a E32 7-series. The 606 swap took 3 months, but I had to wait for the IP for another month after that.. Actually I planned a W123 for this engine setup, but it's so hard to find a decent one here. I bought this for last winter about an year ago and planned to sell it in the spring, but then I had some other thoughts, as I had had the GT37v in the garage for a while. The solenoid on in was broken (propably the reason it was for sale for such a low price), but I managed to disasseble it and remake the coil out of similiar wire as the original.
Old one out, some washing, oil pan mod and there it goes. All mounts etc are selfmade. Transmission is reinforced 722.3 from a W124 300TDT. Exhaust manifold is also selfmade, though a professional welder did the final tigging.
Now I wish I had used a lighter chassis. In this heavy tincan (~1800kg) you just can't feel the acceleration. Not that slow for a daily driver as it gets over 170km/h on a 1/4 mile (compared to 135km/h with the original 3-liter petrol). So after this is finished, maybe another 606 and a lightweight chassis coming up..
edit: About the slow boost build up on this.. as a dd, I want this to be moderate in the amount of black smoke. Tried massive no-boost fueling (typically seen on fast diesel Mercs in Finland) and it helped a lot, but the it just doesn't fit the style and dignity (
) of this vehicle. Plus, one has to was the the rear end daily..
This post was last modified: 11-16-2009, 02:27 PM by muuris.
(11-16-2009, 01:25 PM)ForcedInduction Very nice setup! What chassis is that?
Thanks! It's a E32 7-series. The 606 swap took 3 months, but I had to wait for the IP for another month after that.. Actually I planned a W123 for this engine setup, but it's so hard to find a decent one here. I bought this for last winter about an year ago and planned to sell it in the spring, but then I had some other thoughts, as I had had the GT37v in the garage for a while. The solenoid on in was broken (propably the reason it was for sale for such a low price), but I managed to disasseble it and remake the coil out of similiar wire as the original.
Old one out, some washing, oil pan mod and there it goes. All mounts etc are selfmade. Transmission is reinforced 722.3 from a W124 300TDT. Exhaust manifold is also selfmade, though a professional welder did the final tigging.
Now I wish I had used a lighter chassis. In this heavy tincan (~1800kg) you just can't feel the acceleration. Not that slow for a daily driver as it gets over 170km/h on a 1/4 mile (compared to 135km/h with the original 3-liter petrol). So after this is finished, maybe another 606 and a lightweight chassis coming up..
edit: About the slow boost build up on this.. as a dd, I want this to be moderate in the amount of black smoke. Tried massive no-boost fueling (typically seen on fast diesel Mercs in Finland) and it helped a lot, but the it just doesn't fit the style and dignity (
) of this vehicle. Plus, one has to was the the rear end daily..
11-16-2009, 04:32 PM
#26
(11-16-2009, 02:20 PM)muuris Now I wish I had used a lighter chassis.
Go for a E30. Then maybe a 605 is plenty.? That is where my 602td hopefully ends up mated to a 525tds manual box.
edit: About the slow boost build up on this.. as a dd, I want this to be moderate in the amount of black smoke. Tried massive no-boost fueling (typically seen on fast diesel Mercs in Finland) and it helped a lot, but the it just doesn't fit the style and dignity (
Yes, smoke is not ideal on a daily driver.
OT: Do you know of any successful stories without too much smoke with 7mm and twin turbo`d 60x from finland? sequ or compound..? links?
(11-16-2009, 02:20 PM)muuris Now I wish I had used a lighter chassis.
Go for a E30. Then maybe a 605 is plenty.? That is where my 602td hopefully ends up mated to a 525tds manual box.
edit: About the slow boost build up on this.. as a dd, I want this to be moderate in the amount of black smoke. Tried massive no-boost fueling (typically seen on fast diesel Mercs in Finland) and it helped a lot, but the it just doesn't fit the style and dignity (
Yes, smoke is not ideal on a daily driver.
OT: Do you know of any successful stories without too much smoke with 7mm and twin turbo`d 60x from finland? sequ or compound..? links?
11-16-2009, 04:59 PM
#27
http://www.youtube.com/watch?v=_8lkKZO5vBY
11-17-2009, 12:54 PM
#29
Engine has modified internals (head work, larger valves etc). Last time dynoed 466hp and 6xxNm.
Pulls hard after 3000rpm.
Haven't seen nice wide-range setup on 60x yet. One 603 with GT37v is rumored to have usable rev range of 1500-6000 and +1bar under 2000rpm with no smoke at all. Based on my experiences with 606 and GT37v I'd say it's just not possible. I couldn't get that kind of results even with all the smoke in the world.
Funny thing you mentioned, I'm pretty confident that it will be a E30-606 swap. My previous project was BMW's old M21 diesel (524td). Started out in E28 chassis, later swapped engine into E30. Dynoed 233hp/422Nm. Some videos of my previous and current project can be found in youtube by search word "muuri5".
This post was last modified: 11-17-2009, 01:07 PM by muuris.
Actually jeemu is a friend of mine. Nowadays has just HX40 Super under the bonnet, because compound setup didn't work out and caused a lot of EMP.
Engine has modified internals (head work, larger valves etc). Last time dynoed 466hp and 6xxNm.
Pulls hard after 3000rpm.
Haven't seen nice wide-range setup on 60x yet. One 603 with GT37v is rumored to have usable rev range of 1500-6000 and +1bar under 2000rpm with no smoke at all. Based on my experiences with 606 and GT37v I'd say it's just not possible. I couldn't get that kind of results even with all the smoke in the world.
Funny thing you mentioned, I'm pretty confident that it will be a E30-606 swap. My previous project was BMW's old M21 diesel (524td). Started out in E28 chassis, later swapped engine into E30. Dynoed 233hp/422Nm. Some videos of my previous and current project can be found in youtube by search word "muuri5".
11-17-2009, 01:40 PM
#30
(11-17-2009, 12:54 PM)muuris Haven't seen nice wide-range setup on 60x yet. One 603 with GT37v is rumored to have usable rev range of 1500-6000 and +1bar under 2000rpm with no smoke at all. Based on my experiences with 606 and GT37v I'd say it's just not possible. I couldn't get that kind of results even with all the smoke in the world.
I would agree. Even the 6.xL engines have some bad lag at low-rpms. In the 6.0L ford shop truck I used to drive, I could mash the throttle to the floor and count to 4 before the engine would scream to life.
Even tuned ones have a hard time getting it going. http://www.youtube.com/watch?v=kz6Bn_evP1A
Not much quicker than with a ball bearing turbo: http://www.youtube.com/watch?v=Gn72q_O18GY
Honestly, I dislike the GT35v and GT37v turbos. I don't know what was going through Garrett's engineer's heads when they thought to use airplane wings instead of vanes.
This post was last modified: 11-17-2009, 01:41 PM by ForcedInduction.
ForcedInduction
11-17-2009, 01:40 PM
#30
(11-17-2009, 12:54 PM)muuris Haven't seen nice wide-range setup on 60x yet. One 603 with GT37v is rumored to have usable rev range of 1500-6000 and +1bar under 2000rpm with no smoke at all. Based on my experiences with 606 and GT37v I'd say it's just not possible. I couldn't get that kind of results even with all the smoke in the world.
I would agree. Even the 6.xL engines have some bad lag at low-rpms. In the 6.0L ford shop truck I used to drive, I could mash the throttle to the floor and count to 4 before the engine would scream to life.
Even tuned ones have a hard time getting it going. http://www.youtube.com/watch?v=kz6Bn_evP1A
Not much quicker than with a ball bearing turbo: http://www.youtube.com/watch?v=Gn72q_O18GY
Honestly, I dislike the GT35v and GT37v turbos. I don't know what was going through Garrett's engineer's heads when they thought to use airplane wings instead of vanes.
11-18-2009, 01:01 PM
#31
(11-17-2009, 12:54 PM)muuris Funny thing you mentioned, I'm pretty confident that it will be a E30-606 swap. My previous project was BMW's old M21 diesel (524td). Started out in E28 chassis, later swapped engine into E30. Dynoed 233hp/422Nm. Some videos of my previous and current project can be found in youtube by search word "muuri5".
Oh shit. That was a strong pull from 110-180. It lokked like it had far more power than just 211hp.. Was it smokey? I think it is the combo of a light weight chassi and a manual gearbox that makes the difference. Imagine a 606 in that
. You are not gonna have slushbox in a e30, right? I am afraid it will be a little nose-heavy though.. That is okay for normal street driving, but will require a special tune on the front suspension if you want to take it to the racecircuit. Looking forward to see haw this develops!!!!!!
(11-17-2009, 12:54 PM)muuris Funny thing you mentioned, I'm pretty confident that it will be a E30-606 swap. My previous project was BMW's old M21 diesel (524td). Started out in E28 chassis, later swapped engine into E30. Dynoed 233hp/422Nm. Some videos of my previous and current project can be found in youtube by search word "muuri5".
Oh shit. That was a strong pull from 110-180. It lokked like it had far more power than just 211hp.. Was it smokey? I think it is the combo of a light weight chassi and a manual gearbox that makes the difference. Imagine a 606 in that
. You are not gonna have slushbox in a e30, right? I am afraid it will be a little nose-heavy though.. That is okay for normal street driving, but will require a special tune on the front suspension if you want to take it to the racecircuit. Looking forward to see haw this develops!!!!!!
11-18-2009, 03:58 PM
#32
(11-18-2009, 01:01 PM)lars Oh shit. That was a strong pull from 110-180. It lokked like it had far more power than just 211hp.. Was it smokey? I think it is the combo of a light weight chassi and a manual gearbox that makes the difference. Imagine a 606 in that
In the video the car was with the same settings as in the dyno (233hp) and wasn't very smokey. That time I had problems with IP (self-built), the distributor head started floating after 3800rpm with stock springs and torque went down. Shimmed the springs quite a bit and after that it pulled strong over 5500 (as seen on other video). With stock non-LSD and 185/65/14" tires it ran 14,8sec and 157km/h (about 1400kg with driver).
Starting to get really OT, but I've planned to modify the E30 firewall to be able to get the engine as far back as possible with reasonable amount of work. Not going to be a serious race car, just for fun.
(11-18-2009, 01:01 PM)lars Oh shit. That was a strong pull from 110-180. It lokked like it had far more power than just 211hp.. Was it smokey? I think it is the combo of a light weight chassi and a manual gearbox that makes the difference. Imagine a 606 in that
In the video the car was with the same settings as in the dyno (233hp) and wasn't very smokey. That time I had problems with IP (self-built), the distributor head started floating after 3800rpm with stock springs and torque went down. Shimmed the springs quite a bit and after that it pulled strong over 5500 (as seen on other video). With stock non-LSD and 185/65/14" tires it ran 14,8sec and 157km/h (about 1400kg with driver).
Starting to get really OT, but I've planned to modify the E30 firewall to be able to get the engine as far back as possible with reasonable amount of work. Not going to be a serious race car, just for fun.
joshyak
Unregistered
04-09-2010, 11:47 PM
#33
Turboedmav
Unregistered
04-21-2010, 05:16 AM
#34
I'm still droolling over this project... what's the current status?
I can't wait to see what the end package looks like...
I'd be very interested in buying one of these control units from u... or if you're feeling generous
perhaps you'll want to share a component list and the code too in good will? maybe?... lol ...Whichever way you choose to go, I'll understand you've put hard work into it... but I definitely want one of these units for my own turbo project man!!! please let us know about your latest progress!!!
Turboedmav
04-21-2010, 05:16 AM #34Hey Oel?
I'm still droolling over this project... what's the current status?
I can't wait to see what the end package looks like...
I'd be very interested in buying one of these control units from u... or if you're feeling generous perhaps you'll want to share a component list and the code too in good will? maybe?... lol ...
Whichever way you choose to go, I'll understand you've put hard work into it... but I definitely want one of these units for my own turbo project man!!! please let us know about your latest progress!!!
12-12-2013, 02:21 PM
#35
(05-12-2009, 05:14 PM)oel_brenner I have assembled most of the parts needed to build the control system for the Holset HE351VE on my OM603.
what I plan to use:
Arduino microcontroller w/AT328
http://arduino.cc/en/Main/ArduinoBoardDuemilanove
.....
I have the hardware pretty much sorted, I plan to use the IP rack position sensor
for "throttle" position, an EGT input, and possibly RPM input and maybe others.
static relationship between rack position and amount of boost delivered
......
or a non-linear relationship with a curve matching boost with rack position
based on rack position to amount of boost required to burn that much
fuel efficently
this will probably have to be arrived at with a trial and error process, but the great thing
is that it will only require software changes, and can be done rather quickly.
.....
possibly a static boost control knob that moves nozzle ignoring throttle position signal
for testing/tuning/data collection etc..
full throttle triggers that could :
disengage A/C compressor
engage max cooling fan for A/W intercooler
other ?
any other ideas / feedback ?
Yes, I have some feedback. Good Project, I will follow.
I have been trying to get usable information from IP rack position sensor, got nothing suitable to be used as indicator of throttle position.
My objective is to generate a signal representing real time fuel consumption for feeding a trip computer.
I think failure of IP fuel rack position sensor is due to the fact that EDS (ECU) in Mercedes OM603 (and others similar) only controls idle speed.
Based on Mercedes given info, load is completely controled via throttle pedal.
Due to this, I am looking to test suitability of the cruise control position feedback signal. In case it does not work, or work only with cruise engaged, I would consider to install a throttle position sensor and generate a new signal.
Reading futher down, I discovered you got to the same conclusion.
I wish you success with this. Same wish I to myself
Oldbeaver
PS: I attached to scope graphs of rpm vs IP rack position sensors at 700 and 2500 rpm (a lot different load) and waves stay stable.
Attached Files
Image(s)(05-12-2009, 05:14 PM)oel_brenner I have assembled most of the parts needed to build the control system for the Holset HE351VE on my OM603.
what I plan to use:
Arduino microcontroller w/AT328
http://arduino.cc/en/Main/ArduinoBoardDuemilanove
.....
I have the hardware pretty much sorted, I plan to use the IP rack position sensor
for "throttle" position, an EGT input, and possibly RPM input and maybe others.
static relationship between rack position and amount of boost delivered
......
or a non-linear relationship with a curve matching boost with rack position
based on rack position to amount of boost required to burn that much
fuel efficently
this will probably have to be arrived at with a trial and error process, but the great thing
is that it will only require software changes, and can be done rather quickly.
.....
possibly a static boost control knob that moves nozzle ignoring throttle position signal
for testing/tuning/data collection etc..
full throttle triggers that could :
disengage A/C compressor
engage max cooling fan for A/W intercooler
other ?
any other ideas / feedback ?
Yes, I have some feedback. Good Project, I will follow.
I have been trying to get usable information from IP rack position sensor, got nothing suitable to be used as indicator of throttle position.
My objective is to generate a signal representing real time fuel consumption for feeding a trip computer.
I think failure of IP fuel rack position sensor is due to the fact that EDS (ECU) in Mercedes OM603 (and others similar) only controls idle speed.
Based on Mercedes given info, load is completely controled via throttle pedal.
Due to this, I am looking to test suitability of the cruise control position feedback signal. In case it does not work, or work only with cruise engaged, I would consider to install a throttle position sensor and generate a new signal.
Reading futher down, I discovered you got to the same conclusion.
I wish you success with this. Same wish I to myself
Oldbeaver
PS: I attached to scope graphs of rpm vs IP rack position sensors at 700 and 2500 rpm (a lot different load) and waves stay stable.
Attached Files
Image(s)
12-12-2013, 03:46 PM
#36
(12-12-2013, 02:21 PM)oldbeaver PS: I attached to scope graphs of rpm vs IP rack position sensors at 700 and 2500 rpm (a lot different load) and waves stay stable.
When you say "rpm vs IP rack position sensors", you're not saying that one of the waveforms is rpm, you're saying both waveforms are from the RPS, and that one screenshot is 700 rpm and the other is 2500 rpm, right? Is this voltage at 1V/division? Did you change the load a lot on the 2500rpm screenshot, but not the 700rpm shot? Sorry for all the questions, I'm trying to make sense of what I'm looking at.
To my knowledge, the rack position sensor was on U.S. market OM617's in model year 1985. This engine didn't have electronic idle control. It's my understanding that rack position (load) is used to control EGR.
It would be incorrect to say that load is completely controlled by throttle position. It should be something more like throttle position+governor+alda=rack position
raysorenson
12-12-2013, 03:46 PM #36(12-12-2013, 02:21 PM)oldbeaver PS: I attached to scope graphs of rpm vs IP rack position sensors at 700 and 2500 rpm (a lot different load) and waves stay stable.
When you say "rpm vs IP rack position sensors", you're not saying that one of the waveforms is rpm, you're saying both waveforms are from the RPS, and that one screenshot is 700 rpm and the other is 2500 rpm, right? Is this voltage at 1V/division? Did you change the load a lot on the 2500rpm screenshot, but not the 700rpm shot? Sorry for all the questions, I'm trying to make sense of what I'm looking at.
To my knowledge, the rack position sensor was on U.S. market OM617's in model year 1985. This engine didn't have electronic idle control. It's my understanding that rack position (load) is used to control EGR.
It would be incorrect to say that load is completely controlled by throttle position. It should be something more like throttle position+governor+alda=rack position
12-12-2013, 05:59 PM
#38
(12-12-2013, 03:46 PM)raysorenson(12-12-2013, 02:21 PM)oldbeaver PS: I attached to scope graphs of rpm vs IP rack position sensors at 700 and 2500 rpm (a lot different load) and waves stay stable.
When you say "rpm vs IP rack position sensors", you're not saying that one of the waveforms is rpm, you're saying both waveforms are from the RPS, and that one screenshot is 700 rpm and the other is 2500 rpm, right? Is this voltage at 1V/division? Did you change the load a lot on the 2500rpm screenshot, but not the 700rpm shot? Sorry for all the questions, I'm trying to make sense of what I'm looking at.
To my knowledge, the rack position sensor was on U.S. market OM617's in model year 1985. This engine didn't have electronic idle control. It's my understanding that rack position (load) is used to control EGR.
It would be incorrect to say that load is completely controlled by throttle position. It should be something more like throttle position+governor+alda=rack position
Thank you for replying, Ray.
You wrote: "When you say "rpm vs IP rack position sensors", you're not saying that one of the waveforms is rpm, you're saying both waveforms are from the RPS, and that one screenshot is 700 rpm and the other is 2500 rpm, right?"
In fact it is rpm vs one of the three pulses (wires) coming from the rack position sensor. I did it with all three against rpm.
My car wasn´t sold in the US, but in Japan. 1993.
You also wrote:
"It would be incorrect to say that load is completely controlled by throttle position. It should be something more like throttle position+governor+alda=rack position".
Yes, throttle pedal send the command signal, and Alda modulates it, and ECU modulates it too (idle), so both participate somehow.
I also have scope graphs between two IP rack position signals, and they don´t make sense to me at all. They variate against rpm, but they dont between each other. I even save videos of several seconds of them, with the car in movement, starting from still. I may send them to you or publish them here, if posible.
Divisions don´t mean volts, but they are relative measures only. I use an AudioScope and never used a real Scope before. It was hard to me to put signals in the screen so one can see wave shapes.
Oldbeaver
This post was last modified: 12-12-2013, 06:05 PM by oldbeaver.
(12-12-2013, 03:46 PM)raysorenson(12-12-2013, 02:21 PM)oldbeaver PS: I attached to scope graphs of rpm vs IP rack position sensors at 700 and 2500 rpm (a lot different load) and waves stay stable.
When you say "rpm vs IP rack position sensors", you're not saying that one of the waveforms is rpm, you're saying both waveforms are from the RPS, and that one screenshot is 700 rpm and the other is 2500 rpm, right? Is this voltage at 1V/division? Did you change the load a lot on the 2500rpm screenshot, but not the 700rpm shot? Sorry for all the questions, I'm trying to make sense of what I'm looking at.
To my knowledge, the rack position sensor was on U.S. market OM617's in model year 1985. This engine didn't have electronic idle control. It's my understanding that rack position (load) is used to control EGR.
It would be incorrect to say that load is completely controlled by throttle position. It should be something more like throttle position+governor+alda=rack position
Thank you for replying, Ray.
You wrote: "When you say "rpm vs IP rack position sensors", you're not saying that one of the waveforms is rpm, you're saying both waveforms are from the RPS, and that one screenshot is 700 rpm and the other is 2500 rpm, right?"
In fact it is rpm vs one of the three pulses (wires) coming from the rack position sensor. I did it with all three against rpm.
My car wasn´t sold in the US, but in Japan. 1993.
You also wrote:
"It would be incorrect to say that load is completely controlled by throttle position. It should be something more like throttle position+governor+alda=rack position".
Yes, throttle pedal send the command signal, and Alda modulates it, and ECU modulates it too (idle), so both participate somehow.
I also have scope graphs between two IP rack position signals, and they don´t make sense to me at all. They variate against rpm, but they dont between each other. I even save videos of several seconds of them, with the car in movement, starting from still. I may send them to you or publish them here, if posible.
Divisions don´t mean volts, but they are relative measures only. I use an AudioScope and never used a real Scope before. It was hard to me to put signals in the screen so one can see wave shapes.
Oldbeaver
12-12-2013, 08:09 PM
#39
For a load signal it would be easiest by far to fit a throttle position sensor to the throttle shaft on the outside of the injection pump.
Baldur Gislason
The red signal is obviously RPM. The green signal looks like a high frequency signal that the ECU feeds into the rack position sensor to measure its inductance. It's running at a frequency higher than your sound card can sample, that's why it looks strange, you have aliases in the sampled signal. You need a real scope to see what it really looks like, but that won't do much good because the interesting part really is the current and not the voltage if I understand the operation of this particular ECU correctly.
For a load signal it would be easiest by far to fit a throttle position sensor to the throttle shaft on the outside of the injection pump.
Baldur Gislason
12-17-2013, 02:01 PM
#42
(12-12-2013, 08:09 PM)baldur The red signal is obviously RPM. The green signal looks like a high frequency signal that the ECU feeds into the rack position sensor to measure its inductance. It's running at a frequency higher than your sound card can sample, that's why it looks strange, you have aliases in the sampled signal. You need a real scope to see what it really looks like, but that won't do much good because the interesting part really is the current and not the voltage if I understand the operation of this particular ECU correctly.
For a load signal it would be easiest by far to fit a throttle position sensor to the throttle shaft on the outside of the injection pump.
It's the time it takes to charge the coil in the sensor. That's the rack position for the ecu.
The further a piece of iron is inserted in the coil, the longer it takes to charge the coil which translates to rack position.
(12-12-2013, 08:09 PM)baldur The red signal is obviously RPM. The green signal looks like a high frequency signal that the ECU feeds into the rack position sensor to measure its inductance. It's running at a frequency higher than your sound card can sample, that's why it looks strange, you have aliases in the sampled signal. You need a real scope to see what it really looks like, but that won't do much good because the interesting part really is the current and not the voltage if I understand the operation of this particular ECU correctly.
For a load signal it would be easiest by far to fit a throttle position sensor to the throttle shaft on the outside of the injection pump.
It's the time it takes to charge the coil in the sensor. That's the rack position for the ecu.
The further a piece of iron is inserted in the coil, the longer it takes to charge the coil which translates to rack position.
12-18-2013, 06:55 PM
#43
and each "node" ( the Holset controller is a node ) communicates back to the main system. The master system does not just send a signal to open
and close the VNT it sends full message packets in CANbus protocol. Just too much work for me right now.. someday when I own a OM606
and have a LOT more time I might figure out how to retrofit a native CAN turbo controller.
The Hella CAN bus actuator is easy to control if you need help just ask.
The CAN system is way more complex then just sending PWM signals on the data wires.. its a whole system with a master controller
and each "node" ( the Holset controller is a node ) communicates back to the main system. The master system does not just send a signal to open
and close the VNT it sends full message packets in CANbus protocol. Just too much work for me right now.. someday when I own a OM606
and have a LOT more time I might figure out how to retrofit a native CAN turbo controller.
The Hella CAN bus actuator is easy to control if you need help just ask.
12-18-2013, 07:08 PM
#44
(12-18-2013, 06:55 PM)pemtek The CAN system is way more complex then just sending PWM signals on the data wires.. its a whole system with a master controllerThere is a canbus library for the arduino finally... Even a shield it looks like.
and each "node" ( the Holset controller is a node ) communicates back to the main system. The master system does not just send a signal to open
and close the VNT it sends full message packets in CANbus protocol. Just too much work for me right now.. someday when I own a OM606
and have a LOT more time I might figure out how to retrofit a native CAN turbo controller.
The Hella CAN bus actuator is easy to control if you need help just ask.
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
(12-18-2013, 06:55 PM)pemtek The CAN system is way more complex then just sending PWM signals on the data wires.. its a whole system with a master controllerThere is a canbus library for the arduino finally... Even a shield it looks like.
and each "node" ( the Holset controller is a node ) communicates back to the main system. The master system does not just send a signal to open
and close the VNT it sends full message packets in CANbus protocol. Just too much work for me right now.. someday when I own a OM606
and have a LOT more time I might figure out how to retrofit a native CAN turbo controller.
The Hella CAN bus actuator is easy to control if you need help just ask.
1987 300D Sturmmachine
1991 300D Nearly Perfect
1985 300D Weekend/Camping/Dog car
1974 L508D Motoroam Monarch "NightMare"
OBK #42
12-22-2013, 12:51 PM
#45
Hi yes the code is pretty simple on the Arduino. Five minutes of modifying the code that comes as an example will have the actuator moving http://www.hobbytronics.co.uk/can-bus-shield if you can do basic coding.
12-23-2013, 06:37 AM
#46
(12-18-2013, 06:55 PM)pemtek The CAN system is way more complex then just sending PWM signals on the data wires.. its a whole system with a master controller
and each "node" ( the Holset controller is a node ) communicates back to the main system. The master system does not just send a signal to open
and close the VNT it sends full message packets in CANbus protocol. Just too much work for me right now.. someday when I own a OM606
and have a LOT more time I might figure out how to retrofit a native CAN turbo controller.
The Hella CAN bus actuator is easy to control if you need help just ask.
Do you know the CAN addresses to control the Holset VGT?
Baldur Gislason
(12-18-2013, 06:55 PM)pemtek The CAN system is way more complex then just sending PWM signals on the data wires.. its a whole system with a master controller
and each "node" ( the Holset controller is a node ) communicates back to the main system. The master system does not just send a signal to open
and close the VNT it sends full message packets in CANbus protocol. Just too much work for me right now.. someday when I own a OM606
and have a LOT more time I might figure out how to retrofit a native CAN turbo controller.
The Hella CAN bus actuator is easy to control if you need help just ask.
Do you know the CAN addresses to control the Holset VGT?
Baldur Gislason
12-24-2013, 09:20 AM
#48
(12-24-2013, 08:30 AM)pemtek Sorry I do not, I know the Hella ones have different addresses dependent on the part number.
Would not be difficult to work out using a can bus analzer device like the one from Microchip.
What is the Holsett VGT turbo from?
Newer Dodge trucks have the HE351VE, people are always cranking those up, good used turbos are pretty easy to find.
Newbie-read this: Cheap Tricks
617.952-220k-Getting built up
larsalan I guess I need to look at this stupid ass drip shit. What you have to like mess with those elements on the pump? What a fucking hassle. then use some wire to hold the throttle open or some shit?
Simpler=Better
12-24-2013, 09:20 AM #48(12-24-2013, 08:30 AM)pemtek Sorry I do not, I know the Hella ones have different addresses dependent on the part number.
Would not be difficult to work out using a can bus analzer device like the one from Microchip.
What is the Holsett VGT turbo from?
Newer Dodge trucks have the HE351VE, people are always cranking those up, good used turbos are pretty easy to find.
Newbie-read this: Cheap Tricks
617.952-220k-Getting built up
larsalan I guess I need to look at this stupid ass drip shit. What you have to like mess with those elements on the pump? What a fucking hassle. then use some wire to hold the throttle open or some shit?
Users browsing this thread:
28 Guest(s)
28 Guest(s)
Users browsing this thread:
28 Guest(s)
28 Guest(s)