The Top DSM Community on the Web

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. Log in to remove most ads.

Please Support Morrison Fabrications
Please Support Morrison Fabrication

Holset HE351VE Installed in My DSM (pics)

This site may earn a commission from merchant
affiliate links, including eBay, Amazon, and others.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

kmetiuk

10+ Year Contributor
521
125
Mar 10, 2010
Edmonton, AB_Canada
As far as I know I have the 1st ever Holset HE351VE installed in my DSM with a working electronic controller. Peeps have installed this turbo sans controller....but I'm prettty sure I'm the only one that has it reading sensors and using a algorithim to set where the vanes should be AND outputting a signal to DSMLink that you can then log which makes tuning and playing with the maps in the controller a breeze.

Spool is GREAT! Huge 24 cm (1.85 AR) housing means ZERO choking out the exhaust.....back pressure is low low low. It's crazy when its in quick spool mode...spooling everywhere. When in traffic you can hear it spooling up with just the lightest of taps on the throttle. SOOOOOO much fun ripping around. B/C it has that holset whistle too the reaction I get from trucks and shiiit is hilarious. I was ripping around a few days ago and took it out on the highway and got a huge horn blast from a tractor trailer driver and a devil horns out the window. Apparently you CAN hear that "Holset turbo whistle" from inside other vehicles. He fcuking loved it.

:hellyeah::pray: :applause: :sneaky:


Without the controller it spools around 6000 RPM in 3rd gear (starting from 2500 RPM)....with it I am can make it spool at 3800 RPM but it makes it ride that surge line a lil too closely so I backed it off to 4000 RPM. Without controller it wont spool in 1st and in 2nd it will but at 7000-7500 RPM....with it im spooling at 5800 in 1st and 4500 in 2nd. Basically if your over 4000 RPM in 2nd and higher gears you have full spool in less than 1 second....0.88s according to link actually.

I am currently writing the an application for the controller so that you can plug you PC in to it and set all the vane maps with your PC. Once I am done that I am going to post a new thread with ALL the data from LINK, screen shots of the APP, and lots of pics and videos of it in action.

Here's a few teaser pics of it installed in my car.....its a HUGE bi***! :D

You must be logged in to view this image or video.


You must be logged in to view this image or video.


You must be logged in to view this image or video.


You must be logged in to view this image or video.
 
I think you have a lot to learn about radio and flow turbines. A large nozzle area does not imply no back pressure. as far as I can tell you've not even got listen to the area where it would be beneficial over a smaller turbo
 
I think you have a lot to learn about radio and flow turbines.


Not to start a flame warm, but man your posts are always so salty....I see it frequently in your reply's to others. Take it down a notch and stop being so condescending. I am sure you know a lot but you come across as an azz when you come in to a thread swinging like that. Haters gonna hate I guess so it just means I'm onto something here so I'll take it as a compliment. :)




It would also help if you actually read the post first and maybe absorbed or thought about it a bit b/c you got 2 things wrong:

A large nozzle area does not imply no back pressure.

1st: I didn't say no back pressure....I said low low. Peeps apparently have issues with back pressure in the 0.55 AR on the HX35. You think my back pressure is the same or nearly the same at 30 psi and 8000 RPM on a 1.85 AR vs a 0.55 AR? Clearly you didn't read the post fully or thought about it b/c clearly a larger AR will have lower back pressure which I think you'd agree with.

You're also not the first person to have a electronically controlled variable geometry turbo charger on the DSM mine is actually controlled by a 1GECU

2nd: I didn't say I was the first elec controlled variable....I said the first working elec controlled HE351VE installed in a DSM. Aero installed a HE351VE in a DSM but it was mech controlled (there are others who have done that too). There are also the Fleece controller for Dodges, Capsurocks controller (which I dont believe has installed in a dsm as its $$), and there is one guy on here that installed the HE351VE with a "lil black box" (designed for dodges but he's attempting to get working on his DSM). As far as I know (or can find) he's still fighting with his controller as the lil black box uses only the turbo RPM sensor and TPS to control the vanes which is totally ineffective for quick spooling it up.

The Holset electronic actuator (not a mech type like I believe the VNT is) has taken the community a long time to crack and regardless of wether you agree or not there are some props due there. Took a member on here sniffing the CANBUS codes going to the actuator to figure it out....he is the true hero here. I would not have my controller with out the proper J1939 protocols he sniffed out. If you looked at the amount of coding (i'm at 2200 lines and counting) and the various microchips it takes to run this thing properly and stably you'd realize this has been quite a challenge and absolutely cannot be controlled by the DSM ECU or any other ECU other than the Dodge one.

It's pretty awesome your controlling your GT3788 VNT with a 1G ECU and I'd love to see a thread on that and how your doing it. I don't know much about it, but from a quick google search it appears it uses a wastegate style mech actuator which is not electronic control it is actually electric over pneumatic control but I'm sure you knew that

Taking a stab at it I would guess you are controlling it by using a duty cycle with a solenoid hooked up to a vac/boost source similar to how link controls the boost with the BCS correct? Still neat and I'd truly love to see more if you got time to start a thread on it.




As far as I can tell you've not even got listen to the area where it would be beneficial over a smaller turbo

Ummmm.....what? I can't decipher what you mean here. What small turbo?
 
Not to start a flame warm, but man your posts are always so salty....I see it frequently in your reply's to others. Take it down a notch and stop being so condescending. I am sure you know a lot but you come across as an azz when you come in to a thread swinging like that. Haters gonna hate I guess so it just means I'm onto something here so I'll take it as a compliment. :)

Clearly you cannot tell the difference between constructive criticism and hate. I feel really bad for you, your in for a tough journey in life. Honestly, your lucky the old timers like defiant are note still around. I don't post a thing unless I've been there done that, and in this case I have. I've looked at your posts, and I can easily see the "I just finished ME school I know everything attitude". Let me tell you, having a b.s. In ME doesn't mean shit. I get it, you worked hard, solved a lot of problems. You should be proud, it's far harder than most other paths. But it's not an instant path to enlightenment of all things motorsport - In fact it should be a very humbling experience of what you don't know.

As for my comments on "lot to learn on radial inflow turbines" Yeah I stand by it. A typical undergrad ME student gets a 1 semester course on gas turbines - as an elective, you might also have gas dynamics as an elective. How do two semesters of school qualify you as an expert? I'm far beyond that, and I certainly have lots to learn on the subject. Did you design turbines for years before this?

It would also help if you actually read the post first and maybe absorbed or thought about it a bit b/c you got 2 things wrong:
Here we go let's start splitting hairs.


1st: I didn't say no back pressure....I said low low. Peeps apparently have issues with back pressure in the 0.55 AR on the HX35. You think my back pressure is the same or nearly the same at 30 psi and 8000 RPM on a 1.85 AR vs a 0.55 AR? Clearly you didn't read the post fully or thought about it b/c clearly a larger AR will have lower back pressure which I think you'd agree with.

Ok you got me, I said "no" when I was implying "low low".

Here's the problem with your theory, turbine efficiency isn't fixed, it very much depends on pressure ratio, vane geometry, rotor speed, inlet conditions, outlet conditions, and a whole slew of other things. It's very possible that by opening up the nozzle all the way that the efficiency tanks so far that it will still take a huge pressure ratio to make enough shaft power to run the compressor. However whats far more likely is that at large openings the inlet enthalpy will be far too low to generate enough shaft power to run the compressor. Furthermore, have you figured the area of the turbine exducer? It's probably on the order of 14cm^2. Most performance turbo's that have low back pressure have an exducer area about 2x the nozzle area.

Lastly are you aware what goes on when you start pushing a turbo towards choke flow? The efficiency tanks on both sides, Surely you know the power equations for an isentropic compressor and turbine, and you can see that if it was unwastegated and 100% efficient, the compressor pressure ratio is proportional to the turbine pressure ratio. I hope you can understand what that implies. A real eye opener was a pm with kevin jewer, on his super99 which is clearly designed for flow and power, he still can near a 2:1 exhaust to intake pressure ratio. Even on the compound setups there is still plenty of back-pressure when running the turbo's in the efficiency islands, and with inter-stage cooling. A giant nozzle isn't a golden ticket to low backpressure, sure it can help, but it isn't all there is to it.

For example on the gt3788anvt, the vanes rotate and actually change the angle of attack on the turbine. Opening vanes all the way kills the turbine efficiency by taking away the tangential component of the gas velocity. Here's the net result, boost pressure drops and the exhaust to intake pressure ratio goes through the roof even though the nozzle area is HUGE. In fact the sweet spot is right about 50% open, by 60% it's heading downhill fast. I'm at least assuming tanking turbine efficiency is what is causing this behavior. It could be a loss of sonic conditions in the nozzle area. Who knows.

I do realize that the vanes are fixed in your turbo, and it should not have as much of an impact on the turbine efficiency as the garret, but there are still effects.


2nd: I didn't say I was the first elec controlled variable....I said the first working elec controlled HE351VE installed in a DSM. Aero installed a HE351VE in a DSM but it was mech controlled (there are others who have done that too). There are also the Fleece controller for Dodges, Capsurocks controller (which I dont believe has installed in a dsm as its $$), and there is one guy on here that installed the HE351VE with a "lil black box" (designed for dodges but he's attempting to get working on his DSM). As far as I know (or can find) he's still fighting with his controller as the lil black box uses only the turbo RPM sensor and TPS to control the vanes which is totally ineffective for quick spooling it up.

The Holset electronic actuator (not a mech type like I believe the VNT is) has taken the community a long time to crack and regardless of wether you agree or not there are some props due there. Took a member on here sniffing the CANBUS codes going to the actuator to figure it out....he is the true hero here. I would not have my controller with out the proper J1939 protocols he sniffed out. If you looked at the amount of coding (i'm at 2200 lines and counting) and the various microchips it takes to run this thing properly and stably you'd realize this has been quite a challenge and absolutely cannot be controlled by the DSM ECU or any other ECU other than the Dodge one.

My bad. The point was there are others out there, probably more outside our platform. You can spare me the lecture on CAN and J1939, I know plenty on that. It needs like 2 chips to run it, a micro controller and a can transceiver. The hardest part of it would be sniffing out the can commands and understanding that. I'm assuming the member here is hakcenter. So he basically did the hard work and your putting together a PID controller I hope?

As for controlling by the dsm ecu? Absolutely it could be done. Create a serial to can convertor, and send data out of the ecu using the aldl. The converter sends a byte to the ecu requesting a vane poistion, the ecu responds with a byte of vane poisition. Not that big of a deal, you would be hijacking the diagnostic command section of the serial comms. It could potentially interfere with data logging, but there most definatly are work arounds. I'm currently working in this area, trying to put together better logging. Dumping the request-reply format, for a request-reply with like 10 channels at once, giving a massive speed up in logging.

It's pretty awesome your controlling your GT3788 VNT with a 1G ECU and I'd love to see a thread on that and how your doing it. I don't know much about it, but from a quick google search it appears it uses a wastegate style mech actuator which is not electronic control it is actually electric over pneumatic control but I'm sure you knew that

Taking a stab at it I would guess you are controlling it by using a duty cycle with a solenoid hooked up to a vac/boost source similar to how link controls the boost with the BCS correct? Still neat and I'd truly love to see more if you got time to start a thread on it.

Lol no, it's a hydraulic servo. It does use pwm, but the pwm commands a postion - much the same as a servo in a model airplane. Far more control than a solenoid and air pressure, there is a hydraulic feedback loop inside the controls. It's really very elegant. I can also send vane position for feedback to the ecu to compenstate for drift, but so far it's been a non issue. It goes where I command it to.



Ummmm.....what? I can't decipher what you mean here. What small turbo?

Yeah I was using my phone talk to text feature. Anyway, you haven't even pushed into territory that you see any benefit from what you are doing, or even see if it is a benefit. What I'm getting at is yeah it runs, yeah it works, but you have not done enough with it to even begin to understand it. - that is how the flow and back pressure are going to be with it cranked up.

Like I said you have a lot to learn yet. I've had mine on for over a year, and it's been on kill the whole time. The car makes near 700whp, and I learn new things everytime I take it out. When I first started it had a odd ball compressor, 57mm ind 88mm exducer. With that wheel it would bury a 4bar map at 4000rpm in 3rd gear. 3000 in 5th. It's laggier now with a 63.5mm wheel, but still very very responsive for such a big turbo.
 
[...]Aero installed a HE351VE in a DSM but it was mech controlled (there are others who have done that too). There are also the Fleece controller for Dodges, Capsurocks controller (which I dont believe has installed in a dsm as its $$), and there is one guy on here that installed the HE351VE with a "lil black box" (designed for dodges but he's attempting to get working on his DSM). As far as I know (or can find) he's still fighting with his controller as the lil black box uses only the turbo RPM sensor and TPS to control the vanes which is totally ineffective for quick spooling it up.

The Holset electronic actuator (not a mech type like I believe the VNT is) has taken the community a long time to crack and regardless of wether you agree or not there are some props due there. Took a member on here sniffing the CANBUS codes going to the actuator to figure it out....he is the true hero here. I would not have my controller with out the proper J1939 protocols he sniffed out. If you looked at the amount of coding (i'm at 2200 lines and counting) and the various microchips it takes to run this thing properly and stably you'd realize this has been quite a challenge and absolutely cannot be controlled by the DSM ECU or any other ECU other than the Dodge one.

So I guess I'm the guy that's "fighting" with my controller ? LOL. My full blown controller has Bluetooth and an android tuning app. But good job at the stab mister I don't want to start a flame war.

If anyone did any of the hard work getting this turbo on other vehicles it was Steed and Farm828. I've gotten the turbo out of stupid mode and integrated the shaft speed sensor, which I suggest you actually look at before you think what you're doing is going to last. The rotor is rated to 130k and the controller is more than capable of pushing it into 160k land.

But really, 2200 lines of code ? My arduino control is at 810, and 310 of it is just the serial interface for user control, and another 100 lines for variable setup. Set and forget is 500 lines tops. What is taking up so much space ?
 
So I guess I'm the guy that's "fighting" with my controller ? LOL. My full blown controller has Bluetooth and an android tuning app. But good job at the stab mister I don't want to start a flame war.

If anyone did any of the hard work getting this turbo on other vehicles it was Steed and Farm828. I've gotten the turbo out of stupid mode and integrated the shaft speed sensor, which I suggest you actually look at before you think what you're doing is going to last. The rotor is rated to 130k and the controller is more than capable of pushing it into 160k land.

But really, 2200 lines of code ? My arduino control is at 810, and 310 of it is just the serial interface for user control, and another 100 lines for variable setup. Set and forget is 500 lines tops. What is taking up so much space ?

You created the lil black box did you not? Props to you sir. One of the few who have created a controller from the ground up. :)

The guy who installed it and was fighting with it was called "EvoAuto" on the lil black box forum....I dunno what his name here is tho. I read through his whole thread there and kept up to date on it as I was curious about results on a 4G63 as mine was not finished at the time. As far as I know he has not finished nor posted results. http://community.lilbb.com/general-discussion/he351ve-in-a-4g63/60/

So, no your not the guy I was talking about... I didn't say @hakcenter , @bastarddsm did. And it was mos def not a stab as your not even the right guy. I actually love what you did with the frequency code to watch shaft speed. I had never thought or tried measuring an RPM with an Arduino so it was cool to read through your code and learn. Didn't know you had an Android app either....i just checked it out. Super super cool man. That is a good idea for sure.

I believe the guy who originally sniffed the CAN messages and figured out the protocol and posted it here was @ILiveForBoost. He's the one that I got the original info from and was the one I was giving mad props too. Also I believe that Steed created the Banshee and never did share what the CAN messages were in the end. Not sure who Farm828 is.

I do plan on using the shaft speed sensor I just ran out of inputs on my current micro controller so I need to finish the PC app so I can drop the clunky joystick and LCD. But ya, I am worried about shaft speed....i have an extra turbo so even if it does go i have a backup. Not a great plan at all, but I'll fix that pretty quick here.

I have seen your code before....haven't looked at it in a while tho. I believe you run off purely shaft RPM and then set the vanes at position XX correct?

I am using a more sensors and currently it has only been running on my DSM. Using those sensor my algorithm is considerably more complex than looking at shaft speed and setting the vanes at position XX. I also have a whole menu system, different modes, and allow you to save maps and reset them, etc,etc.....that takes quite a bit of room.

Also I am running a lot of math and a lot of variables and I have a several huge user editable 3-D maps (like the SD table in LINK) so that I have maximum tunability. It currently rides the 75% level on the flash memory for the variables and about 90% of the EEPROM is used up too. 2200 lines is just my main program, that does not include any of the libraries. Once I finish the PC app I will be able to purge a lot of the code as some is for LCD's, joystick, and some other peripherals that will be phased out.
 
Were your ears burning? LOL.

I've gotten the turbo out of stupid mode and integrated the shaft speed sensor, which I suggest you actually look at before you think what you're doing is going to last. The rotor is rated to 130k and the controller is more than capable of pushing it into 160k land.

Are they really that sensitive to rotor speed? What is the speed sensor like? 2 flats on the shaft? is it inductive or hall effect? I've toyed with the idea of the holset, as I like the nozzle better, but the damn actuator makes it a pain to package in a DSM.
 
Are they really that sensitive to rotor speed? What is the speed sensor like? 2 flats on the shaft? is it inductive or hall effect? I've toyed with the idea of the holset, as I like the nozzle better, but the damn actuator makes it a pain to package in a DSM.

If you look at my pics it was a biiitch to get in there.....I have zero room for a rad so I tried out the Scirrocco style rad. But it cannot keep up to more than 2-3 minutes of abuse. Not having a turbo/mani blanket or down pipe wrap is adding waaay to much heat so I'm hoping once I have those in it'll solve the heat issue.

Definitely some hurdles to overcome packaging wise but not impossible as I have one installed and working now. :)
 
EvoAuto fell off the planet. It is flat you know ? His main issue is trying to control boost with rpm. It's never really going to work out and I've touched on this subject on my forum. There's really no point to putting this giant turbo on a vehicle then not go for max flow. If it flows too much and you need to lower boost pick a smaller turbo.

130k is Holsets recommended max, I've seen the videos of them over speeding and grenading. Probably why a lot of high power Cummins break these shafts IMHO.

Bastard, the he351ve shaft has a single slot in it and a VR magnetic pickup. Very simple to count.

The whole point to using rpm over other inputs is that it is indicative of load which makes setting curves super easy. Its caveat is of course is latency but it works in your favor because you start tighter and open up so it just spools faster.

A few others use my code and add other inputs which work out really well. Closing up on tps increase etc.

The cool part about my LBB is that it is not vehicle specific and super easy to install.
 
Ops I lied....I know who Farm828 is now....I actually linked his thread in Ilive for boost after running into some issues initially.

I think Farm828 may have taken the CAN code from @ILiveForBoost 's thread as iLive posted his code in Feb/14 and Farm828 started his thread an April/14. Maybe not tho, not like it really matters....way back then I was searching for solutions to talk to the actuator non stop and ILive was the first one that I found who posted the full meal deal for the CAN.


There's really no point to putting this giant turbo on a vehicle then not go for max flow.
The whole point to using rpm over other inputs is that it is indicative of load which makes setting curves super easy.

Maybe in a diesel truck you want max flow as you are looking for raw pulling power but in a gas car that is going for as long a power band as possible you need boost as early as possible. The holy grail for DSM turbo's (in my humble opinion) is a turbo that spools up as fast as a 14b or 16g and has the max flow of a Super HX40. The HE351VE is close to delievering that and in order to ride that surge limit and get max spool as early as possible you need to know how much it is flowing. Boost and RPM can get you close to knowing how much it is flowing.

Also one specific shaft RPM can give you many mass flow rates at various PR as is seen on the compressor maps.....your engine could also be at various load points as well. I am not sure how you get load from just shaft RPM.

Either way, for a gas 2.0L engine the way to go is lots of sensors inputs to do the math and place the vanes at a close enough position to optimal to give you that max spool and then open up to try and ride that right turbine PR right to redline.
 
Shaft speed increases and decreases depending on motor and load. It's a very simple mechanic. I should of probably just gone with the DSM ecu "load". You know g/rev...

You can't get 60k rpm on my motor under 1700rpm, nor can you ever have 20k rpm with the motor over 1500rpm... so it's a super simple point to use for vane control. It's basically alpha-n substitute rpm for tps. But then remove engine rpm too LOL.

Using boost, or back pressure alone is a super bad idea. TPS and engine rpm can get you places, but realistically just the shaft speed alone accounts for so many variances that it's just super awesome.

Have you watched my YouTube video ? 30k rpm super loaded motor, 10psi. Or happy motor and give up a little.

You'll have to excuse my rudeness in the next part. But seriously why would you not go all out flow. It seems almost criminal. Getting boost earlier isn't going to make more power if you sacrifice flow to get it. Watch my YouTube video. I don't get why anyone would want to take this turbo then dumb it down to 20psi. I would just choose something else. My truck reaches peak boost and drops off but still continues to climb the shaft speed numbers just FYI. I'd have to look back but I think I hit 40psi by 120k and then bottom out at 35psi but 128k rpm.
 
Shaft speed increases and decreases depending on motor and load. It's a very simple mechanic. I should of probably just gone with the DSM ecu "load". You know g/rev...

You can't get 60k rpm on my motor under 1700rpm, nor can you ever have 20k rpm with the motor over 1500rpm... so it's a super simple point to use for vane control. It's basically alpha-n substitute rpm for tps. But then remove engine rpm too LOL.

Using boost, or back pressure alone is a super bad idea. TPS and engine rpm can get you places, but realistically just the shaft speed alone accounts for so many variances that it's just super awesome.

Have you watched my YouTube video ? 30k rpm super loaded motor, 10psi. Or happy motor and give up a little.

You'll have to excuse my rudeness in the next part. But seriously why would you not go all out flow. It seems almost criminal. Getting boost earlier isn't going to make more power if you sacrifice flow to get it. Watch my YouTube video. I don't get why anyone would want to take this turbo then dumb it down to 20psi. I would just choose something else. My truck reaches peak boost and drops off but still continues to climb the shaft speed numbers just FYI. I'd have to look back but I think I hit 40psi by 120k and then bottom out at 35psi but 128k rpm.

Makes sense about Alpha-N. I have never messed with this on a large diesel like yours so I do not know how running it off different sensor would work. I've watched a few of your vids, but not in the last year or two....I'll check em out for sure. :)

I don't just use boost, I am using several sensors....some stock some aftermarket, to caclulate where I want the vanes. It all runs very very slick. I had problems with the calc running to slow in the beginning but with some creative coding it's running well under the 10ms the holset actuator requires. I am not running enough boost or in a high enough RPM to worry to much about over speeding it. My main issue is getting to much boost to quick and it surging. Its quite violent when it was surging at the start of this journey but I have that under wraps now.

Its a fair question and no rudeness taken. The reason why is to extend the power band. When road racing you want a nice flat and wide power band. The sooner you can get into the boost the wider it is. It's all about the area under the curve on the torque graph. I am getting lots of air flow at 4700 RPM and 22 psi....I get about 29 lb/min and around 270 ftlbs at the wheels according to Link (which isn't super accurate but its an okay ball park). So I get 29lb/min early in the RPM range and b/c this turbo can flow 69 lb/min before the comp chokes (or rather efficiency is dropping off fast) that torque will last till redline.

Basically you want to hit peak torque sooner and want it to last as long as you can. That's why the holset+contoller is beauty. It generates lots torque down low and pulls like a train right to redline. The the bigger the power RPM band the better.

I am only running it at 22 psi right now b/c my motor is shiiiit. It's the last and final phase of my build. Once that is done (in the next month i hope) then I will turn it up to 30 psi. I agree holding it back under 30 psi is not what it likes....it's still moving that car good but it could be loads better. The Holset can make a lotta boost but efficiency starts dropping off over 35 psi according to the compressor map. It's prolly why your boost settles down to 35 psi...but I don't know enough about your setup to stand behind that so its just a guess. At 30 psi that thing is right in the sweet spot of all the efficiency islands so that's where I am gonna run it. I will then up my redline to 10,000 rpm to get up into the 60+ lb/min range. In the end I will have around 6000 RPM of peak torque (not quite as torque drops off at the higher end due to cams and intake and head efficiency dropping off) so it's gonna be a blast at the track. :)
 
Clearly you cannot tell the difference between constructive criticism and hate.

Hmmmmm.....

I think you have a lot to learn about radio and flow turbines.
You're also not the first person to have it electronically controlled variable geometry turbo charger on the DSM mine is actually controlled by a 1GECU

These two offer almost no advice....they simply so "no your not the first I did it before you.....and you don't know enough to do this" That is not constructive at all....its clearly salty.

I'll play along ....let see how much of this is salt and how much is constructive......

I feel really bad for you, your in for a tough journey in life. Honestly, your lucky the old timers like defiant are note still around.

Aaaand were straight into the salt on this first sentence.

I've looked at your posts, and I can easily see the "I just finished ME school I know everything attitude". Let me tell you, having a b.s. In ME doesn't mean shiiiit. I get it, you worked hard, solved a lot of problems.

Saaaaaaaaaaaaaaaaaaaaalt.
Now your attacking my education (assumed correctly). You sure you don't wanna bring my family or religion into this too? Pull the perfect Salt Trifecta?

A typical undergrad ME student gets a 1 semester course on gas turbines - as an elective, you might also have gas dynamics as an elective. How do two semesters of school qualify you as an expert?

Saaaaaaaaaaaaaaaaaaaaalt.
Never claimed to be an expert...in fact I said I have a lot to learn. Still attacking my education.

Did you design turbines for years before this?

Saaaaaaaaaaaaaaaaaaaaalt.
Actually this is flexing in a condescending tone.

Here's the problem with your theory

What theory? I didn't put forward a theory. I did say "low low low" and to clarify I was implying that is was relative to a small housing such as 0.55 in boost at high RPM.

turbine efficiency isn't fixed, it very much depends on pressure ratio, vane geometry, rotor speed, inlet conditions, outlet conditions, and a whole slew of other things. It's very possible that by opening up the nozzle all the way that the efficiency tanks so far that it will still take a huge pressure ratio to make enough shaft power to run the compressor. However whats far more likely is that at large openings the inlet enthalpy will be far too low to generate enough shaft power to run the compressor. Furthermore, have you figured the area of the turbine exducer? It's probably on the order of 14cm^2. Most performance turbo's that have low back pressure have an exducer area about 2x the nozzle area.

Constructive....FINALLY. Took a lotta salt to get here but we finally got some content we can discuss.
Clearly turbine efficiency is based on a bunch of variables....but I didn't state anything about efficiency at all. The infamous Strawman tactic but I'll bite regardless....

"It's very possible that by opening up the nozzle all the way that the efficiency tanks so far that it will still take a huge pressure ratio to make enough shaft power to run the compressor"
I can see the merit in this however the data would show that and it doesn't with the way I am doing it and the algorithm I wrote. If you get into boost and go 100% open at XX psi then for sure its gonna drop off....and we can see that with the work that Aero and others did. They stated that using a mech vac pot at a set boost psi opened the nozzle to quick and they lost spool. So building on that I knew (and they knew) that you need to open it gradually instead of instantly at some boost threshold. What I did is found empirically what works through the logs. From the logs it was easy to go and find what positions produced quick spool and what positions choked the exhaust. From that it was easy to find at what rate I could open the vanes back up without letting all that built up energy escape suddenly or without choking the exhaust.

"However whats far more likely is that at large openings the inlet enthalpy will be far too low to generate enough shaft power to run the compressor."
Totally agree except for the enthalpy part....as the RPM and boost goes up so does the mass flow rate....as the mass flow rate goes up the pressure ratio across the turbine goes up too...as pressure goes up so does temp....the overall affect clearly increases the enthalpy of the mass flowing through the turbo. Therefore at high RPM and boost with the vanes all the way open there is tons of enthalpy in the mass flowing through the turbine. This can be seen physically as without the controller and the vanes all the way open it spools up to full boost at 6000 RPM.

"Furthermore, have you figured the area of the turbine exducer? It's probably on the order of 14cm^2. Most performance turbo's that have low back pressure have an exducer area about 2x the nozzle area."
By reducing the area of the nozzle I can see the turbo making boost lower so I know I am make it to that special PR sooner in the RPM band....then once I hit it full boost I slowly open it back up until its all the way open. From testing with it all the way open for a whole pull, as stated above in the enthalpy part, I know that if I have full boost and I am above 6000 RPM I can just leave the vanes all the way open and she breathes awesome. I am betting that at 10,000 RPM and 35psi of boost back pressure will start to climb but I haven't made it that far....this is just the beginning of this device and I'll get there sooner than you think.

"Lastly are you aware what goes on when you start pushing a turbo towards choke flow? The efficiency tanks on both sides"
I don't agree with this. If you are choking the flow on the turbine side, that does not mean that the compressor side suddenly loses efficiency as soon the exhaust start choking. As you start to choke the flow on the turbine side, initially the compressor could be operating in its peak efficiency island (say a 78% island for convo sakes). As you keep moving up in RPM the choked flow will be producing XX shaft power but in order to keep the compressor mass flow rate moving up at the same PR that shaft power will not be sufficient and you would start to see boost drop off (and mass flow rate as it is function of the PR) as there is not enough shaft power being produced by the turbine. As the PR starts dropping off it does not mean that the compressor is not still in a 78% island....maybe its in a 76% island...that's not exactly "tanking". I guess "tank" is a subjective word tho so we could debate what "tanking" efficiency is all day. That being said if you keep climbing in RPM while choking the turbine the PR and mass flow rate will drop and eventually you are sub-70% efficient and I'd agree that its starting to tank. I guess what I am saying is that just b/c the turbine is starting to choke does not mean the compressor efficiency suddenly starts dropping off....the PR would drop and the mass flow rate would stop climbing and likely plateau at a certain point.

"Surely you know the power equations for an isentropic compressor and turbine, and you can see that if it was unwastegated and 100% efficient, the compressor pressure ratio is proportional to the turbine pressure ratio. I hope you can understand what that implies."
This is worded to bait me into asking "what does that imply" as all you have asked is "do you know the equations (which I do)" and then said "if its under these conditions" and then said "the comp pressure ratio is proportional to the turbine PR (well they are directly connected by a shaft and the flow from the comp goes to the turbine after combustion so obviously they will be proportional)". Then you ask what that implies? You haven't stated enough for anyone to figure out where you're going with this....so no I do not know what you are implying. Can you elaborate a bit more please?

"Even on the compound setups there is still plenty of back-pressure when running the turbo's in the efficiency islands, and with inter-stage cooling. A giant nozzle isn't a golden ticket to low backpressure, sure it can help, but it isn't all there is to it."
AAAAAAaaaaahhhh I think I FINALLY know what your on about now. You are interpreting me saying "low low" as its low everywhere all the time through out the boost and RPM range. I can see how you got there now as my statement is more over arching than I meant it to be....perhaps I may have even used the completely wrong wording. To clarify: I was referring to the excessive back pressure that peeps get when they run the 0.55 on the HX35 at higher boost levels and start seeing it choke at 6500-7000 RPM (not my numbers but "quoted" from other threads)....I don't see that excessive pressure at those RPMs/boost levels b/c that 24cm housing is too damn large. I am measuring it, no....but you could rightful assume its lower than a 0.55 correct?

What I should have said was "I am closing the vanes early in the RPMs and at low or no boost to build back pressure earlier in the RPM band to reach a PR across the turbine to promote as quick a spool as the HE351VE will allow without causing it to surge, while managing that back pressure from reaching excessive levels by opening them gradually to maintain spool/boost until a target boost is reached and at an RPM that will allow the vanes to be fully open. When the engine is over a certain RPM and at target boost the vanes are fully opened and the back pressure is much lower than you would get from a housing that would net the same spool but would choke out at higher exhaust mass flow rates such as the 0.55 AR housing on the HX35."

Considerably more wordy than "low low back pressure" and to be totally honest you make a solid point: "low low" was totally vague and I can see how you could jump to the conclusions you did. Lesson learned....I'll try and be much more precise in my wording. I guess that's my fault....I've working on it so long articulating what its doing and why it works so well is not something I have had to do. Been in my own bubble with this so feed back is good (even if it is saltyAF). :)

"The point was there are others out there, probably more outside our platform."

That is why I was soooo specific in the wording there. I know there are others on other platforms (lots of dodge guys run the fleece one) and I have looked extensively at all the electronic controllers out there b/c if someone had come up with what I have before I ended up reaching my goal I would have just bought thier's. But there isn't. The banshee is close but the guy who made it wants you to pay him for map changes (or he did at one point)....and it costs to much and uses all these special pressure sensors and stuff.

You can spare me the lecture on CAN and J1939, I know plenty on that.

Back to the saaaaaaaaaaaaaaaalt and with a little flex on the end of that one.


"I'm assuming the member here is hakcenter. So he basically did the hard work and your putting together a PID controller I hope?"
No not hakcenter, it was iliveforboost. I also acknowledge his work there as well. I wouldn't call it the hardest work of this though....not to steal any credit from him either I'm sure it was a challenge for him to get it originally. Bot for me it was the main algorithm and the main program. Took maaaany hours of coding and several 100 km's of pulls to get it to work as smoothly as it does now.

As for controlling by the dsm ecu? Absolutely it could be done. Create a serial to can convertor, and send data out of the ecu using the aldl. The converter sends a byte to the ecu requesting a vane poistion, the ecu responds with a byte of vane poisition. Not that big of a deal, you would be hijacking the diagnostic command section of the serial comms. It could potentially interfere with data logging, but there most definatly are work arounds. I'm currently working in this area, trying to put together better logging. Dumping the request-reply format, for a request-reply with like 10 channels at once, giving a massive speed up in logging.

Can't touch this one....I am totally out of my league. I'm a MechE not and EE remember? :)
That being said it does not sound like something the average user could do. Not a big deal to you, but sounds like greek to me. Mine is plug in play right into my car and Link now. SlickAF.
I'm wrong you absolutely could do it I guess.....but with enough knwoledge, $$, and time I am sure you could get an Apple watch runnning it as well. Again, average DSMer very likely could not do it. Some guys still struggle with the concept of deadtime on injectors even though there are many writes ups on how to do it and what it is.

Lol no, it's a hydraulic servo.

Saaaaaaaaaaaaaaaalt.
This is what I find condescending in your posts. You can't just say "No its a..." you feel the need to laugh at me first for making an incorrect assumption by saying "Lol no, its a....". I did a quick google search on your turbo, saw a pic, made an assumption (incorrectly) and you come back in swinging again.

Either way it is still cool. Where do you get the hydraulic pressure from then? Is it built in to the actuator some how?

"Anyway, you haven't even pushed into territory that you see any benefit from what you are doing, or even see if it is a benefit."
No? I think it spooling at 4000 RPM in 3rd with my controller running the show is a benefit when it previously spooled at 6000 RPM. Personally I think a 2000 RPM reduction in spool is a benefit.

"When I first started it had a odd ball compressor, 57mm ind 88mm exducer. With that wheel it would bury a 4bar map at 4000rpm in 3rd gear. 3000 in 5th."
So you had a 42 trim GT3788 running at 43 psi of boost (where a 4 bar would be buried) then? From what I can find this is the stock Ford Powerstroke 6.0L turbo on the 05-07. There are no full compressor maps but I did find the comp map on Garrett's site and it does have the stock surge line on it.

So at a boost pressure of 43 psi (which is where you would bury a 4 bar sensor) and assuming you at the air pressure it currently is in Mendota, Illinois right now (which is 14.724801 psi), and this is still a 2.0L DSM, and you had similar VE's to the SD table in Link....then your comp map would look like this for a "bury a 4 bar at 4000 RPM in 3rd":

You must be logged in to view this image or video.


I filled in the spool up boost pressure so that it would ride the surge line and not just be instant 43 psi.

You seem to be way off the map here. Its pretty funny how you were calling me out on efficiency earlier and from what you claim you are running your turbo at (overrunning a 4bar) that you are clearly outside the all the efficiency islands for your compressor.




Now if we look at the "overrunning a 4 bar.....and 3000 RPM in 5th" comp map:

You must be logged in to view this image or video.



For this one I had to leave it with sudden and instant 43 psi of boost as you are so so far outside the surge line that it would be physically impossible that this happened. Past that you are once again waaaay outside any and all efficiency islands on this one.

From the above I'm guessing that was a typo and you meant to say "burying a 3 bar map sensor" but I dunno.


aaaaaaaanyways.... GO OILERS!!:D
 
Last edited:
I'm just glad people are still making progress with this turbo, I have had mine running for awhile and love it. Blew a worn out motor before I got to push it too hard though. New motor now, and swapped the m45 to a m62 and new SMIM which meant all new piping. Should be fun with a lot of boost, hope she doesn't explode :(
 

Attachments

  • IMG_20170312_121729.jpg
    IMG_20170312_121729.jpg
    366.6 KB · Views: 335
I'm just glad people are still making progress with this turbo, I have had mine running for awhile and love it. Blew a worn out motor before I got to push it too hard though. New motor now, and swapped the m45 to a m62 and new SMIM which meant all new piping. Should be fun with a lot of boost, hope she doesn't explode :(

That's awesome man! Hadn't seen you post in a while. Glad you are still around and kicking azz. :D

As stated above; thank you so much for your work on sniffing the CAN data and posting in on here. Most def would not been anywhere with this controller with your ground work.
 
Not a problem at all, glad I could help. I did send my turbo off to have it upgraded, I originally was told they would put on a 67mm comp wheel and a larger exh wheel. Long story short it just ended up getting a 67mm comp wheel same exh wheel. It seemed to spool the same still, never got the chance to turn it up as my motor gave out. Anyways, anyone have any speculation on what this wheel combo may do on this vgt turbo? I'm hoping for as much hp as my 2150's will allow on e-85 and wondering if this turbo may not be up to the task.
 

Attachments

  • IMG_20170429_150438.jpg
    IMG_20170429_150438.jpg
    299.4 KB · Views: 271
Not a problem at all, glad I could help. I did send my turbo off to have it upgraded, I originally was told they would put on a 67mm comp wheel and a larger exh wheel. Long story short it just ended up getting a 67mm comp wheel same exh wheel. It seemed to spool the same still, never got the chance to turn it up as my motor gave out. Anyways, anyone have any speculation on what this wheel combo may do on this vgt turbo? I'm hoping for as much hp as my 2150's will allow on e-85 and wondering if this turbo may not be up to the task.

The HX52 is 67mm and from the holset info thread here is supposed to max flow at 88 lb/min. Not sure if you'd get that with it in a HE351 compressor housing (assumed you just put it in the same comp housing).
 
I really don't have time to respond to all the little things, but you are very very wrong in may places.

Attacking your education? get real. I never said one bad think, only implied that you have a good introduction to these concepts. There are phd's that have studied turbines all their career, and they wouldn't make as bold of statements as you have. I see this kind of shit with students that take 1 semester of cfd, and think they are ready to be a CFD engineer yet have no understanding of any turbulence models.

At the 22psi your running this at, you are nowhere near the choke point of the compressor. Your data doesn't mean shit, your not in an area that begins to push that turbo. That compressor wheel is enough to make 700+whp, your making 400 tops. You can't linearly extrapolate. You also realize guys have made close to 700 with an HX40 in the bep housing, until you start pushing all these assumptions you make are meaningless.

I love your comment about it makes full boost at 6000 with the vanes open. 22psi is bi*** boost. How much will it make? and what backpressure does it have with the vanes wide open? That's right, you haven't tested at those conditions. So here we are taking the bits of info that benefit your statements, but ignoring everything else. Again the lower PR


"However whats far more likely is that at large openings the inlet enthalpy will be far too low to generate enough shaft power to run the compressor."
Totally agree except for the enthalpy part....as the RPM and boost goes up so does the mass flow rate....as the mass flow rate goes up the pressure ratio across the turbine goes up too...as pressure goes up so does temp....the overall affect clearly increases the enthalpy of the mass flowing through the turbo. Therefore at high RPM and boost with the vanes all the way open there is tons of enthalpy in the mass flowing through the turbine. This can be seen physically as without the controller and the vanes all the way open it spools up to full boost at 6000 RPM.

I think you forget that the enthalpy will be so low that you will need HUGE amounts of flow to generate enough shaft power. Don't believe me, calculate it. I'll be generous and let you assume a constant efficiency of 70% on both sides and an engine ve of 100%, What rpm will a 4g hit 30psi of boost at? Assume steady state, so rotor acceleration isn't relavent

I don't have the data on a 4g, but I have real data on others. On a 400cu in engine 26cm holset housing non wastegate it takes about 3000rpm for it to reach 30psi, and the back pressure is about 45psi. back of the napkining, thats about 9800rpm, and it's about 75lbs/min.

Actually you can size the nozzle at each rpm with this method.


"Lastly are you aware what goes on when you start pushing a turbo towards choke flow? The efficiency tanks on both sides"
I don't agree with this. If you are choking the flow on the turbine side, that does not mean that the compressor side suddenly loses efficiency as soon the exhaust start choking. As you start to choke the flow on the turbine side, initially the compressor could be operating in its peak efficiency island (say a 78% island for convo sakes). As you keep moving up in RPM the choked flow will be producing XX shaft power but in order to keep the compressor mass flow rate moving up at the same PR that shaft power will not be sufficient and you would start to see boost drop off (and mass flow rate as it is function of the PR) as there is not enough shaft power being produced by the turbine. As the PR starts dropping off it does not mean that the compressor is not still in a 78% island....maybe its in a 76% island...that's not exactly "tanking". I guess "tank" is a subjective word tho so we could debate what "tanking" efficiency is all day. That being said if you keep climbing in RPM while choking the turbine the PR and mass flow rate will drop and eventually you are sub-70% efficient and I'd agree that its starting to tank. I guess what I am saying is that just b/c the turbine is starting to choke does not mean the compressor efficiency suddenly starts dropping off....the PR would drop and the mass flow rate would stop climbing and likely plateau at a certain point.

You clearly do not have a solid grasp of compressible flow, nor compressor maps or any of it. Although, I should have said coke the compressor. When you choke a compressor typically this is considered to be when you get to less than 60% compressor efficiency. As you try and push the mass flow higher the efficiency drops rapidly past that point. Rotor speed also gets really really high as well - turbine efficiency takes a shit as well. Now on choked flow, this is defined as the point where the flow rate across a nozzle does not increase with decreasing downstream pressure. In this case (compressor) The nozzle is the compressor inducer, upstream is air pressure in the aircleaner, and down stream is just past the inducer. There is a finite limit to the flow through the compressor. Now on the turbine side, the nozzle never stops flowing, once you reach a PR of 2:1 the nozzle is choked, and the velocity in the throat is fixed at the local sonic velocity. As the PR increases further the mass flow increases due to the increase in upstream density. Upstream in this case meaning manifold.


Then here you go spouting off info about compressor maps, and you don't quite get what the axis really indicate, ie corrected flow. Using the wrong compressor maps, and bad assumptions to try and prove something I have the data for as wrong. Lol using the base map from link as what my car is? Who the hell would put a GT37 on a stock car? I'll give you a tip, anyone going fast on a particular turbo is way outside the compressor map. Maps are for nerds, time-slips are for winners.

Now here for some salt, you can't keep a (two?) stock motor together at 22psi, and your trying to tell me how these turbo's work.
 
Attacking your education? get real.

No?....

Let me tell you, having a b.s. In ME doesn't mean shiiiit

^^^ Here's some reality^^^

You must have a different meaning for the word "attacked" as this is clearly a jab at Eng education. You attacked it right here then go on to guess at what course work I took and how that wouldn't be shiiit either. I guess technically you just attacked all MechE BSc degrees....so any other MechE's reading your post will likely feel attacked as well. Good job on piissing off even more people.

At the 22psi your running this at, you are nowhere near the choke point of the compressor.

Never said I was choking the comp.....good try at yet another Strawman tactic. You should get into politics you're semi-good at it....better than most politicians for sure.

That compressor wheel is enough to make 700+whp, your making 400 tops. You can't linearly extrapolate.

Stawman again.....I didn't extrapolate nor did I claim any power numbers at all. I just posted pics and stated the spool results I was getting. You're the one pulling a Strawman and trying to make this about something else I wasn't even claiming.

"Although, I should have said coke the compressor."

But you didn't and you had been going on and on about turbine efficiency up till this point so its a fair assumption to assume you are still talking about the turbine. Under that reasonable assumption that we are still discussing turbines, choked flow on my turbine with the vanes closed (as you keep trying to point out is "tanking" the effieinecy of the turbine when I close the vanes some) does not necessarily mean the compressor efficiency has tanked. Since that is correct, you are trying to change the subject again and trying to be right about that instead. Basically, everything after this sentence is still Strawman. I mean you are right about what you are talking about in that paragraph, but I never even touched on ANYTHING you are stating....so are you just flexing again.

Why don't you write a tech article about it...you are clearly smart and experienced enough that there are more constructive ways for you to flex than to do it by Stawman-ing the shiiit out of this post. I would love to read a tech article about all of this so please take the time to do it as I, and many others, would benefit from you passing your knowledge forward.

Maps are for nerds, time-slips are for winners.

I have seen you post about comp maps so I don't get this....are you calling your self a nerd? Are you saying that people that we should all just ignore compressor maps and just huck on a turbo and see what the timeslip says? Why not just throw darts at a Garrett catalog to pick our turbos. Pretty shiiit advice in my opinion.

you don't quite get what the axis really indicate, ie corrected flow. Using the wrong compressor maps, and bad assumptions

FIIIIIIIIIIIIIINE.....you baited me enough finally hahahah....I will show I the corrected flow....I'm just lazy and losing interest....but I am a bit curious what diff it will make. I will go plug in the corrected flow into the calc and state my assumptions (send me your VE's if you want this to be more accurate) and replot and post. You say you have a GT3788AVNT which is a Garrett which has the correct maps for the 57/88 comp you claim you were using on their site so they are the correct maps. Its not an odd ball comp its from the 05-07 6.0L powerstroke....unless that was a typo or "another I should have said" too.

"using a base map from link as what my car is? Who the hell would put a GT37 on a stock car"
Had to make some assumptions and I stated them so if they are sooo way off then why are you getting so defensive and salty?Post a log and I'll suck the VE's from it and use that.

"anyone going fast on a particular turbo is way outside the compressor map"
Quite the overarching statement. I didn't realize you can only go fast with a comp operating at sub 60% efficiency as you stated. I find it amusing that you are pontificating about efficiency this and efficiency that and then back pedalling now and saying efficiency is bullshiit and you can't go fast while operating a compressor efficiently. Ummm....what?

Now here for some salt, you can't keep a (two?) stock motor together at 22psi, and your trying to tell me how these turbo's work.

HAHAHAH!!! I was waiting for you to say that about my motors. Toooooooooo funny. Even told my GF that would be in your very next post on this. I have not even tried keeping those 2 together at all so it really doesn't matter to me.

I ran the 7-bolt into the ground b/c I have two 6-bolts laying around and I never planned on staying 7-bolt anyways. The 1st 6-bolt I bought I got ripped off on as when I took it apart it had melted pistons skirts on 2 cylinders due to the guy who sold it to me RTV every single joint and getting so much excess RTV in it that the oil scrapper rings it even had bits of RTV stuck in there which is crazy as I dont know how it made it through the filter. Brutal build from that guy no wonder it was cheap cheap cheap. It was so bad that when I pulled the OFH off it had the protector caps that come with it when you buy it jammed down inside the outlet side of the housing b/c he didn't even remove those. It's pretty messed up and I wanted a motor to finish the firmware on my controller before winter so I could write my PC app for it while I wait for the snow to melt. So I did what I could to hobble it together and clean it out...but its got terrible blow by and compression in 2 of the cylinders BUT it does move the car and did allow me to finish the firmware.

The 2nd 6-bolt is getting planned out as we speak as my controllers firmware is pretty much done and the data proves I can drop spool up by 2000 RPM so its time to move to the next step. You won't have that string to pull at much longer so enjoy it while you can. :)
 
Support Vendors who Support the DSM Community
Boosted Fabrication ECM Tuning ExtremePSI Fuel Injector Clinic Innovation Products Jacks Transmissions JNZ Tuning Kiggly Racing Morrison Fabrications MyMitsubishiStore.com RixRacing RockAuto RTM Racing STM Tuned

Latest posts

Build Thread Updates

Vendor Updates

Latest Classifieds

Back
Top