Adding Boost Control

[Guest Loose_Screws]

I added the GM boost control valve today and MSnS 'Extra' version xxxU3. Gonna take it out for a spin tomorrow and see if I can 'tune' the boost controller.

 

Oh yeah, been running MSnS with MSD 6A and 14psi of boost. Traction in 1st and 2nd is purely optional, just roll on the throttle. 3rd gear is fun, but doesn't last very long. 4th gear is great, as long as the clutch holds. 5th is even worse, so I can only cruise and lightly accelerate. Not bad for a stock trim T3 and mostly '77 engine!

[mario_82_ZXT]

Could you explain the GM boost valve?

Are you gonna control it with MSnS-E?

 

EDIT: So you are gonna use MSnS-E to control (title just maked sense)

If it works good could you post more information?

 

Thanks,

Mario

[Guest Loose_Screws]

I will post up results, but may be a few days. I have some heavy tuning to do before weekend, heading to dragstrip.

 

Take a look on http://www.msefi.com under MSnS extra, page 3, boost control thread.

[ihatejoefitz]

im having trouble finding the page... can you link me to it directly?

[Guest Loose_Screws]

im having trouble finding the page... can you link me to it directly?

 

 

Thread is 13 pages long, happy reading!

 

http://www.msefi.com/viewtopic.php?t=3155

[mobythevan]

cool, I have been waiting for someone to try boost control. Keep us posted.

[Guest Loose_Screws]

cool, I have been waiting for someone to try boost control. Keep us posted.

 

Well, it's not much at the moment to keep posted about! It seems this is VERY experimental and not much info on tuning it. Maybe you can help me with something.

 

I am using the GM boost control solenoid from a Typhoon/Cyclone, part number 1997152. Do you have any idea what pulse rate this solenoid needs? I think it may be 31hz, but cannot find anything definite. There are basically four settings for 'tuning' the valve, PULSE/HZ, update rate, proportional gain, and differential gain. From there you can choose to use a target KPA table or duty cycle table. I would prefer to use the target KPA table. Have any suggestions?

[mario_82_ZXT]

I'd just like to say good luck!

 

I made it through 11 pages at school before the internet went out and it looks like it is very hard to get nice responsive stable boost.

 

I think the SAAB valve they were talking about was less Hz then that, IIRC. Maybe see if they have an FSM online for a model that has it, as it might be stated there.

Mario

[Guest Loose_Screws]

Well, I drove the Z today and tried to tune the boost controller with no luck. From what I have read and from experience, no one has gotten the target KPA table to work, so I am going to try and tune tomorrow using only the duty cycle table.

 

Also, hitting overboost protection is quite violent on the driveline! I hit it several times before deciding the testing was too rough on the driveline and stopping.

[mario_82_ZXT]

What over boost protection are you running?

 

I think they decided dropping spark is the nicest on the driveline IIRC, but it can cause a giant backfire causing your exhaust to explode.

 

Do you have a pop-off valve? Maybe you can use that as overboost protection without turning on the MS one just for tuning.

 

My $0.02,

Mario

[midnightzxt]

pop off valve is pretty much useless above 11 psi.. the spring doesn't compress anymore to adjust it

[Guest Loose_Screws]

I am running overboost protection at 14psi, cutting spark and fuel. I would have liked to set boost at like 6psi to start, but my WB works at 8psi and the boost controller would have no effect.

 

I have 3" turbo back crush bent exhaust. I may try spark only and see what happens.

[mobythevan]

I don't know any specs on the valves. I haven't looked into it before so I'm not sure how hard any of that info is to find. good luck

[Guest Loose_Screws]

Tuned the boost controller using the Duty Cycle table for about 12psi. I found that I had to ramp up the duty cycle as rpm's increased.

 

Also, the duty cycle vs. boost amount is NOT LINEAR! Upping duty cycle from 25% to 30% may increase boost by 2 psi. Increasing from 30% to 35% may increase as much as 4 psi. This is something to consider when tuning your boost controller!

 

I'll post up my duty cycle table later on, right now I am headed to the strip.

[Guest Loose_Screws]

Went 13.13 today at 106mph, running around 13psi of boost on standard T3. Having trouble with traction and clutch.

[mario_82_ZXT]

Spill the beans already!

 

What year make and model for the valve and what are your settings at.

Does it spike? How stable and responsive is the boost?

 

Congrats on the times, now go beat those 350z's and new mustangs

 

Mario

[Guest Loose_Screws]

Spill the beans already!

 

What year make and model for the valve and what are your settings at.

Does it spike? How stable and responsive is the boost?

 

Congrats on the times' date=' now go beat those 350z's and new mustangs

 

Mario[/quote']

 

 

It's the GM boost control valve from the Typhoon/Cyclone. It's set at 19hz PWM and I used duty cycle table. The boost is stable, but need to set more of a ramp and have it only come on after 50% throttle.

 

And as far as the 350Z's and new mustangs, already have a few under my belt. I actually ran an 03' S/C'd cobra yesterday at the strip and almost beat him. I cut a 2.0' 60' to his 2.3 60'. Time difference was .01!

[Tony D]

Do you have any idea what pulse rate this solenoid needs? I think it may be 31hz, but cannot find anything definite. There are basically four settings for 'tuning' the valve, PULSE/HZ, update rate, proportional gain, and differential gain. From there you can choose to use a target KPA table or duty cycle table. I would prefer to use the target KPA table. Have any suggestions?

 

Do you understand how a basic control loop functions?

You usually have

(R) Reset (times the controller looks at the Set Point in a given timeframe---usually per minute) This sounds like "Update Rate" in this application

 

(P)Porportional Gain (the user-set percentage of the scale the controller will move the control valve when looking at the "offset from SP". For best and most hair-left-in-your-head methodology you need to know how they define P... Is it 1/P or is it P/1? Usually it's 1/P meaning P is the divisor, so with a bandwith of 10psi, and a setting of P=1, with an SP of say 7, and a real measured value of PA, your offset is 3, so what your controller woudl do is move 3/10 of it's travel, then look at the reset rate, and then make another adjustemnt. With PORPORTIONAL ONLY Control, you will NEVER reach your setpoint! You will ALWAYS have an offset equal to your Poprortional band below or above your setpoint.

 

(I) Integral Gain (don't know if this is the "differential gain they speak of, but it probably is...) is another divisor setup that is equal to time. So the longer you have a setpoint offset, the more this will take into account and make an additional correction to the move size of the control valve.

In the above case where you would have a porportional offset, the integral will let you close that gap to exactly setpoint in a very short order.

 

(D) Derivative. This is a rate of change correction factor. "Derivitave" and "Devil" both start with the same letter for the same reason "Despicable", "Destruction", & "Debauchery" do, they all connotate that bad things are happening, and you best stay clear from it. Derivative can send even the most carefully tuned system into Hysterisis from even the smallest jump... Be glad it was not included (or at least I think it wasn't included)...

 

For your best response fastest, I would use a Quick Reset Rate, initially a Large Porportional Band (move wise, not necessarily numerically!), and a somewhat medium Integral Rate. From this if you are seeing boost overshoots and undershoots on recovery, start slowing your porportional rate, till the overshoots go away. if that does not do anything, go back to your initial setting, and halve your reset rate. If it cuts it out, you will have to determine if you want a quick reset rate, or one that only makes slow porportional corrections. I prefer to keep P and I to the smallest numbers possible for finest control, and to do this a quick reset rate is usually required, ESPECIALLY when you have fast varying load conditions (like in an automotive turbocharger!)After you are satisfied with your P response, Then find a hill and see what happens when you stay in boost for a while. If it "creeps" and it's not mechanically caused then play with integral by lessengin it's correction factor.

 

Basically in any tuning loop you first make it unstable, then make it stable. Then upset it and check for response. Tuning the control loops on Centrifugal Compressors can be rather daunting. They pay me pretty well to do it on large stationary turbocompressors. Remember if you can have the luxury of a variable reset rate, USE IT. Start fast and over responsive, and work to the stable end of the spectrum. Basically start jittery and go towards null response.

 

I hope some of that made some sense!

[Tony D]

Tuned the boost controller using the Duty Cycle table for about 12psi. I found that I had to ramp up the duty cycle as rpm's increased.

 

Also' date=' the duty cycle vs. boost amount is NOT LINEAR! Upping duty cycle from 25% to 30% may increase boost by 2 psi. Increasing from 30% to 35% may increase as much as 4 psi. This is something to consider when tuning your boost controller!

 

I'll post up my duty cycle table later on, right now I am headed to the strip.[/quote']

This makes sense. The air production from the turbo is not linear, either. And the PWM equating to a larger waste gate opening as RPMS rise is also logical. The reason is the flow coefficient across the valve is non-linear. In the controllers I work on they have a "valve characterization screen" that allows you to plot flows from the actual positioning of the valve so you can normalize the blowoff valve on a non-linear cage to ACT like it has a linear cage installed.... Odd trivia fact, when I went for initial roll-out training on this thing in April 2002, it became clear they were using a MOTOROLA chip in the control computer. Then later this year going for advanced controls school, I was able to cut apart the training center's potting compound to reveal...... Guess what Processor? How to tell the instructor about Ver2.1.1 and Ver2.55 being incompatible with V3.2.1 and Ver 3.2.1 not having to use a bootloader jumper because the software now allows that! Sound familiar? I passed...LOL And he wanted a photo of my Megasquirt!

 

I would try the Target KPA Table. I will bet the reason nobody got it to work is they don't understand the PID loop requirements and how the porportional and Integral interact over time. What exactly was it doing? Overboosting liek your P was too small? What are the maximum values available for the range, and what are they based upon?

 

I should go take a look at this code and how the user interface is set up. I am not a fan of the electronic controllers, but this would give me another gadget to play with (like I need any more complexity, eh? LOL)

[Guest Loose_Screws]

Tony D, I appreciate the information. Need to let the whiskey wear off and I will try and comprehend it a little better. What I did recognize was that MS doesn't have an "I" setting, only PWM, sample rate, proportional gain, and differential gain. I am also using the xxxxU3 code, which is not the latest. Maybe the later versions have refined control loops that may allow the KPA target tables to work better.

 

Again, let me read more and try to comprehend the settings and try again on my setup. I'll post results and maybe we can come to a conclusion as to what will make it work or what code changes may need to be made.

 

*Early Times and Coke*