Improving stock boost control?

[InvisiBill]

Just Datamastered my first 1/4 mile run at the track. I noticed how much my boost wavered up and down. It was only down to 13.3psi, which is better than some people's trucks, but a nice solid 14.7psi would be awesome. This is a daily driver, and I don't have a lot of spare cash. So what are some things that can be done to improve things?

Tuneup: I know that other broken parts (vac leaks, etc.) can screw up boost control. Making sure everything else works is a prerequisite, I know.

Adjusting the WG rod: This was a TSB or something right? It was originally specced at 1/2 overlap on the holes, and they changed it to 7/8? Anyone got any good pics of this? I'm a very visual person. Being a newbie, I like to see how things are set up before I start tearing into them. I've never really gotten a good look at the WG stuff... A very detailed explanation of which way to turn what, and where it should be would be handy. Pics are even better, but I'll take those myself if I get a good enough description.

WG hole: I've heard that the puck is quite a bit bigger than the hole, so you can simply open the hole up some more to improve WG flow. Again, I'm not familiar with these parts, but I thought someone said you had to take the turbo off and apart or something to do it? I saw a pic of Christian Dionne's setup, with the larger downpipe and extra port for the WG, which looked cool, but like a lot of work too.

Exhaust: Will a less restrictive exhaust let the WG work better? Or does that really only affect spoolup and stuff?

Other" What am I forgetting? Edumacate me!

 

[syntax]

less restrictive exhuast hinders the wastegates ability to work properly

 

[Mephisto]

so if youre using an ATR exhaust does that mean you should tighten the wastegate to hold more boost?

 

[syntax]

no, less restrictive exhaust usually causes an overboost condition

 

[smeagol]

Less restrictive exhaust has less backpressure
Less backpressure means turbo will spin faster
higher speed turbo means more boost, unless the wastegate steps in to help control the turbo speed (by lowering pressure in the turbine housing).

If the turbospeed is too great, and the wastegate doesn't have the capacity (aka every stock Syty setup with opened up exhaust and whatnot) then boost is difficult to control.

Because turbo speed is dependent on exhaust pressure (turbine pressure mainly) a wastegate may have to work harder on a smaller turbo. A well thought out combination (properly sized turbo) will not stress a wastegate as much as an undersized turbo/bad combination.

 

[InvisiBill]

It was late when I posted that... I guess I was kinda thinking of somehow making the exhaust less restrictive for the WG, without making it less restrictive for the turbine. I do understand that a less restrictive exhaust means the turbo will spool up faster, and thus the WG needs to be able to respond more quickly.

Yeah, what about the ATR? I still have the stock exhaust which is just starting to rust. Will the ATR exhaust affect boost control noticeably (either good or bad)?

So there's not really a whole lot of "tinkering" type stuff that can be done to improve the stock setup? I'm not really ready to go to a PTE or an external with a separate boost controller yet... Is it possible to make the stock stuff control an external? Within reason, that is; I know that anything is *possible* with enough time and money...

 

[syntax]

replacing the mufflers usually does not cause a boost control problem for stock trucks, my first sy had the borla cat back, and my 2nd has a dynomax, both have good boost control. The problem starts when you gut the cat, and or run a straight pipe on a stock truck, your most likely going to run into a overboost problem.

Some people have tried to port the stock wastegate, im sure that helps a little bit, but it requires removing the turbo, which is a big hassle, and its still a far cry from the performance of a external wastgate, which was the best mod i ever did to my nearly stock syclone. A external wastegate can not be controlled with a stock setup, but all i ran with my external was a bleeder valve (about 15 bucks) and it controlled boost perfectly, just because you get an external wastegate dosnt mean you need a fancy electronic boost controller that predicts weather. HTH

 

[smeagol]

Yeah, what about the ATR? I still have the stock exhaust which is just starting to rust. Will the ATR exhaust affect boost control noticeably (either good or bad)?

So there's not really a whole lot of "tinkering" type stuff that can be done to improve the stock setup? I'm not really ready to go to a PTE or an external with a separate boost controller yet... Is it possible to make the stock stuff control an external? Within reason, that is; I know that anything is *possible* with enough time and money...

ATR exhaust is good high flow stuff, but the restriction is still in the cat.

Stock stuff control an external? Are you talking about the stock solenoid/computer programming? Then yes it is possible.

You have to clarify and specify what exactly you are asking
- you have the stock turbo/wastegate
- the stock solenoid
- the stock calibration
- the stock boost control strategy

The stock turbo/wastegate is a major issue, especially with modifications.

The stock solenoid is fine, I use it to control my PTE, I've used it to control an external.

The stock calibration is ok, but you need to tweak this for your goals/needs.

The stock boost control strategy isn't any good IMO. Dig and I modified this to use MPH to help some, but it is still a PITA. It needs a better algorithm IMO. If you want to keep it simple for boost control and you are doing your own chips, set the boost control to open loop, and use the MPH hack (again my opinion). We need to use a PID control strategy for the boost control, then maybe it'd work better.

You can play with the calibration on the stock setup, but I think you'll go insane if you have a high flow cat etc

 

[InvisiBill]

I didn't realize the stock boost control algorithm was that bad. I just thought the wastegate was undersized. I was wondering about using the rest of the stock parts to control an external WG. I'd like to mess with chipping, so I'll look into that MPH thing when I get to that point.

For us dummies, can you post a simplified explanation of the "stock boost control strategy", and what the ideal solution would be? I'm the type who likes to get to the root of the problem and overanalyze things before realizing it's way more work than I want to mess with. =)

 

[kwick6]

I don't know if you have seen this...

http://www.geocities.com/w_bowley/Green/boost1.html

Save yourself all the aggrevation and get in on the external WG group purchase. :wink:

Scott

 

[smeagol]

I didn't realize the stock boost control algorithm was that bad. I just thought the wastegate was undersized. I was wondering about using the rest of the stock parts to control an external WG. I'd like to mess with chipping, so I'll look into that MPH thing when I get to that point.

For us dummies, can you post a simplified explanation of the "stock boost control strategy", and what the ideal solution would be? I'm the type who likes to get to the root of the problem and overanalyze things before realizing it's way more work than I want to mess with. =)

The wastegate is undersized, and the turbo is undersized, those are the 2 MAIN issues. With a good sized turbo & wastegate, boost control is fairly easy, which is why I'm a fan of a simple boost controller when you use a PTE or external. I'm using a very simple one with my PTE setup, and it works great (I won't speak about my combination, since it sucks).

The control strategy, in a nutshell is this.

Boost goes into open loop when coolant goes above xxx deg F, RPM goes above a threshold value, and manifold pressure goes above xxx kpa (typically set at low boost, 5-10psi). This means that the wastegate duty cycle won't start moving, and the ECM won't start 'controlling' boost until those two parameters are met.
Once it goes into closed loop boost control, an initial wastegate duty cycle is set (usually 60-70%), and the desired boost for the conditions at that point is looked up in a table (stock table is TPS vs RPM). Ok, we determine we want 15psi at this condition. The actual boost level is read, and compared to the desired boost level. So we see if we are overboosting or 'underboosting'. So let's say there is a difference of 3psi, we are overboosting by 3psi (18psi actual, 15psi desired). The difference value is compared to a deadband value. Let's say that value is 2psi. The 3psi is greater than the 2psi deadband, so the computer takes action and is now going to step down the wastegate duty cycle. It steps the wastegate up or down (depending on whether you are overboosting or underboosting) in what is named positive or negative step adjustments to wastegate dc in promgrammer. These are a set value, usually set to 2-5% dc. So the scenario occurred like this, we hit closed loop boost, wastegate duty cycle went to 65%, we spooled up, we saw too much boost for our conditions, an adjustment was made to lower wastegate duty cycle by 3% (for example), and now our wastegate duty cycle is at 62%. This closed loop update (check actual boost level, compare to desired boost level, adjust wastegate duty cycle, start the cycle over) occurs at a set frequency as well, which we can adjust in the chip. The update rate is set in tenths of a second, and a typical update rate is 0.1-0.3 sec, so an update is made that often. Still more - the wastegate duty cycle can be reset, usually done at a shift point, and the RPM drops by XXXX RPM. So a negative RPM change, when greater than another threshold value, will force the wastegate duty cycle to reset to a value that is set in another table, a single line table versus TPS. The closed loop cycle begins again. Dig has been using a hack of this last table, so the reset wastegate duty cycle table is versus MPH. This is good since now we can adjust and reset the wastegate duty cycle when a shift or converter lock occurs, and do it versus MPH, to try to help account for boost spikes.

So, in my opinion, here are the flaws with this system. The wastegate duty steps are SET values, unchanging unless you reburn a chip with different values (there is one for increase in dc, and one for decrease in dc). So if you are overboosting by 10psi, the duty cycle will still only step down by 3%, same as if you are overboosting by 3psi. This value should be a proportional term, so if you are overboosting by 3psi, the duty cycle stepdown is by 2%, and if you are overboosting by 10psi, the duty cycle should step down by 10% (these are just examples, testing & tuning would have to be done if such a strategy was ever integrated). You can imagine the same situation at spool up. We can't be as aggressive to spool the truck up when the duty cycle can only do up or down in set increments. Now do you set the step in a large or small values? Too large, and you can help avoid those overboosts, but you will overshoot your target often, and you will have VERY wavy boost control. Too small, and the control is too slow, and the boost typically gets out of hand. Update too fast, and again, you get steps up & down in the wastegate duty cycle, and in the actual boost level.

I've talked to people in the calibration business about this... the better way to do it is with PID control, proportional integral derivative control. It is very complicated, but takes into account the rise & fall of the term you are trying to control (boost level in our case), and also has the ability to be adjusted & tuned, so you don't have overshoot or undershoot, or the wavy result where it cycles overboost/underboost. The steps are proportional to the desired change the computer has. If boost is way off target, the step is bigger, if the boost is close to target, the step is smaller, or no change is made.

If anyone sees anything that needs correcting, post it, or if you have any better descriptions of PID control.

 

[Maxtor]

Great stuff, Brian. Thanks for the post.

I worked with the code to try to get boost under control with very limited results. I could improve it but not really fix it.

Installing a Mitsu 15psi actuator and a manual boost control valve from "boostvalve.com" dramatically improved the results I was seeing with a basically stock setup - turbo upgrade using a mitsu 20g 12cm2 turbo. Up to 16-17psi boost was rock solid with this setup. Higher than 17 and I still have some flututation so I'm going with an external at this point - enough of the headache.

 

[smeagol]

Yep I hear ya.

Just a note though (since I'm trying to not confuse people, but will end up doing so anyways ) - you were working with the calibration tables, not the code.

The code is the actual software, the algorithms, that control the ECM outputs. The calibration tables are the lookup tables, constants, that are used by the code.

If you edit the code, you are changing the software, the control strategy. They are both contained in the eprom, but in different locations. If you are editing in Promgrammer, you are editing calibration values/tables. To edit code, you need to disassemble, and reassemble, or do other painful steps to edit it.

By going to the turbo with the larger exhaust side, that had to help ease the strain on the wastegate, but like you have seen, it is still a PITA.

I am a fan of the PTE setup, since I've seen it work. I am way unimpressed with any Mitsu wastegate setup.

 

[Maxtor]

You are very correct, Brian. I was making changes to the tables, not the code. I get used to looking at lines of machine (PLC) code at work and referring to all such things under a blanket statement of "code".

The larger exhaust housing did help with boost control. If I wanted to only run 15psi - I could just run the mitsu 15 psi actuator with no other controller and would get a rock solid 15psi with the 12cm2 housing.

When making changes to the tables, it seemed like I was reacting in the correct direction, but by the time the change would have given better control another change occurred. It's like trying to maintain solid control with a tug-of-war match. I know part of it was inexperience on my part in dealing with changes to the tables but it sure seems like there just isn't enough control there. I talked to a couple of people who were very helpful, Barry Tisdale and Bill C. who both gave me ideas on what to work with - it improved but never was "fixed".

I agree, a real PITA.

 

[HighPerformanceTrucks]

I have a cat delete, ATR exhaust and stock turbo. I run 17 psi at all times on water injection since anything below that will cause an overboost situation and the Ultimate chip will shut the truck down. I have descent boost control from 17 psi-20 psi except some small spikes when the air is really cold and dense, like San Francisco foggy nights.

Adam Auerbach
Sy #2419