Boost control opinions...

[ShelGame]

I'm looking for opinions and data regarding methods for boost control. Our stock systems basically have 2 types - source control and bleed control. The T2 uses the source control - it only allows the WG can to 'see' boost when the solenoid is active. The late T1 ('88-up) use the bleed style - the WG always 'sees' manifold pressure, but the signla is bled off until the solenoid is active; then it sees the full signal and opens to control boost.

The later Mex cals use a different version still. It uses 2 solenoids in a bleed/signal type of combination. it may be the best of both worlds. I have not fully decoded the Mex boost control routine, though. So, I'm not really sure how it works. Additionaly, they use a solenoid to control the BOV. I attached a diagram of the Mex vac system.

Basically, I'm looking for ideas on what to do with boost control in the Turbonator codebases. I have not been able to find out much information at all about how electronic boost control systems work. There's a little documentation about the Megasquirt system, but it seems really basic compared to our stock stuff...

[risen]

I can fill you in a little on the MS system as my brother and I just got the boost control working on our glh-t a month ago or so. In the later ms2-extra code there are 2 types of control: closed and open loop.

The closed loop control seeks a boost target based upon a tpsxrpm=kpa map. MS is so flexible it could probably be setup any which way you want it. There are 3 settings in addition to a rpmxtps=kpa map that set how hard it seeks it's kpa target. That setup is actually rather complex, or at least I think it is every time I look up PID controller on wikipedia. I have no experience with this setup on MS yet. We didn't mess with it because we just wanted to get the control working correctly.

The open loop one is fairly simple, it takes a rpmxtps=Duty cycle table. MS looks up what your tps and rpm are, and interpolates a duty cycle value for the boost control solenoid. We have this setup in the glh-t and after about 20 minutes of tuning it's pretty good. No spikes in high gears and the boost climbs pretty hard even in 1st and 2nd gears. It's actually a pretty effective system that's easy to tune in.

The solenoid itself is alot like our standard solenoids, but it's from a GM typhoon. It's setup to have a bleed on the wg line when the solenoid is on (like the T1 setup). I'm not sure if given the proper adjustability that we really need the 2 solenoid setup. The bb60 boost control in my csx is good @ wot but is not so good on the highway (spikes to 25 psi easily).

I'd love to have something like the tpsxrpm =kpa target table for our cars, but I'm not sure how easy it is to implement a PID algorithm in ASM or how much cpu it uses. Although, even 2 duty cycle maps (one for tps and one for rpm, and larger or lesser of 2 would be used) would be a hell of an upgrade I think.

[Tony Hanna]

I'm probably stating the obvious, but it seems to me like the engineers were looking at how fast the turbo spools when they designed the boost control systems for the TI and TII. The bleed type makes perfect sense for the small, quick spooling mitsu. Since the wastegate is always seeing signal, and the vacuum plumbing constantly has at least some pressure when in boost, solenoid events would translate into wastegate movement almost immediately. This seems like the way to go with something that spools that quick.

The TII system on the other hand seems designed to help the turbo spool as quickly as possible by keeping the wastegate actuator from seeing any pressure at all until it's necessary to control boost. This makes sense considering the larger Garrett is slower to spool.

It seems to me like the TII setup would be the way to go for anybody running a TII or larger turbo since the bleed type may cause the wastegate to open prematurely resulting in slow spoolup, and possibly even limit the maximum attainable boost depending on the restrictor orifice size.

I ran into this on the Sundance last weekend. With the actuator unhooked, it would make 21 psi. With the stock wastegate plumbing connected and the cal setup for a 20 psi target, it would make 16. I had to switch to a smaller restrictor to get to 20.

Just tell me to STFU if I'm babbling or not making sense.
Tony

[GLHNSLHT2]

+1 on using the T2 setup. I still can't get your T1 boost control in your old programs to work right.

[ShelGame]

Here's my opinion of the 2 types our stock systems use:

T1 - Advanages
1) Failsafe mode: when the solenoid fails, you get only whatever boost the WG will support. In the case of the stock Mitsu, that's 7psi. If the WG Solenoid fails on the T2 style system, you get unregualted boost. Of course, we have OB protection just for that reason.
2) Slow boost rise: The advantage related to slow spool is that it's less likely to spike. When using a small turbo at low boost (like the Mitsu on a 2.5), this can be an advantage.

T1 - Disadvantages
1) Tuning: The restrictor and WG duty cycle are dependant on each other. If you increase the boost target, you will get to a point where the WG Duty Cycle is 100% and you will not be able to increase the boost any more without increasing the size of the restrictor.
2) Slow boost rise: Slower spool (generally) than the T2 style. But, this can be an advantage with a small turbo - see above.

Does anyone have a feeling for accuracy of the 2 systems? I know on my race car with the T1 style BC, that I get >1psi delta to the boost target. Now, I haven't done much tuning with the system to try and improve this accuracy. Any ideas on how accurate the T2 style is? I'll probably just build it in to a test cal and see how it works with the Mitsu turbo.

Another note - both types (T1 and T2) are closed loop. They set a boost target and use that to determine the WG Duty Cycle. The Duty Cycle is also modified based on the boost error (ie, delta to boost goal). Although, I still don't completely understand the mechanism for the feedback to affect the DC.

[GLHNSLHT2]

On my 87 T2 setup I set the boost at 14.5 and it was dead nuts. Now I went to smec and changed turbo's and it would bounce from 14-10psi at random even though it was set to 14.5psi. Maybe the arm is to tight, wastegate too big. I don't know how to play with the wg tables to make the wastegate do what I want so I just slapped on a grainger.

I could never get the T1 boost control to work at all on my 2.5 with a garrett. It either ran minium boost or maximum boost. Tried different vacuum sources, restrictors, everything. Solnoid was cycling. Good thing my BOV's were leaking or I would of seen way more than 20psi.

I want my big turbo to spool as fast as it can. So I like/love the Garrett style boost control when it works.

[ShelGame]

Does the T2 system use a restrictor at all in the WG vac plumbing?

[BadAssPerformance]

I do not remember seeing one on any of my cars even before modding them.

My boost control - grainger straight off the intake plenum to the WGA

[ShelGame]

There's one shown in the vac diagrams. Not sure why it would be needed.

I built a T2-style WG control routine for Turbonator SMEC. I'll test it in a couple of weeks at the Eaton Drag Day outing @ Milan. If it works good, then I'll post it as part of Turbonator SMEC v13...

Nah, that's too far away. I'm sure I'll blast it thru the neigborhood before then...

[risen]

Another note - both types (T1 and T2) are closed loop. They set a boost target and use that to determine the WG Duty Cycle. The Duty Cycle is also modified based on the boost error (ie, delta to boost goal). Although, I still don't completely understand the mechanism for the feedback to affect the DC.

FWIW, the turbonator cal wouldn't hold 20 psi on my CSX. It would hit 20, then slowly drop back to 15psi or so. All the tables other than the duty cycle map matched my bb60 cal. Which leaves me with this question: how much does the duty cycle table have to do with it? I can pm you if this is sliding off topic.

[ShelGame]

Not off topic at all, I think.

The reason is it dropped back down is because the adaptives are set for that 'direction' - for lack of a better word for it. The system is expecting a bleed style WG system. With that, higher WG Duty Cycle = more boost. But, you copied the T2 duty cycle tables. In the T2 system, lower duty cycle = more boost. But, the adaptives are still geared to increase the duty cycle if boost is too low and vice versa. So, in effect, the adaptives are counter-acting the base tables the way you set it up.

I built a modified routine for T-SMEC v13 that will allow T2 style WG controls, and still keep the knock adaptives that were added into the original T1 code. I'll test it out ASAP and post it if it seems OK.

[risen]

Not off topic at all, I think.

The reason is it dropped back down is because the adaptives are set for that 'direction' - for lack of a better word for it. The system is expecting a bleed style WG system. With that, higher WG Duty Cycle = more boost. But, you copied the T2 duty cycle tables. In the T2 system, lower duty cycle = more boost. But, the adaptives are still geared to increase the duty cycle if boost is too low and vice versa. So, in effect, the adaptives are counter-acting the base tables the way you set it up.

I built a modified routine for T-SMEC v13 that will allow T2 style WG controls, and still keep the knock adaptives that were added into the original T1 code. I'll test it out ASAP and post it if it seems OK.

Sorry, should have been more clear. It was turbonator LM v6 I believe.

I should also mention that my car had a grainger on it when I bought it, and I put it back to the 87 t2 setup. I believe I have the solenoid plumbed correctly, but it's always possible that it may be wrong. BB60 will hit 21 and hold just below 20 psi on WOT runs through the gears. It spikes bad on the highway in 5th though.

[ShelGame]

... then there's the T3 WG control. It's a source type control (like the T2). But, the plumbing is different from the T2. The T2 is plumbed normally open (The WG sees atmosphere and the intake manifold sees nothing). The T3 is plumbed normally closed (the WG sees MAP normally, and the solenoid closes to vent the WG to atmosphere). So, the WG control for the T3 is kind of a hybrid between the T1 and T2. It also has more adaptive cells (for finer control over the RPM range), and a default DC for transient conditions to help spool the turbo.

[GLHNSLHT2]

T2!!! And yes the earlyT1 and T2 (garrett turbo) vac lines had a yellow restrictor in them just before the solenoid. I've heard it's to help increase spool up. On a modded setup I wouldn't run it though. Why? Because when the boost goal would go from 10-12psi at the 3500ish mark the turbo would spool so hard it would hit overboost cutout. Without the restrictor there was enough flow in the WG line to keep it from cutting out.