Think it depends on your setup, I’m using a 3 port valve and it works great controlling boost levels at the high end. Running a .7 bar spring and without any duty cycle at the ebc I get a max of .7 bar. Adding duty cycle increases my boost to my target no problems. Also making 300 ft/lbs torque at 3300 rpms with a T4 1.06 hot housing. Funny thing is I had the opposite problem with my old header, overboost when tested with the wastegate removed.

As I remember you are running a Borg Warner EFR turbo in true twin scroll mode right? they use big turbine exducers compared to other brands, and you are running also T4 1.06 housing, so very effecient setup with very LOW backpressure. What boost are you running? and what size WG's?

It depends. :D But from what Jakob is saying about early WG opening (especially against an overly soft spring), I believe that a 4 port (which would hold pressure to the top port until target reached) would likely work better for me. My 4 port will be turning up Monday and I'll be fitting it almost immediately.

46mm is certainly in the "larger" range of WGs.

Last night I ordered a large green (.8) and small yellow (.2) to go with the existing .5 (whatever color that is), so should have some flexibility to play around. And the ability to just fit stronger springs in the WG and get boost, regardless of what ends up being the result with boost control.

Darnit. Seems almost every time I touch the car I end up waiting on parts or tools to proceed... :)

Yeh, I probably explained it poorly. :)

At .2 bar manifold (120 kPa) and below, solenoid runs minimum duty cycle- anything that happens is down to the spring.

At .3 bar manifold (130 kPa), solenoid activates, immediately goes maximum duty cycle, diverting air to the top port.

That's the usual MoTeC advice/configuration (which, in almost all respects I've found to be pretty awesome[*]); set activation pressure 20-30 kPa below spring rate.

I believe idea is to have the solenoid involved in the process early as boost is building. I can't see, from you've explained, that boost to the top port would hurt in any way - in fact, as I understand it, seems to be exactly what is needed.

Target boost is 1 bar manifold - or 200 kPa. It just never gets there.

Never noticed this before because I've been tweaking part-throttle/cruise/idle, warm-up compensation, adding sensors and other diversions. And getting a few miles on the motor.

Now that I've actually tested "floor and hold" scenarios, can clearly see that boost doesn't build beyond spring pressure. Car is not exactly slow ya understand - just much of the boost shove you'd expect towards the end of a gear isn't there. Just a corner of the performance envelope I hadn't been focusing on while working on everything else...

[*] standard MoTeC warmup compensations very clearly not what the 930 wants. Not at all.. :D

Exactly, perfect!

Ok now understood - Perfect! Yes the all boost should go the top of the WG port as early as poosible. Good!

You will have the issue fixed with the new springs and 4 port valve, interesting will be whether if the spool-up will do better also;)

JakobM

Yes, twin scroll GT35 with 2 38mm wastegates ran up to 1.1 bar.

Ok, great package! GT35 is a smaller turbo around 62mm exducer. I mixed you up with a BW 366 turbo.
Twin scroll is KING and will hold a lower overall average back-pressure, but it also builds higher peak pressure in system per puls (hence quicker spool-up), so also higher peak back pressure. So you are properly running around a steady 1:1 in ratio of boost vs. back pressure up to the 1.1 bar target boost level with the 62mm turbine wheel exducer and appr. 16-17cm2 in turbine area (the T4 A/R 1.06). Maybe you even have a slightly lower back pressure than boost due to the Twin scroll setup and still "only" 1.1 bar of target boost, but I believe not much.

Reaching target boost is one thing, and as mentioned MANY people are running happy with the 3 port valve setup. However, my experience - and I beleive it is simple psysics - the WG(s) cannot hold the WG valve closed between the DELTA of spring load and boost pressure with a 3 port boost solenoid setup (asuming 1:1 boost vs. back pressure).

You have a delta of appr. ~0.4 bar gab/time area between 0.7 bar and the 1.1 bar if running a 1:1 boost vs. back-pressure ratio. My estimation is you do not have as low as 0.7 bar back pressure all to the end of your boost target. I guess your back pressure climes past 0.7 bar before you reach 0.8-1 bar of boost. You do not hear it, it does not impact on target boost level in your case, you still have awesome results overall, however gains on spool-up in the Delta area are still there to be achieved. The gains can be irrelevant or quit major. Just my experience;)

You have a very efficient setup with twin scroll and two WG's building boost fine, but I bet you, if you where to test a cover plate on those two WG's, you would find a difference on spool-up time in that particular area between 0.7 - 1.1 bar boost build-up;) If you where to aim for higher boost, the difference would be even greater.

With a 3 port valve, any back pressure above spring level WILL have to open WG valve regardless how it is set up in EFI management. There is simply no "amo" left mechanically to hold it back when the back pressure force exceed the spring load.

The greater the DELTA becomes between WG spring load and boost level, the greater the need for a 4 port valve;)

Question for Jakob,

How does all this apply to a EBC? I assume it’s the same as your 3 port comments? When I installed mine boost response got MUCH better and I was able to dial in more. I was encouraged to install a smaller wastegate spring since the magic EBC box can control 2x base rate spring. I opted for a 7psi base spring for 10-11psi target. It just made more sense to me.

It’s a older Greddy Profec A. It matches the simple ness of the car.

I noticed I can’t seem to get boost much higher than the 11psi. Was thinking my IC is at its limits for flow. Now I wonder if I just need more WG spring. Thoughts?

JakobM

Ok I see what your saying, so the more boost you add over spring pressure the more chance you have of the valve cracking open before you want it too.
bet I'm almost out of spring pressure.

[QUOTE=JakobM;10530260]My advise - go with a 4 port MAC valve and you get a better solution - you get to add your boost pressure as your "amo" to fight against your back-pressure. A 3 port valve can not do this. A 4 port valve will provide you a 2:1 force ratio (double force compared to a 3 port valve) to fight back-pressure.

Excellent post thank you very much. This is exactly what I needed to know and it seems I've wasted my money buying a 3 port. Never mind it's a small amount of $$ and I'm glad I'll now have the functionality of the 4 port EBC.

Godos to Jacob for mastering with this thread. Very clear advice and for future everyone should just order the 4 port to begin with. I had a 3 port on shelf so I tried it but with low 7psi springs it just cant keep up the pressure and I hope/bet the spool is also much delayed. I should have pretty low backpressure with my 18cm2 hot housing but those 7psi springs just wont cut it.

Before I will buy and install stiffer springs (turbosmart instructs that one should not change springs after wg is being used for best operation?) I will test the MAP to top port test and I just ordered the 4 port MAC too. I also see the point that users dont care if the 3 port leaks power as the car is a hanfull already, but I have 315 wide r888 so I have some grip unused : )

I have tried all the options on tunerstudio for better boost but I only get a spike and its all over with 3 port. HW on my car now is holset hx40/18cm2 with two turbosmart 38 WGs/lower port/3portMAC-valve and true split design. Zork after turbo.

With my slipping clutch (new clutch is under paypal resolution, duh) at this moment I only need best spool and higher boost and no taper on 1st and 2nd gears.

Dont worry about the intake (CIS filter), it was just for the first test drive and it droppep to newer be found : ) even though I had a worm band on it plus another to catch it if it loosens and falls.

http://forums.pelicanparts.com/uploa...1595400852.jpg


http://forums.pelicanparts.com/uploa...1595400852.JPG

Has anyone got a good strategy on how to dial the PID with 4 port mac valve on 930 with smallish 7200 turbo? I've done about 100 WOT runs trying to get good result on all gears but still work to do.

hz is 19.5 and control interval 50ms. Tunerstudio has initial dc map from where the PID starts its work. Obviously I need to get the initial DC map close on the map bút different gears behave differently with rpm/kpa table. On tunerstudio boost control I'm now at p=20,i=0,d=200 and sensitivity 34 and 4th gear is spot on but lower gears stay under target. Next I will raise the initial dc table values to see if that helps.

Pic from earlier test with lower P gain where one bar was just before 3000rpm.


http://forums.pelicanparts.com/uploa...1597854247.png

Here is the last WOT where boost hits target and stays there. I have 7psi springs so there is lot to be gained with say 10psi springshttp://forums.pelicanparts.com/uploa...1597855246.png

Idk if this helps, but it helped me understand PID when building an automated PID controller for my friends Green Egg :D

Hey Smurfbus!

I came to the conclusion that it takes dyno time to tune PID... Dunno, maybe I'm just a quitter... :D

I have PID values that I know will hold set values perfectly for WOT on the rollers - but they' don't seem to work that well for short periods of boost (little/no enhancement over spring). Trying to improve PID values either produced no effect or overshooting target by 20 kPA, Added to which, boost simply couldn't be held long enough to observe the behavior to see if PID would correct to the set value...

This also suggests that Hz may be set too low (depending on your solenoid, of course):

From https://www.motec.com.au/forum/viewtopic.php?f=11&t=1326, but I've seen other references to MAC valves having a similar Hz & duty cycle range to that described (although there is a "high-speed" variant available with a wider range, I believe).

The first post in this https://www.motec.com.au/forum/viewtopic.php?f=11&t=210 thread contains a PDF explaining how to tune PID, and how the algorithm works - at least on MoTeCs..

Thanks for the links. Will read them later but it should be doable just dialing p then cut the spike with d. Dial i for possible taper. My mac valve has a range from 10 to 20hz and I already started with 19.5hz. Went for another test drive and I think the answer lies in finetuning the initial dc map. I had crazy oscillation which stopped when I lowered target so it was wrong initial dc map values that made boost duty jump up and down at higher boost.

Maybe I should first dial open loop where I want it and then use that map to iterate initial dc map for good starting point for closed loop.

After reading this post in it’s entirety I purchased a Mac 4 port to swap out the 3 port. I could not be happier. I knew I was going to overboost while using the same settings I was using with the three port but decided to “ ease” into it and see. Yep overboost thresholds were hit and I took a stab at lowering the duty table by 16 and hit my target perfectly at 20#sboost. Boost comes in much sooner and MUCH scarier than before. Thanks Jacob glad I spent the 45.00 ;)

Did you go closed loop or open loop? If closed where did you end up with PID and sensityvity lever and what kind of DC does your GT40 work with? Which WG or WGs?

My valve is this https://www.ebay.de/itm/MAC-Valves-USA-4-port-Boost-Solenoid-BCS-46A-AA1-JDBA-1BA/231676454835?hash=item35f0fe27b3:g:AfQAAOSwB4ha7Mr H
It says I think I need a stiffer spring to ease with the targets. 7psi is too weak for about 15-18psi targets to work nicely on all gears at the same time. Before I get the springs I will try to massage the initial DC table so its a best possible compromice.

Newest ms3 firmware should have PID logging parameters too but last time upgrading I had some problems with odd changes on my tune so I'm on the fence about upgrading it.

Here is a scatter plot of all of my 4port WOT runs. I hope to get working initial dc map with these numbers. Automated histogram of the same runs gives a bit more higher DC numbers than looking at the data.

http://forums.pelicanparts.com/uploa...1597922699.png

http://forums.pelicanparts.com/uploa...1597923406.png

Smurfbus, I’d sure like to see a pic of your intercooler tucked under the tailless engine cover.