rx7jocke

Cant get my open loop boost control to keep my boost stable...it keeps going up and down +-2psi(ish) in even pulses.

My setup is dual 38mm gates with a 4 port solenoid, the vac hoses are about 20inches long all 4 of em.

This is my map:

rx7jocke

the frequency is at 15 hz

vrx8

iTrader: (18)

Try 33Hz

rx7jocke

i remember the tuner tried both 25 and 30hz and it made the boost go bananas, and oscillate between 14 and 18 psi in 5 big waves over the powerband

Vicoor

Someone else might want to chime in,
But I'd say since you are in open loop I wouldn't expect a problem with the frequency setting to cause an oscillation. Now having the wrong frequency setting might keep boost solenoid from responding and result in either not reaching target or just over shooting it all together.
The MAC types of boost valves are supposed to work well around 33hz and I suspect that turning that way down may be an attempt to compensate for another problem in the system, possibly even just bad values in the table.

arghx

iTrader: (3)

Please post a datalog and your Haltech map.

At 130kPa boost your duty cycle is set to zero (presumably as an overboost protection). It's possible you are bouncing between the 120 and 130kpa columns. What happens when you extend the duty cycle to higher load?

KNONFS

iTrader: (14)

Had similar problems, and created a thread:

https://www.rx7club.com/haltech-foru...oblem-1058984/

I am far from knowing what I am doing, with that said, I was able to linearize the boost on my open loop map, after doing what C Ludwig posted on my thread. His suggestion was to lower the duty % on the second to last column, and raising the duty % on the last column. In your case, it will be lowering the 120kpa duty cycle and raising the duty % on the 130kpa column.

Also, while you figure things out, I would linearize the columns, I see different duty % on each column as the revs go up. I will post a screenshot of my map after I got things working.

rx7jocke

arghx: i will look in to it tonight!

knonfs: yes please do so and i will also try your suggestions!
thank you vm!

Jobro

iTrader: (2)

It does not make sense to vary duty based of manifold pressure in an open loop system? I think the Open loop duty cycle % should be based off RPM and Throttle Position, and the closed loop target pressure if you use it later should be in manifold pressure based off RPM and throttle position.

You you configure the system in this way?

Slides

Technically if an input to the 2 axis wastegate DC table is manifold pressure (ie the measured system output variable)then it is closed loop boost control, just not using a single target boost/base DC figure and PID numbers, different ways to skin at cat.

I guess the point of having a map reference in that case (really is to run closed loop without scaring people with the concept of PID control by just populating the table with less scary DC numbers) is to increase boost response from transition into boost (if you don't have any other transition to OL boost control conditions) by running a higher DC at low pressures so that the gate will not creep open, but if DC is relatively high anyhow it shouldn't be a problem unless using a very light spring.

If I was tuning that "OL" boost table I would be running very high DC at low pressures trailing off to the steady state DC at boost target (ie the DC that roughly maintains your target boost if the whole table is populated with that DC) then dropping off slightly above boost target for at least 2 column division to stop any interpolation down to 0 DC throwing it wildly. This would effectively make it a closed loop system with (variable) proportional gain. You just need to tweat DC for rpm (due to engine VE and turbine efficiency changes/EMP change as flow increases with RPM).

Which clown at Haltech called that manifold pressure referenced table open loop? Fair enough if you make every pressure entry at an rpm point the same it is open loop, but then why have a 2d table, just have a single line rpm / DC entry with a minimum manifold pressure for wastegate output threshold.

Given the lowish DC I'm guessing you are running a spring pressure not too far off your target boost, any reason to go to the effort of 4 port? They are more sensitive to DC change than 3 port. You would usually only go 4 port (combined with very light spring pressure gates) if you wanted to run a very wide boost target range, say from like 4 PSI to 30PSI+

Slides

That closed loop table setup would allow you to do some tricky stuff like "split" high rpm boost target for driving in traffic. You could contour the DC across the boost levels so that there were two "lanes" which you could not cross from above a certain rpm point (if holding WOT). Would mean if you got right into it at low rpm it would transition along the top boost curve, if you rolled on steadily into revs it could give you a low limit so you could still go WOT for a smooth rev out but with much less power, could be good for wet weather driving if you didn't want any additional switch gear in the cabin to run a more traditional Closed loop PID & variable target setup.