PWM Closed-Loop: Controlling static idle as well?

This is a forum for discussing the development and testing of alpha MS2/Extra code. Documentation
(Runs on MS2 and Microsquirt)

Moderators: jsmcortina, muythaibxr

Post Reply
cguenther
Helpful MS/Extra'er
Posts: 47
Joined: Sat Mar 03, 2007 5:41 am
Location: Germany

PWM Closed-Loop: Controlling static idle as well?

Post by cguenther »

Hi,

just experimenting with this new (at least for me :-) ) PWM closed-loop stuff. Read wiki about PID. Tested some setups on the desk.

After reading the forums one question is left open: As i understood the configuration-options for closed-loop, this is indended to have an adapted, controlled idle with different loads (e.g. air condition on, eletric) when coming from higher rpms with throttle idle (tps below threshold). ok.

but how does this logic react to a "static situation": throttle idle, rpm steady, lets say at 850RPM. now the A/C kicks in. Is the pwm-control able to react fast enough to control the higher load leading to lower rpms to get idle again at 850 rpms through controlling the FIdle-Valve?

i was not able to find PID/dashpot etc. parameters which seems to let the idle-valve (2 wire bosch) react quickly enough to such disturbances in load.

in other posts i find the suggestion to have a dedicated input for the air condition, telling MS to increase idle.

another post was suggesting using a seperate on/off-air-valve for the higher idle....

but, with PWM closed-loop idle control just this should be done through the fidle-valve as well, or am i wrong?

did anybody has some working paramters/setups for PWM closed-loop to have the system working as described: throttle idle, static rpm dropping below target, fidle-valve is controlling idle back to default value.

best regards
carsten
hassmaschine
Super MS/Extra'er
Posts: 1331
Joined: Mon May 21, 2007 8:36 am

Post by hassmaschine »

a well tuned PID loop will be able to compensate for changes in load pretty well. most of this will be dependent on your P value - if it's too low, the valve won't react fast enough (of course too high and you have oscillations).

it could be that the 2 wire valves aren't as fast as the 3 wire valves, but I only have experience with 3 wire valves so I don't know - but my hunch is you need to spend more time tuning the PID loop. My cooling fan is electric and when it kicks on, I barely even notice it (I have to be paying attention to tell that it came on). adding another input is redundant when the loop is already built to handle changes in load.
muythaibxr
Site Admin
Posts: 8228
Joined: Thu Oct 14, 2004 12:48 pm

Post by muythaibxr »

Yeah, tuning the PID loop properly it should react to changes in load pretty well.

I've also only tuned the 3-wire valve on the current algorithm though.

Many OEMs do idle-up the engine when the AC button is pushed though.

If we had the extra input, I'd add that, and a VSS input to the algorithm, as those would make it more reliable.

Ken
cguenther
Helpful MS/Extra'er
Posts: 47
Joined: Sat Mar 03, 2007 5:41 am
Location: Germany

Post by cguenther »

Hi Ken, hi hassmaschine,

i tried some desktop-tests, and yes, it seems to work quite properly and fast enough to compensate idle for sudden load changes.

however: Ken, would you be able to add this feature - idle-up by xxx RPM when "high" on input y - to MS2E?

in fact, my mercedes g-wagen has a decent air-valve whick kicks idle a bit when A/C jumps in. id like to get rid of all those thnigs, which could be managed from the MS itself.

thanks!

best regards
carsten
muythaibxr
Site Admin
Posts: 8228
Joined: Thu Oct 14, 2004 12:48 pm

Post by muythaibxr »

There really aren't any spare inputs that could be used for that, which is why I have not already done it.

I mean we could probably reuse another input but I'm trying to avoid that.

Ken
cguenther
Helpful MS/Extra'er
Posts: 47
Joined: Sat Mar 03, 2007 5:41 am
Location: Germany

Post by cguenther »

Hi,

yes, there are no spare inputs. but i think, that one has to decide already, which pin is used for what. An option to select pin x as an input for our idle-speedup, is only an option, not a fixed pin, that has to be wired to get MS working.

with GPIO we will have no more problems, so why not implement this feature now for those, who only need this and 2 other features for which are pins available?

best regards
carsten
mops
Master MS/Extra'er
Posts: 705
Joined: Sun Sep 25, 2005 5:13 pm
Location: Auckland, NZ
Contact:

Post by mops »

I dont have AC, but if I had I'd wire up a small aux idle air valve and would hook it up to ac compressor solenoid, so if the ac compressor turns on, valve opens and allow bit extra air to engine... that doesnt require any special tuning or extra outputs... i know kinda crude, but should work ok... but what do i know...
BMW, 1985, E30, 325i, 2-door, 5spd. Lots of custom work. Turbo build in progress: http://www.e30tech.com/forum/showthread.php?t=55733
natesully
Experienced MS/Extra'er
Posts: 226
Joined: Mon Jul 25, 2005 9:13 pm

Post by natesully »

That is what I am considering. Also, an adjust valve to set how much air the solenoid flows, so I can get it perfect at normal intake temp. Right now, PWM idle is very stall-prone with A/C on, especially if the compressor kicks in at a really inopportune time.
95 Miata M-Edition, GT28 Turbo
cguenther
Helpful MS/Extra'er
Posts: 47
Joined: Sat Mar 03, 2007 5:41 am
Location: Germany

Post by cguenther »

hi,

the way mops described how to kick in some additional air is exactly the way it was already designed by the car manufacturer. but if you see the amount of rubber-hoses etc. for the k-jetronic together with this air-valve, youll have the idea to simplify this with a single pwm-valve.

im shortly prior first start with spark & fuel (for the time it was spark only), so this will be the next "highlight" to get this working.

best regards
carsten
Post Reply