Tweet
You are a trailblazer that's making it easier for the rest of us. Thanks.
-
I'm still learning -
I'm try'in! =POriginally posted by Bob Holmes View Post
Also, I turned on the steering wheel to load up the engine, and it loaded up and the rpm dropped for about a second and the IAC opened up a tad and bumped the idle right back to 900. Pretty cool.Escaped on a technicality.Comment
-
PID? should have asked my kids...they know all about PID loops. robots, you knowMy fabulous web page
"If it don't go, chrome it!" --Stroker McGurkComment
-
PID is Proportional, Integral, and Derivitive - three stages of control for a process control system. I used the system particularly for throttle/rack control on our dyno engines that were following a transient dyno cycle (constantly varying speeds, loads, and therefore throttle settings). It's been a LONG time but here's what I can remember:
P= rough setting to (in our case) pull the rack to roughly the anticipated position for the upcoming point
I= fine tuning based on the error signal between the anticipated setting (say, 100 ft/lb) and the feedback (say, 105 ft/lb)
D= using the P & I to anticipate future adjustments even if the settings are not the same values - essentially learns how the system acts regardless of the specific desired setting
As I said, this is foggy as I haven't used this in YEARS.
An alternative to PID is fuzzy logic which we never got to use when I was involved with this. Maybe Jim or others will have some input here.
Hope this helps
DanComment
-
Damn, Dan is further ahead of the curve than me!! I got some catching up to do!
BTW, that is a good description of each! Thanks Dan
Escaped on a technicality.Comment
-
Shucks, twarn't nothin' - just what I did for a living. We were having a ton of trouble with this and the Gov bought me a book on the subject. Last I knew there isn't a "PID Tuning for Dummies" but there would be a small but eager audience, for sure. The book I read and tried to absorb was a sure-fire cure for insomnia and I was INTERESTED in the topic. Between me and our super electrical engineer we beat on it until it gave up.
DanComment
-
PID for high school kids
http://www.chiefdelphi.com/forums/sh...ad.php?t=37955
but they're definitely not dummies!My fabulous web page
"If it don't go, chrome it!" --Stroker McGurkComment
-
Originally posted by TheSilverBuick View PostAlright, I think I'm finally happy with the auto idling settings in the Skylark. I was having issues with it taking forever to lower the idle, which after every burn/update to the MSII/extra it opens up the IAC all the way, and having it sit there running at 2,000rpm for 30+ seconds sucked. But now it starts dropping rpm right after it peaks and settles into a 900rpm idle pretty decently with enough room on the IAC to be able to lower the rpm further for when I go to low elevation.
I had to use google to figure it out, I didn't even know what PID stood for nor what it did, and it turns out to be critical to the idling. I still am not perfectly clear on it, but I figured out settings that made it 1000x better.
These are the three links that I think helped me out the most. It seems quite a bit of the MS2/extra coding is similar to the MS3 coding...
http://www.msextra.com/forums/viewto...?f=101&t=40371
Translation: It starts and idles good in the morning.BS'er formally known as Rebeldryver
Resident Instigator
sigpicComment
-
Another memory cell has fired its synapse and I remembered this (no clue if it's useful to you). In a manual PID loop (no computer self-tuning) you turn the "D" all the way down (essentially off) and same with the "I". Get the "P" to respond accurately in a steady state situation. Then add in "I" until it can react when the system is disturbed - the goal is to get the system to return to the set point without much ringing (searching up and down until the set point is finally found). "D" may or may not be needed depending on what you want the system to do but if needed add slowly and see what happens when the system is disturbed. Too much "D" will result in either an unstable system or sluggish response depending on the system.
My usual disclaimers apply - these are old memories from long ago.
Hope I fogged this issue fully. Are all asleep out there?
DanComment
-
Originally posted by DanStokes View Post
Dan, great info on tuning PID loops!! There are guys who are professional PID tuners out there. Next time I see Doc McEntire I am going to try the closed loop PID idle like Randal is running. Right now I just have his IAC set a fixed openings based on coolant temperature (fast idle for warmup)...
Good info - again, thanks!!www.realtuners.com - catch the RealTuners Radio Podcast on Youtube, Facebook, iTunes, and anywhere else podcasts are distributed!Comment
-
Is this still the Buick we're talking about here and not the space shuttle ? .Comment
-
Yeah, PID control loops are used for a lot of stuff....cruise control is a very common example.
The I and D are terms from calculus--integral and derivative. Integral "adds up" the small errors, derivative finds the slopeMy fabulous web page
"If it don't go, chrome it!" --Stroker McGurkComment
-
Yup, still a late 70's BuickOriginally posted by langleylad View Post
Escaped on a technicality.Comment
-
At this point it has more computing power then the original Apollo rockets.Originally posted by langleylad View Post
Dan, we have members here display fuzzy logic all the time.
I'm still learningComment
-
And I'm chief among the fuzzy logic brigade!
Scott's right about those pro PID tuners. Those are exactly the guys we didn't have the budget to contract with so Mike (the EE) and I were stuck figuring all this out. Before we took it on as a project we were just turning pots until something happened and we could never get back to where we started.
The new dyno control systems cheat big time. The whole deal is computer controlled and set the PIDs based on sensor feedback - no humans involved. But that assumes sensors and enough computer power to run the system as well as the 11 billion functions needed to make the test run. The new systems HAVE the computing power plus a ton to spare. Technology marches on.
As Jim said, there's way more of this stuff going on than we know about as it's in the background just making stuff like cruise control happen. I understand that the new systems that control the wipers based on the amount of rain hitting the windshield also use this logic. Or automatically putting up your convertible top (what, you don't have a Mercedes convert??!!).
DanComment
Comment