Olds / Wildcat PID is ready to ship.

[iceman]

Hey guys. Sorry it took so long but your PID's are ready to ship. Send me a PM or e-mail with the address you want it sent to. Thanks.
Pat

[Wildcat]

Thanks Pat - will do so as soon as I catch up on reading these posts.

[bullsi1911]

Another newbie question... what's a PID?

[iceman]

Another newbie question... what's a PID?

Here is a definition that Arcs posted a while back....

"PID stands for Proportional, Integral, Derivative. It is a type of control scheme that does a bunch of math based upon the difference between what the temperature is, and where it should be. PID control schemes reach the setpoint fastest, with the least overshoot, and least error.

Having said all that, other control schemes work quite well. Your house thermostat is an on/off control that is very accurate (it helps that, to change the house temperature, it takes some time; there is a lot of thermal inertia)."

[pensrock]

"PID stands for Proportional, Integral, Derivative. It is a type of control scheme that does a bunch of math based upon the difference between what the temperature is, and where it should be. PID control schemes reach the setpoint fastest, with the least overshoot, and least error.

PID controls also require knowing the responsiviness of the unit being controlled. That's where the autotune comes in, you should simulate a load in the device when doing the autotune. What it does is: raises the unit temperature till it overshoots the setpoint, then it cools till it undershoots the setpoint. It will do this several times, each time it should overshoot/undershoot less. By doing this it learns how fast and slow the unit heats and cools, this is why it is necessary to simulate a load. Even after doing an autotune sometimes you still need to tinker with the actuall PID settings selected by the instrument. Changing the load and temperature should have little effect on a properly tuned controller. It takes time and practice to find the perfect PID settings. Even once you have the PID's properly set, it is not unusual for the temperature to overshoot the setpoint, then drop below the setpoint as the instrument zeros in on the desired temperature the overshoot and undershoot will keep getting less and less till it is controlling the best it can with the PID's in the instrument.

Now that I have totally confused everyone. PID's are difficult to try to explain. Even when you work with PID controllers every day. When I was first exposed to PID's, I asked a tech from the company where we purchase our controllers from to tell me what numbers to put in for each use. He laughed and said he has been tuning controls for 30+ years and still has trouble getting it perfect right away.

My advice is: If you are not happy with the way it is controlling. Either try another autotune or try tinkering with the actual PID numbers in the controller. I have personally done several autotunes on the same controller till it was close to what I wanted. But I normally play with the numbers once it gets close so I can fine tune it to my liking.

The higher the 'P', 'Proportional' number is, typically the slower the controller should react to a temperature change. This is just  a rule of thumb, I have never even seen one of the controllers you are using. Each controller has its own algorithm used to do the math functions. So if the unit seems to take longer than normal, all things being the same as before, to get to temp and the "P' is equal to say 7, then try to make it a 4 and see what happens. The most that will happen is you will have to do another autotune.

We rarely even use the 'I', 'Integral'. normally I just set this to a zero. It seems to take some of the jerkyness out of the controls. Most of the time I have an ampmeter clamped to the power wires to watch how it is reacting, you may not even see any differance.

Some instruments have a function, sometimes called 'Loadline' that you can use if the controller is functioning the way you want but it is maintaining a steady deviation from setpoint say + or - 20 degrees. By playing with the loadline you can keep the control but raise or lower the process temperature till it reaches the setpoint.

Well so much for the short lesson in PID controllers, now that everyone is confused...... keep staring into the pretty light.... do not notice the man sneeking into your smoker taking samples..... allow him to slip away into the wormhole with all your great food.

Good luck.

[iceman]

Bump..... You around Old's.

[Arcs_n_Sparks]

"PID For Smokers"

For those interested, let me elaborate on the quote provided by Iceman. The PID is a closed-loop control scheme that looks at what you want (setpoint) versus what you are getting (measured parameter). This can apply to many processes, but we'll stick to temperature.

The difference between what you want and what you have is an error term. The error term, in this case temperature error, is used to adjust how much power is applied to a heating element. Now, here comes the fun part. The P term is a multiplier to the error; it represents "gain" or amplification of the system. Low P represents small adjustment to the error, large P represents big adjustment. For cruise control, a 5 MPH error could be a soft acceleration or having the engine floored depending upon P.

Both courses of action close the error, but either slower or faster (rise time is the term used). Now, hitting the setpoint may result in overshoot depending upon P. If you keep it floored until the error goes to zero, you will be over the setpoint as you lift your foot off the gas (or electricity for us doing heating). If you gradually applied gas, you will not overshoot, but may not hit the setpoint as quickly as you might otherwise. This term is overshoot or over damping, depending upon upon being conservative or aggressive. If you overshoot, you will cycle back under and oscillate under you reach steady-state (this is settling time).

This approach works fine, but has a mathematical problem. You can have a residual error that is not zero, but the system will not correct if it is purely a Proportional control system. This is where I (integral) comes in. It will sum the residual error term over multiple control cycles to the point that it will correct the error to zero. This helps settle out the error (settling time) and reduce it to zero (steady-state error) At this point, we could call it good. However....

Some like to take advantage of knowing what the user is intending when changing the setpoint. This is the Derivative (D) term. The D term is used when the setpoint is changed relative to where it was, and signals the control system to react appropriately.

All these terms: P, I, & D are multipliers to the error term and are used to control the process parameter, in this case power to the heating element. All affect rise time, overshoot, settling time, and steady-state error. A tuned PID will reduce all of these to the best and optimal value.

For more detail, this is a very useful tutorial: http://www.engin.umich.edu/group/ctm/PID/PID.html

Whew....Now I need another drink.   8)

Arcs_n_Sparks

[Ontrack]

Uh...OK... :-\

[Arcs_n_Sparks]

Uh...OK... :-\

Ontrack,

I'm with you: forget all that complicated cr@p and just let the electronics do their thing. In the mean time, select the next adult beverage. In this case, explaining the operation of the Porter Indicating Device would have been the better tutorial.    8)

[Ontrack]

Arcs, that I can do! How about Wild Turkey Rare Breed (108.6) and Coke? Just happen to have one within arm's reach...probably the last one of the night. 11:20pm, been up for many hours, and eyes are beginning to close... 8) and go with the porter!

[Arcs_n_Sparks]

I second your motion. All in favor? (Aye). Those oppose? (silence). Motion passes.

Glad some of this stuff is still simple...   ;D

[KyNola]

With all due respect to you guys, what in the wide wide world of sports are you talking about?  ???  I fully understood the drinking part but beyond that, I am a total moron.

Ontrack, love to ride the train from Illinois to New Orleans.  Cool way to travel. 8)

KyNola

[Arcs_n_Sparks]

With all due respect to you guys, what in the wide wide world of sports are you talking about?  ???  I fully understood the drinking part but beyond that, I am a total moron.

Well, I cannot comment on the second point of your question...  :D However, I thought Ontrack and I were talking about:

P: the Proportion of alcohol (Wild Turkey) being distributed within a mixer (in this case, Coke)
I: the Inebriation factor being considered for the evening. Note: higher on weekends
D: the Distillation factor. In this case, the more select Rare Breed at 108.6 proof stuff

I think this topic is getting overly complicated....   :D :D :D

[Wildcat]

Thanks all for explaining how it works.  I think I have a basic understanding of how it works.  Hopefully I will not have a problem making it work when mine gets here.  Would not want to try to build one sober (or un-sober for that matter) though.  Electronincs is not my fortitude.

[DrunkenMick]

I so want to build a PID but me and electrical things usually don't get along too well.  I'm seriously considering just buying this thing http://auberins.com/index.php?main_page=product_info&products_id=44&zenid=23711349586570e08e567c062108a9b1.  Seems to just be a pre-built PID.  Not sure if it has the auto-tune features though.  Might be a deal breaker if it doesn't.