JLD612 overshoots

[robinhood]

I am trying to get my JLD612 to hold a constant 100°F after entering Pv and Sv values the Pv values wonders all over the place while heating,,
I am sure I got something screwed up..here is a photo attached to see...any ideas Pleeeeeaaaaaseee...

[adrenalynn]

After seeing a ton of overshoot on my hot-plate-heated bath, I found an article that basically has a lot of P, and not much ID. Their specific heat was the same and their heat capacity wasn't far off from mine, their graphs looked convincing, so I gave it a try. Worked well in testing. Some initial overshoot as the controller "caught its stride" but within a half hour or so it flattened out.

Yesterday was my event, and I used that revised programming at the event. From 7am to 2:15pm, it maintained 63C +/- 0.5C. Acceptable for my application.

Here's that article: http://qandabe.wpengine.com/2011/revise ... libration/

[robinhood]

Thanks for your help...I'll give it a shot..

[richiem]

My first rule for problem solving is: Check the settings, especially Inty and Outy.

After that, auto-tuning *may* provide the best result.

As noted above by adrenalynn, P is a good thing. P -- proportional control parameter -- works by making large changes when the PV is far from the SV, then makes proportionally smaller and smaller changes as PV converges on SV.

In the JLD612, the proportion is the duty cycle -- the ratio of on time to off time -- of the controlled element, usually a heater. As PV approaches SV, the controlled element is on for shorter and shorter periods, and is off for longer and longer periods. So for most heating apps, the P part is the dominant part of the control algorithm.

The I, or integration parameter, controls the amount of time that information about PV is gathered as an average value over the sampling period. This helps smooth out small or rapid variations in PV that essentially constitute noise in the system. But large I values, which slow down the system response time, make the system insensitive to large sharp changes in PV.

The D, or differential parameter, looks at the rate of change of PV. If the change is very sharp and fast, it forces the system to respond, and is able to compensate for the slow response forced by the I parameter.

Obviously, good control requires a balance of all three parameters, and that balance is highly dependent on the characteristics of the system being controlled. For example, a big thermal mass being heated takes longer to heat, so having a control system that tries to force fast small changes will be useless. Same result if the heat applied is small in comparison to the thermal mass being heated.

The opposite is true if the heat applied is large and the mass being heated is small -- here, the temp changes rapidly when heat is on, so the control system has to respond very quickly or large overshoots/undershoots occur; or the system completely fails to stabilize (fails to converge) and instead oscillates around the SV.

The 612 manual shows these factors graphically, which is helpful, but doesn't give any practical hints as to good parameter value ranges for different applications. That's why just messing around can be informative; but it's a slow hard way to get to a good result.

[robinhood]

richiem, Thanks a bunch will work on it...I have a small strip heater I want to heat up to 100F..I almost had it working once...
Thanks again..