building a temperature control unit for mercury pot

[christian_rice]

For my purposes, does anyone know if it matters if a thermocouple junction is grounded or ungrounded? Is it just a matter of speed of measurement, or is the fact that we’re submersed in an electrically conductive liquid a factor?

Currently I’m planning on using a K-style 1/16″ diameter, 6″ long stainless steel thermocouple with 24″ leads.

[jgmotamedi]

I use the grounded version of the K-style thermocouple from Omega (TJ36-CPSS-18G-6-SB), and it works very well. I had a conversation with an electronics engineer before buying it, and decided on the grounded version, but can’t recall why.

[christian_rice]

Thanks a bunch. From what I read, the grounded unit responds to temp variations faster–that should tighten the feedback loop.

I ordered the following from omega.com:

$20 KTSS-116G-24 thermocouple (K type, stainless steel, 1/16″ diameter, grounded, 24″ leads)

$26 SRFR-3/5 silicone/fiberglass heating disk, 3″ diameter, 5W/inch^2

$97 CN7533 1/32 DIN controller

I’m wondering if you used an external relay, or connected your heater directly to the controller. Perhaps the watt density of the heater also matters?

[jgmotamedi]

With the CN7533 you should be able to drive a 3″ heater using the internal relay. I made a mistake and ordered the CN7523 and was forced to use an external relay. Not a big deal, I picked one up at Radio Shack for $4.

I have been very happy with my system, it keeps the mercury within .2*C in the autotune mode. It took me a while to figure out the initial programing, the manual was pretty confusing and largely inadequate, but with some help I managed to set it up.

[christian_rice]

Once again, very helpful. I downloaded and read the manuals before buying, but what I’ve got from you is orders of magnitudes more useful for me.

Thanks!

[jgmotamedi]

I will see if I can find my notes on programming the controller. I know I wrote down all my settings in a notebook, but that was before the move, and I haven’t seen the notebook recently.

[christian_rice]

I managed to get the unit wired and programmed such that it keeps a steady temp without use of an external relay. I’ll provide some lessons learned for the record…

 

For anyone following this thread–don’t buy the thermocouple I mentioned earlier–I thought I was getting 24″ lead wires, but I got a 24″ probe lead. Ugh, ridiculously long probe. I consequently ordered a TJ36-CPSS-18G-6-CC instead, with a 6″ probe. It’s a type T, which is for temps much closer to what we’re doing, instead of a type K. That’s one small change of program on the controller, hopefully I’m getting the right thing this time.

 

Something one needs to know to make this work is that the relay function on the CN7533 completes a circuit from terminal 9 to terminal 10–the CN7533 does not internally provide power to any terminal lugs. I’m sure this is totally obvious to electronics buffs, it being a relay and all, but I was wondering about that for a little while. The controller instructions are not going to help an abject novice build anything useful.

 

I included a switch on the positive leg of my power cord, as per manual recommendation. Basically I bought a 12′ extension cord for $3 and an in-line rocker switch for $4 from Home Despot. I discarded the female end of the extension cord. The positive wire goes to terminal 1 and terminal 9. The neutral wire (aka “wide blade on a US power plug”) goes to terminal 2 and one wire of the heater element (in reality I used an additional terminal block so I could avoid splicing and wire nuts in my enclosure). Terminal 10 connects to the other wire of the heater element. The thermocouple connects to terminals 4 and 6 (omega.com has a reference page that illustrates the polarity color coding of their thermocouples).

 

Terminals 9 and 10 are OUT1. In the initial setup, OUT1 should be set to HEAT, which means when PV (process variable = temp measured) is less than SV (set variable = target temp), the relay closes, and the heater becomes active.

 

The actual opening and closing of the relay is further affected by the operation mode. PID autotune mode is very helpful, especially because there’s a lot of mass in my mercury cup heater connection assembly, especially relative to the volume of mercury being heated. The controller has to account for the fact that relay operation is not linked in real time to the mercury temperature–there’s a significant ramp lag. Autotune mode does some real math in the background, so “magically” makes this work.

 

I used a small ABS plastic project box from my local neighborhood electronics store. The heater pad leads were only about a foot long, so I soldered some extensions on there, and used shrink tubing to keep the wires neat. I wanted to keep the controller unit outside my portable fume hood, so it would be less likely to be attacked by whatever is in there, for what that’s worth…

 

Cheers.

 

[Andy Stockton]

Christian-

Thanks so much for writing up your experience on the forum. Great entries like this save others who follow in your footsteps so much time and wasted effort!

Andy

[Andy Stockton]

Christian-

Thanks so much for writing up your experience on the forum. Great entries like this save others who follow in your footsteps so much time and wasted effort!

Andy

[Mercury]

Christian,

It looks like you’ve exactly replicated my beloved “Dr. Fuzzy.” When I first built “Dr. Fuzzy” in 1992 the fuzzy-logic PID circuitry was pretty new. I was amazed at how it ‘learned’ the ramp-up for my Hg pot, and would get there in minutes without over-shooting. (I have a fairly thin-walled Hg pot, which has the advantage of being a poor heat-sink, which in turn challenges the PID controller to work harder, more cycles per minute, so to ‘learn’ the process more quickly.)

As was chronicled in CDags, my first 1/4 DIN “Dr. Fuzzy” died suddenly after more than a decade of dedicated service. Expect that. They don’t last forever.

My PID controllers have come from Love Controls, now a subsidiary of Dwyer Instruments. They are great to deal with. I encountered all the customer service questions about Manuals and connections that you did, and Dwyer was very helpful. So, I’m glad you’ve had a good experience with Omega; I’ll be glad to add Dwyer to the CDags list for recommended temperature controller suppliers. (BTW, the Dwyer output contacts are the same as for your Omega controller… contacts 9 and 10, but all the same, make sure of your polarity!)

Ken

[Author]

Posts