JLD612 and temperature offset

[AndDroid]

Hi all, I'm hoping someone can shed some insight into my setup. Let me preface everything by saying I've used my sous vide setup once, and it worked perfectly. It measured almost identically to my separate Thermapen thermometer, and it held my water bath at 135 degrees for over 3 hours.

I'm using the JLD612 controller and this PT100 RTD: http://www.lightobject.com/Premium-Stai ... -P657.aspx

Today I turned on the machine, and the temperature reading was very low. It seems to be constantly reading about 25-30 degrees lower than the actual temperature. I know that there is the PSb offset setting, which I've modified and seems to help the temperature be accurate, but I'm still concerned with what may have caused this offset after working perfectly one time.

I have modified the probe cable by attaching a 3.5mm plug/jack in between the probe and the PID controller. I don't think the probe or modifications are faulty since, as I said, this setup worked once, perfectly. I've also measured the resistance of the probe at room temperature while connected to the jack, and resistance seems to be consistant with what I'd expect to see at room temperature (about 108-109 ohm).

If anyone has any insight into what may have happened to my setup, I'd appreciate it very much.

Thanks!

[richiem]

I'm assuming that this plug/jack combo is a "stereo" set, with three connections. The connections on such a plug or jack are conventionally labelled "tip," "ring," and "sleeve" from old telephone days. I would make the red wire the tip, one blue wire the ring, and the other blue wire the sleeve -- is the "sleeve" connection, which is the longest part of the plug and the outer rim/housing/retaining nut of the jack, making contact with the vessel or a ground of any kind? These three connections all need to "float" relative to any of the other components. If these three connections are not isolated from everything else, please fix that first and try again. You also can try spritzing the plug and jack with some WD-40, which will help prevent humidity and moisture from causing contact resistance problems.

The problem could be caused by a high(er) resistance contact (an Ohm would be high in this application) on one or more of the plug-jack contact points, which at the very low current of the 612, could cause a big change -- this contact resistance change might disappear if the current from the ohmmeter is sufficiently high to cause the connection to "make." I'll just venture a guess, though, and say that the JLD612 has a problem -- but it is very hard to be sure there are no intervening variables.

Just to wrap up the practicalities, the probe receives a constant current from the 612 on the red wire -- this constant current (roughly 200 microAmps) then flows through the resistance of the platinum sensor and out the blue wires, and this current flow the causes a variation in voltage across the probe's resistance between the red and the blues wires, according to Ohm's Law -- voltage = current * resistance. As the resistance increases, so does the voltage, which is converted into a temperature reading. The probe is part of an internal bridge in the 612 which helps to prevent the resistances of the probe wires from being part of the measurement. this lets the probe wire length be changed without upsetting the basic accuracy of the system.

The resistance reading you measured seems right, although very small changes in resistance make a large difference in the reported temperature.

[AndDroid]

Thanks for the input. It is indeed a stereo TRS connection. Currently I have the red wire as the sleeve, and the blues as the tip and ring. I plan on rebuilding the connector since the jack part I have is pretty low quality, and I'll take your connection advice into consideration when I do. All the connections seem fine, however I did notice that between measuring the resistance of the probe while it's connected to the jack vs disconnected from the jack gives me about a ~1 ohm drop in resistance (from 108.6 ohm to 107.6 ohm at room temp of about 70F), however I don't think this shouldn't give me a ~25F drop in the reading.

I guess I'll chalk it up to either a fluke that it worked the first time, or something has gone wrong with the PID controller (though that doesn't fit my gut feeling quite right), and be thankful that I can still use it, just with a little more setup of verifying the reading.

Thanks again for the input!

[richiem]

Right on temp variation -- a 1 ohm change will be about a 5F change. Re-do the connections on the plug and jack -- I would use a 1/4" TRS combo for easier wiring, or even a set of XLR microphone connectors. Then be sure everything probe-wise is floating.