View Full Version : Phidgets Thermocouples freaking out...

01-24-2008, 06:25 PM
I use 2 Phidgets thermocouple controllers connected to standard oven thermocouple probes to control my BBQ smoker. While I've questioned the results at times, this last weekend really went nuts. When I fired the smoker up, it was 2 degrees F. The temperature readings on both probes freaked out and read over 600 degrees F (in the cold ambient air......no heat yet at this point). I warmed the phidgets devices with my hand and the temperature went down to normal. So I wrapped them with a sock and all was well (actually freaked out again later, but was managable). So I ask you, 1.) Is this normal. I'm guessing so because it happened to both. I normally don't run the smoker in "snot cicle" weather, but needed to this weekend. 2.) and this is a more serious question: does normal ambient temperature change on the phidges devices effect the results of the thermocouple probe? I only ask this because I've seen strange results that I couldn't explain before....including burning the crap out of some fudge I was trying to make, using the probes to measure the temperature of the molten chocolate. I knew it didn't look right, but the temperature on the computer looked good. Oh well, anyone have any ideas?

01-25-2008, 11:43 AM
Hey elcray!

So, you only seen this happen when it was really cold out huh? I haven't heard of any problems like this before, but I emailed Phidgets to ask them. You're second question I have not clue about...Phidgets should get back to me soon.

01-26-2008, 12:05 AM
Thanks for the help Alex. I've got some milage on a thread over at the Phidgets forum. If/when we get a solution, I'll post the details here so your users can have a heads up as well.

02-01-2008, 01:27 PM
any word on this elcray? I'm curious to know what had happened...

02-03-2008, 04:20 PM
any word on this elcray? I'm curious to know what had happened...

This was the basic response from the Phidgets guys:
"The ambient sensor is an IC on the board - but the temperature that matters for compensation in the copper thermocouple connector - since these are not actually connected, only close to each other, you just need to make sure the board as a whole is at a constant temp - but it shouldn't matter what that temp is (within reason) - below freezing should be fine, though we haven't actually tested this"

Doesn't really answer why the low temperature caused the problem, but they did validate that the boards need to be insulated. I have since wrapped my boards in pipe insulation and stuffed in a project box. They seem to run stable now. However, we since haven't had the extreme weather we were dealing with when the problem occured. I have mounted the box in my freezer and seem to run fine.

The issue above also explains why I had bad results with my fudge disaster. Since the final temperature can be effected by changes in the onboard sensor, and since I had the board sitting on the oven top, as the ambient temperature increased, the accuracy of the probe was effected; theoretically speaking of course, as I haven't had a chance to test this.

11-10-2008, 03:39 PM
Just as a follow up to this topic from the days of old, it appears that Phidgets did in fact recognise a problem with the thermocouple controller and have made upgrades to correct them. For those interested, Phidgets has told me that there are at least 2 ways to tell the difference:
1.) the new devices come shipped in a box, and the old ones are in a bag
2.) The new devices are firmware version 300+, and the old one's are 200+

Unfortunately, the 2 new devices I received on Friday came in bags, and are therefore the old versions. So I can't confirm that the changes made will resolve the problems outlined at the top of this thread. If someone else can, it would really be helpful for me to know

Josh (aka Geek With Fire)