1. Home
  2. Electronics Repair
  3. Thermistor Puzzle

Thermistor Puzzle

Given all this free time, I'm fixing stuff around the house.

I have a Brookstone Grill Alert. Stick a probe into a chunk of meat, plug the cable (2.5mm phone plug) into the box and it happily monitors my steak on the barbecue. Except it stopped working with unreasonably high readings (150F reading at room temp). Wisdom on the 'Net says that the probe is probably broken. But the replacement part (Brookstone) is no longer available. No problem. I'll plug my decade resistance box into the grill alert, start cranking in some resistance values and figure out the thermistor curve. Then find another probe.

It appears to expect an NTC thermistor with 180K ohms giving an 80 F reading and 60K ohms reading 125F. But then I pick up the old (broken?) probe and measure it. At room temp (65F) it has a 220K resistance. Holding the probe in my hands drops its resistance to 140K. Seems to be OK. I wiggle the cord, expecting an intermittent short (low resistance giving a high temp reading). Nope. It's fine. Maybe it's something wrong with the plug/jack interface. So I open the box, unsolder a lead from the jack and measure the probe resistance plugged in at the back of the jack. It reads 220K. Hook it all back up and it's still wonky (probe gives a 150F reading sitting on my desk). But if I plug my decade box in to the grill alert and dial it to 220K ohms, the thing reads room temp.

The Grill Alert appears to just not like that probe. What have I missed?

--
Paul Hovnanian     mailto:Paul@Hovnanian.com 
------------------------------------------------------------------ 
Not all those who wander are lost. Except for the 2nd Lt. 
He is lost as shit.
Reply to
Paul Hovnanian P.E.
Loading thread data ...

My guesses, officially classified as S.W.A.G.:

a) Clean the battery compartment, and replace the battery with a fresh one. Analogy: An audio speaker has a resistance that may be measured by a VOM, but in use, it has impedance that varies by input frequency. Could it be th at the probe does the same under actual use? If so, available current will be critical for accurate readings. And these things live in a harsh environ ment.

b) Cold-solder or a component is intermittent. Again, these things live in a harsh environment, and something could have worked loose. Fine when sitti ng nicely on a bench, not so fine in use.

c) Try the probe in a pot of water. Bring it gently to a boil. It will neve r get above 212F, but you should be able to follow the change in resistance as the water heats, and it should be as gradual as the heating. The differ ence between a test at fixed points, and a test of behavior over actual ran ge-of-use conditions.

I would start with c) first, as it will be the most immediately revealing.

Best of luck with it!

Peter Wieck Melrose Park, PA

Reply to
peterwieck33

It might be that there's some oscillation at the sense amplifier; the impedance of the probe wiring might be the issue.

If so, good news; that means it's not gonna need a new probe. Bad news, you have an oscillating consumer device to debug.

Reply to
whit3rd

Could it possibly be a semiconductor probe that uses the forward voltage drop of a diode?

--
~ Adrian Tuddenham ~ 
(Remove the ".invalid"s and add ".co.uk" to reply) 
www.poppyrecords.co.uk
Reply to
Adrian Tuddenham

contamination inside the probe, subject to position of probe (vertical / horizontal, which jiggling the *wire* might not have changed). Grind off the crimp at its neck or whatever holds the wire, pull out sensor.

Reply to
malua mada!

I chased down the curve of a microwave cooking probe. The way I did it depended on the temperatures of interest.

most temps Stuck good and bad probes in a potatoe Used peanut oil and the stove

Low temps Stuck good and bad probes in a potatoe Water and added ice cubes

I got an excellent curve match. It also turned out, that a random probe I purchased had the same curve.

The bad probe actually had an intermittent contact. The actual readout was broken at the time.

Reply to
Ron D.

I checked the probe with both polarities of my ohmmeter. Also used the diode test function to look for this.

The thing is that the temperature the device displays seems to work well with a resistance decade box. And all the parasitics that introduces.

--
Paul Hovnanian     mailto:Paul@Hovnanian.com 
------------------------------------------------------------------ 
Only through suffering comes wisdom. -- Zeus
Reply to
Paul Hovnanian P.E.

Most failures are from wire fatigue, either at the neck of the probe or at the neck of the plug. Reasonably easy to fix, never really lasts though.

Reply to
malua mada!

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.