Thermometer (In)Accuracy... (Briefly Revisited)

Yesterday, I received three new stove thermometers ordered directly from the manufacturer. The single flue probe I've been using is about twenty years old, and not only did I think it might be time to replace it, but with several monitoring points on my stove, I thought it would be of interest to compare simultaneous readings at the different locations.

But first, I put all the thermometers - old and new - in the kitchen oven at 250*, along with an oven thermometer as well. Here (approximately) is what I got from each:

Oven - 225*
(New) stovepipe surface - 300*
(New) cat probe - 500*
(Old) flue probe - 600*
(New) flue probe - 700* (+)

In actual use, the thermometers also provided disparate results and different readings when swapped from one stove location to another for direct comparison. Even the two flue probes - old and new - didn't agree in temp ranges over 500*.

Needless to say, these 'findings' inspired a *whole* lot of confidence... and I have NO idea which thermometer I can actually trust... if any.

Yes, I was aware that the thermometers would not be perfectly accurate, and that only relative readings from a single thermometer in a single location were really meaningful for stove operation, but I sure didn't expect such a wild range of readings.

I would have spent my money better either buying two <identical> thermometers or a single IR to make periodic readings to compare with the old flue probe.

Lesson learned. I've reverted to simply using the old flue probe... because at least it provides 'familiar' readings.

Don't know offhand what to do with the others...

Peter B.

-----

BeGreen said:

Wow, that's certainly confidence inspiring. I bought a cheap Drolet surface thermometer from our local True Value for a friend last year. According to my IR thermometer, it checked out with 20 degrees at 550.

Have you confirmed the accuracy of the oven thermometer? Have you contacted the manufacturer of these thermometers?

Click to expand...

I don't have a method of verifying the accuracy of ANY of the thermometers at present, I'm afraid.

No, I haven't contacted the manufacturer... yet. I'm not out a huge pile of cash, and I wonder if it's really worth doing so... but I confess, I'm not real happy with the products supplied, when I had previously trusted their name (deliberately unmentioned here).

--

Edthedawg said:

I wonder more about the disparity between Tmetal and Tgas. Were the measurement contact points of the probe thermometers hanging in air? Or resting on a sheet? on a rack?

Click to expand...

The oven test was done with all thermometers resting on a metal sheet (with probe tips touching the sheet)... and you're right, that may have provided some bogus results, but as an example of actual use disparity, in the same location, the new flue probe and the new cat probe read about 300* apart. The stovepipe surface thermometer reads (roughly) half of the flue probe thermometer at the same height (to be expected ???).

--

EngineRep said:

This makes sense to me. Perhaps the very high reading was not incorrect, it was that the probe had in fact heated to that level by absorbing radiant heat of the calrod.

The best way to calibrate (although it is for only a single temperature) is to suspend the probe into boiling water. It must be suspended so that it doesn't pick up heat from the container. Obviously you're looking for 212F (minimal variation for pressure altitude).

Click to expand...

Yes, I could try to verify that specific single temperature point, but if there is such wide variation in readings at other temperatures, it's not of that much use.

Peter B.

-----

Jags said:

Get enough of them and you can use the little round ones for poker chips. Eh, just trying to help. :shut:

I get my confidence by using the IR thermo pointed at my little round rutland. They read within ~20 deg F of each other at almost any temp that I can expect. Do you have any burning friends that may have an IR to borrow?

Click to expand...

I guess it's (long) past time I bought one of my own...

I know there are dozens of threads here on recommendations (and sources) for IR's, but if no one minds too much, I'll ask again. (From another thread of my own, I learned that the less expensive ones won't read much beyond 1000*, but I'd like to be able to do that at least.)

Thanks Much.

Peter B.

-----

For what it's worth, I thought I'd provide this further observation...

I 'tested' another set of thermometers in the kitchen oven at various temps and discovered that probes don't (remotely) reflect accurate temperatures when fully enclosed within a heat source. On the other hand, stovetop/stovepipe (magnetic) thermometers should read reasonably closely to the oven temperature.

Again, probably <none> of the inexpensive thermometers we use with our stoves are likely to be dead accurate, but can only provide relative reference temperatures... unfortunately.

I'd prefer better overall accuracy, but I guess it's not available on the cheap.

Peter B.

-----

Just to clarify...

I've owned a similar flue probe for twenty years and have relied on it to provide reasonably accurate readings... though I've never had a way of calibrating it.

In actual use (not kitchen oven tests), one of the new probes reads about 100* less than the old, while two other new probes (one of them essentially identical to the old) read 200*-300* higher under the same circumstances... usually at temps above 500*... a rather important range for monitoring purposes.

At this point, I have no idea which - if any - of the thermometers to trust as 'accurate'.

So, the old probe with the 'familiar' temp references remains in service.

--

Why so many thermometers in the first place?

I had hoped to find two that closely agreed, so that I could monitor two different points - stovetop internal and flue internal (at 12" above stovetop) to make some sort of meaningful comparison of temperatures at those points.

I can now do that with two of the new probes, but their actual accuracy is still very much in question.

Plus or minus 5% error through the range is NOT what I'm seeing.

Again, they still provide <relative> references, but accuracy is what I was seeking.

Peter B.

-----