What kind of Mash temp thermometer?

[AnOldUR]

. . . but they integrated circuits that compensate cost next to nothing today and are very accurate. The two I have are accurate to less than a degree at 32F (ice bath) and 212F (boiling water).

It's nice to have comments from someone in the industry. This leads to the question, what two (make/model) are you referring to? What do you think of the Comark DT400 that Pol mentioned above?

 

[ForRealBeer]

RTD's are not necessarily that accurate. It's old technology. I used to design interfaces for them for high speed data aquisition. They were liked, IMHO, because they produce a linear relationship between temperature and resistance, and they were rugged in industrial applications and you could do long, remote lead lengths with 3 and 4 wire systems. Easy to design for for most applications, where +/- 5 degree absolute accuracy is more than good enough. One can get better accuracy, but platinum is not cheap and 4 wire versions are a bit more pricey.

Huh? Not to be argumentative, but my experiences with PRTDs were quite different than yours. Like you, I designed and implemented DACs - for laboratories and manufacturing settings in situations where temperature knowledge was critical to our control processes.

The industry standard for platinum RTD's according to DIN IEC-751 is Class A or B, that is, a typical accuracy for Class A ±0.15°C at 0°C, and ±0.35°C at 100°C, with ±0.3°C at 0°C, and ±0.8°C at 100°C in a Class B setting. This is quite good when compared to the ±2.2°C of a standard Type J or K thermocouple. And as you state, the DFL - dead frakking linear.

You may recall accuracy:

Class A: Dt °C = ± ( 0.15 + 0.002 • | t | )
Class B: Dt °C = ± ( 0.30 + 0.005 • | t | )

where: | t | = absolute value of temperature in °C

As for two-wire versus four wire, the two wire setting is subsceptible to resistance changes in the lead wires and/or connections, whereas the resistor bridge in a four wire setting offsets and nulls that out. I never bothered with two wire setups, given the ease of implementing four-wire.

Another reason the PRTDs are well liked in industrial and laboratory conditions is that cycle to cycle differences normally can’t be measured and are considered lumped into stability specifications. Further, you can typically you can count on the probes to be almost totally non-reactive. Some J and K thermos were pretty weak in that regard and were not long lasting. Not only that, they tend to require relatively constant calibration in my experience.

 

[RichBenn]

Huh? Not to be argumentative, but my experiences with PRTDs were quite different than yours. Like you, I designed and implemented DACs - for laboratories and manufacturing settings in situations where temperature knowledge was critical to our control processes.

The industry standard for platinum RTD's according to DIN IEC-751 is Class A or B, that is, a typical accuracy for Class A ±0.15°C at 0°C, and ±0.35°C at 100°C, with ±0.3°C at 0°C, and ±0.8°C at 100°C in a Class B setting. This is quite good when compared to the ±2.2°C of a standard Type J or K thermocouple. And as you state, the DFL - dead frakking linear.

You may recall accuracy:

Class A: Dt °C = ± ( 0.15 + 0.002 • | t | )
Class B: Dt °C = ± ( 0.30 + 0.005 • | t | )

where: | t | = absolute value of temperature in °C

As for two-wire versus four wire, the two wire setting is subsceptible to resistance changes in the lead wires and/or connections, whereas the resistor bridge in a four wire setting offsets and nulls that out. I never bothered with two wire setups, given the ease of implementing four-wire.

Another reason the PRTDs are well liked in industrial and laboratory conditions is that cycle to cycle differences normally can’t be measured and are considered lumped into stability specifications. Further, you can typically you can count on the probes to be almost totally non-reactive. Some J and K thermos were pretty weak in that regard and were not long lasting. Not only that, they tend to require relatively constant calibration in my experience.

I agree RTDs are better than thermocouples. The poster I was replying to was getting terrible results with his RTDs, which were probably either faulty or the resistance measurement circuit is bad, or ???

When I designed circuitry, it was such that it would have a margin of error good enough for the worst case component aging over a 5-10 year period and at a cost as low as possible while still meeting the needs. So even though the RTD itself might be highly accurate, the rest of the system might be far less, including the tolerances and aging of the resistors, transistors and capacitors used to measure the resistance to the thermal coupling of the RTD or thermocouple to the product being measured. But it depends upon the application. My products were in the turbo-machinery field, measuring case and bearing temperatures. A few degrees was good enough. Alot of process control requires much better accuracy than I would typically be required to deisgn for.

My point, and I was hoping to not get into much detail, was that cheap thermistors are OK for our application (home brewing, not professional process control), not to argue the merits of RTDs vs. thermocouples.

For 140-180 degrees, in the range where we mash, a thermistor device can very easily have .5 degree F resolution for $20. Here is one example here:

Taylor USA

although there are many more.

 

[ForRealBeer]

Rich, agreed about thermistors. Like I said, I absolutely am not trying to be argumentative, but as you might have guessed, electrical engineering and instrumentation are real passions of mine. You and I could probably have some very interesting discussions about these things between the two of us over a couple of our homebrews... but the average reader here would lose interest quickly. What the heck, use the best stuff you can afford, do it the best you can and enjoy the beer. It'll be good.

Also, please let me say I hear 100% about what you are saying about cost containment in your work. I was lucky in that in my company's heading days of Big R and little D for stuff coming out of blue sky, we tended to have bigger budgets and in the case of materials engineering we were doing, no corners were turned. Nowadays, we're as cheap as anyone and I am the curmudgeon in the corner saying "are you sure that what you measured is accurate?"

Anyhow, Happy Thanksgiving and look forward to talking to you about this some more sometime.

 

[RichBenn]

It's nice to have comments from someone in the industry. This leads to the question, what two (make/model) are you referring to? What do you think of the Comark DT400 that Pol mentioned above?

I'm using primarily a cheap Taylor wine thermometer, but have other cheap thermistor based products I use around the home.

The product you reference seems to have a 1 degree F accuracy. If the repeatability is good, then it seems good enough for me, although maybe not for others. The parent company of Comark is Fluke, which had a good reputation, AFAIK. But no experience at all with this product or Comark.

Hey, and for anyone else that wants to argue the merits of more expensive vs. cheap measuring devices -- this is my opinion and what works for me, a guy on a budget who likes deals. Alot of people have different values, like having the best, and I have little doubt the Thermopen is a better device.

Reliability? I think there are other threads here that one may want to peruse for user experiences with various devices.

Rich

 

[RichBenn]

Rich, agreed about thermistors. Like I said, I absolutely am not trying to be argumentative, but as you might have guessed, electrical engineering and instrumentation are real passions of mine. You and I could probably have some very interesting discussions about these things between the two of us over a couple of our homebrews... but the average reader here would lose interest quickly. What the heck, use the best stuff you can afford, do it the best you can and enjoy the beer. It'll be good.

Also, please let me say I hear 100% about what you are saying about cost containment in your work. I was lucky in that in my company's heading days of Big R and little D for stuff coming out of blue sky, we tended to have bigger budgets and in the case of materials engineering we were doing, no corners were turned. Nowadays, we're as cheap as anyone and I am the curmudgeon in the corner saying "are you sure that what you measured is accurate?"

Anyhow, Happy Thanksgiving and look forward to talking to you about this some more sometime.

The same to you!