Thermocouple accuracy question

nelstomlinson · Feb 19, 2020

I found a source that explains it: https://infogalactic.com/info/Thermocouple#Type_K

The short version is that the thermocouple voltage is produced by the temperature difference between the hot and cold ends. CDX, I think you were saying that the thermocouple assumes that the cold side is 70F?

laserjock · Feb 20, 2020

nelstomlinson said:
Digital gauge. I see your problem.

You would like to know temp to the nearest 50F or so, but it's telling you temp to the nearest 1F or so. That precision is probably bogus. With an analog gauge, you would never know if it was 10 or so degrees off at ambient, and you would be happy.

You want it to be accurate up around 900F to 1100F. If this thermocouple was calibrated for EGT use, it probably is accurate up around the range you care about. It is probably not very accurate down around ambient, but you don't care about that.

This. ^^. The gauge is probably giving you a reading based on a calibration curve. Typically you adjust that curve to be most accurate in the critical range. +\- 50 degrees at 200 on the pyro is really of no consequence. Preferably you want it within a few degrees in the 800-1100 degree range. Type k thermocouples cover a huge range so expecting the thermocouple to follow a reasonable function over that range is just unrealistic.

Clb · Feb 20, 2020

^^^ or just wait for the tapppppp
Then know it got HOT!
It would rock to see some scientists proof the units...

laserjock · Feb 22, 2020

Clb said:
^^^ or just wait for the tapppppp
Then know it got HOT!
It would rock to see some scientists proof the units...

Ummm..... what do you want proofed?

Clb · Feb 22, 2020

^^^ just how accurate brand xyz are side by side at the critical temp. Points.
Just a fun thought seeing how the glowshift I put in the 88 lasted but a short time before wetting the bed.... and the ? Autometer? In the 93 is still ticking years later.

laserjock · Feb 22, 2020

Clb said:
^^^ just how accurate brand xyz are side by side at the critical temp. Points.
Just a fun thought seeing how the glowshift I put in the 88 lasted but a short time before wetting the bed.... and the ? Autometer? In the 93 is still ticking years later.

Gotcha. I’m not 100% sure how you would check calibration besides putting it side by side with a known calibrated sensor. For temp the easy way is to do ice and boiling water. That’s somewhat reasonable for a water temp sender but for an egt thermocouple, not so much. The thing about thermocouples (and most sensors) is that they are usually treated as having a linear response which is typically not the case. They may be linear over some portion of the range. It depends on the type of gauge too. The modern steppers with microprocessors in them can follow more difficult (less linear) sensors more accurately. If I can ever get enough time to work on it, I’ve got something cooking where I’ll post the curves for some popular sensors and how to read them out.

Thermocouple accuracy question

nelstomlinson

Full Access Member

laserjock

Almost there...

Clb

Another old truck

laserjock

Almost there...

Clb

Another old truck

laserjock

Almost there...

Similar threads

Latest posts

Forum statistics

Staff online

Members online

Share this page