Error due to lag in cold junction compensation

Post any questions you may have about our current range of USB data loggers
Post Reply
John Mitchell
Newbie
Posts: 1
Joined: Wed Jan 07, 2009 12:26 pm

Error due to lag in cold junction compensation

Post by John Mitchell »

I have been using a TC-08 USB temperature logger to log temperatures on a large steel structure. I noticed a peculiar effect: when the air temperature changed rapidly, the readings of some of the sensors on the structure recorded a blip in the opposite direction. I think I have tracked the problem to the fact that the cold junction compensation sensor in the TC-08 is poorly coupled to the cold junctions themselves. As a result, when the temperature of the environment changes, the cold junctions and sensor catch up at different rates which gives rise to an error in recorded temperatures. I'm thinking of maybe installing a miniature fan inside the TC-08 to help matters. Perhaps Pico could think about addressing this problem in future. Otherwise an excellent piece of kit.

ziko
Advanced User
Advanced User
Posts: 1705
Joined: Fri Dec 01, 2006 10:03 am
Location: St Neots

Post by ziko »

Hi and thank you for your post.

The TC08 works best with slow moving ambient changes which allows the cold junction sensor to more accurately track the connector temperature changes.

Kind regards
Ziko

Technical Specialist

olepagh
User
User
Posts: 7
Joined: Sat Jan 01, 2005 9:39 am
Location: Copenhagen, Denmark

Re: Error due to lag in cold junction compensation

Post by olepagh »

John Mitchell wrote:Perhaps Pico could think about addressing this problem in future.
I think they should do so.

We, as part of an accredited lab, uses the TC08-USB with all sensors (type-T) in some applications. The TC´s are of course calibrated and via calculated fit and some look-up tables and other neat software, we can measure accurate to less than 0.1 °C, but this has only been achieved by placing the converter in an isolating foam enclosure.
Without insulating, the uncertainty is more than 0,15 °C alone from drift of the CJC. Especially the inputs 1 to 4 differs from 5 to 8 (the two rows) if the converter is near some heat source (a human for example).

The input connectors should ideally all have been embedded in a copper rail and the internal temperature sensor bored into that rail.
Alternative a CJC sensor for each input, placed close to the terminal instead of one common thermistor placed in the center of the input terminal area.

On the other hand, I agree that the unit is value for money.
/ Best regards
Ole
Denmark

ziko
Advanced User
Advanced User
Posts: 1705
Joined: Fri Dec 01, 2006 10:03 am
Location: St Neots

Post by ziko »

Hi, The TC-08 offers good value for money. It is accurate and has a good resolution. Under some operating conditions the product does need to be isolated/protected from rapid ambient temperature changes to get the best performance from the product.

Kind regards
Ziko

Technical Specialist

Post Reply