50KG Button Load Cell Thermal Drift Over Spec

[alexanderkjones]

[font=arial, sans-serif][highlight=#ffffff]I recently purchased 2 CLZ204E 50KG button load cells and I am currently testing their thermal performance to compare with the certificate. I am not able to replicate the thermal zero shift of 0.185 %FS/10C or 2g/C as noted on the load cell’s certificate and product page from the supplier phidgets.com[/highlight][/font][font=arial, sans-serif][/font][font=arial, sans-serif][highlight=#ffffff]I have been temperature cycling two load cells in a bench-top thermal chamber between 50F and 110F with a low gain of 1000 @ 3.3v using an INA126 opamp that has been tested and operational with a similar load cell the FX1901 from Measurement Specialties. The results can be seen on Ch1 & Ch2 of the following chart. Ch3 is a an unloaded FX1901 button load cell.[/highlight][/font][font=arial, sans-serif][/font][font=arial, sans-serif][highlight=#ffffff]Chart Results:[/highlight][/font][font=arial, sans-serif][highlight=#ffffff]docs.google.com/spreadsheets/d/1h5Se9_M9h0u46m5aTOgcr54mVyXuTHTYU3jFOzePu4U/pubchart?oid=1320126634&format=interactive[/highlight][/font][font=arial, sans-serif][highlight=#ffffff]

The
CLZ
is calibrated to approx. 3mv/lb and I am seeing temperature drift with a static load of 20mv across the temperature range roughly 1%FS/10C not0.185%FS/10C.

I understand that this application is out of spec for the load cell, specifically operating at 3.3v excitation vs the 5v as specified on the website. I also understand I am operating at a very low gain for this sensor.

Please Advise on the best way to alter the design:

1. Is the 3.3v excitation not powering internal circuitry for temp compensation?

2. Are these effects due to the low gain? I would anticipate this having the inverse effect.

3. Other thoughts to remedy this situation would be very helpful
   [/highlight][/font]

[scharette]

Hi,

Welcome to the RobotShop forum!

Here is the response from the supplier of the load cells:

Brian is going to collect some data from our chamber - but, I believe your problem is the INA126.
±0.5uV offset typical / degree celsius drift means 20uV over 40 degrees C, or 20mV after you amplify 1000 times. 2% error.

Use a better op amp - for example,analog.com/media/en/technical-documentation/data-sheets/AD8293G80_160.pdf

This one would require a follow-on amplifier to get you to 1000 gain.
A quick test would be to run just the load cells in the chamber, keeping the temperature of the amplifiers constant.

Attached is the data I have gone ahead and collected. The actual drift is fairly low, on the order of 0.03% or so over a range of about 25°C or so. This was at a gain of 128 on our 1046 bridge.

See the files attached at the bottom of this post.

Hope this helps.
CLZ40 Temp Testing.xlsx (404 KB)
button test.xlsx (87 KB)