PDA

View Full Version : 28X ReadADC10 Problem - AGAIN!



HamRadioAddict
18-01-2006, 14:03
Hello All,

Really enjoyed reading the thread “28X ReadADC10 Problem – AGAIN! I hate to kick a dead dog but there was one thing that instantly caught my attention.......somewhere around reply 27(ish). This was the comment about the temperature sensor instrument cable shielding earthed or grounded at BOTH ends.

In my line of instrumentation work, where we work with the usual industry standard small signal sources such as 4–20 mA, 0-5VDC & 0-10VDC, earthing the screen or shield of the instrument cable is only done at ONE end. And that is at the source or supply end. This would be at the PICAXE or OpAmp end not out at the temperature sensor end. Earthing at both ends can cause unwanted current loops. These are unwanted currents flowing along the instrument cables shield. These currents could be anything from pulsed DC or 50 Hz etc. etc. (60Hz elsewhere;-) what ever decides to hitch a ride and use the shield as return path. When and if these currents are strong enough, they will cause noise on the very signal wire and circuit you are trying to shield.

I’m not sure if shielding at one end only was practical in the temperature sensor, OpAmp
PICAXE arrangement but thought I’d throw my 2 cents in just for the heck of it. Sorry I didn’t respond earlier but I didn’t read the post until the problem was solved.

73’s
Steve VK6HV

BeanieBots
18-01-2006, 16:50
The LONG co-ax sort of slipped in.
I agree, (and did comment) about creating the dreaded earth loop.
One of the drawbacks with using PICAXE for accurate analog work is the inabilty to seperate analog supplies from digital which is another source of noise.

nbw
19-01-2006, 21:24
I prob. need to clarify this a bit... the ground at the sensor end was the shielding braid in the cable, which was grounded at the main circuit board end. The small length of cable from the sensor-cable's socket to the op-amp was shielded also, grounded at both ends (only about 15cm long).