Causes of noisy TDR traces?
Bruce Metelerkamp
I wonder if any of you has faced the following:
Traces obtained from three-rod TDR probes installed in both soil
(sandy loam, 8 probes) and tree stems (citrus, eight probes) in an
experimental orchard showed a highly variable noise in the form of
many saw teeth which completely distort the traces beyond measurement.
The phenomenon comes and goes as it pleases, without any obvious periodicity.
When the noise "disappears" the trace are stable and 'perfect' in shape.
I must add that there are situations where the stem traces are 'quite'
while the soil probes are clearly noisy.
(Intuitively it looks like some induction from communication
transmissions (the army is active in the area) or high voltage
induction (which is far away).
A "grounding" precaution does not seem too promising: the probes are
buried 20 to 80 cm deep.
I would appreciate the reactions of whoever has an idea or has
experienced this phenomenon.
Sincerely
arie
Arie Nadler, Dept of Soil Physics, Institute of Soil, Water, and
Environmental Sciences. Volcani center, ARO, Mins. of Agriculture,
State of Israel.
POB 6 Bet Dagan, Israel, 50250.
Phone: 972-3-9683865 Fax: 972-3-9604017; Home: 972-8-9404161
dbar...@ecs.umass.edu
Ah, two of my favorite topics: TDR and "grounding, bonding and shielding."
Without an image, it is difficult to say for sure, but the description sounds like plain old intermittent AC mains interference. It could also be a "ground loop:" current flowing on the shield connecting pieces of equipment at different potentials. This is usually caused by plugging equipment into different AC mains circuits.
Check and recheck cables, especially shields.
As Robert A. Heinlein wrote, "It's not what you don't know that hurts you, it's what you do know that ain't so."
If the equipment worked correctly before, see what has changed and fix that, rather than redesigning the system.
If you are using unbalanced cable (coax), a broken or intermittent shield is the most likely culprit. A connector failing to carry the shield effectively could also be at fault.
If you are using shielded twisted pair, make sure the shield is not coming in contact with one conductor of the pair. Shields don't need to be grounded ("earthed") at all to work against EMI, but to provide protection against hazardous voltages while avoiding the dreaded ground loop, ground them at one end only.
The beauty of twisted pair wiring is that it is inherently "self shielding."
Finally, if your data acquisition system or data logger powers your sensors and transducers as well as receiving data, stick to a single point ground at the CR10 or other equipment.
Another quote from American popular literature, advice from the "The House of God," goes something like "If you hear hoof beats, it's more likely to be a horse than a zebra." Now, I know some of you are actually IN Africa and zebras may not be uncommon, but you get the point: Look for the mundane explanation first. ;-)
All the best,
Dave B.