Issues beading a thermistor?

[Pack34]

I have a thermistor that's on a sub-assembly that I'm reading using an INA330 thermistor amplifier. What the amplifier does, is it puts a current through the thermistor and a separate sense resistor. It then compares the two currents and outputs a voltage to be used by a PID controller.

Now, I'm getting some noise in the thermistor at a frequency that's seen in a clock that goes down the same ribbon cable. My main concern are any inductive qualities since the current through the thermistor is what's used as the driving factor for the PID.

[T3sl4co1l]

You don't have an EMI filter on the thermistor input pin?

Also, you're running clocks down unshielded cables?

Tim

[Pack34]

At this point, no. In the past this did not seem to be an issue but it's suddenly come up. I'm planning on putting on a ferrite bead to target the 50-MHz noise.

On my To-Do list is to move from the ribbon cable to a flat flex to help mitigate noise.

[T3sl4co1l]

Geez man, at least add an RC, to filter outside crap, as standard practice!

Tim

[calexanian]

Geez man, at least add an RC, to filter outside crap, as standard practice!

Tim

+1 for adding a little RC filter. at  that high of a frequency and what i am assuming to be a standard 10K or so thermistor a couple pf across the line would probably do the job.

[T3sl4co1l]

+1 for adding a little RC filter. at  that high of a frequency and what i am assuming to be a standard 10K or so thermistor a couple pf across the line would probably do the job.

More than that -- remember you're working against the impedance of the cable, not the thermistor.  (The thermistor is all but irrelevant at 50MHz!)  The rule of using only what bandwidth you need, would suggest that 0.1uF would still be "small" for the application -- but more than good enough for RF purposes!

Tim

[Conrad Hoffman]

How fast is the thermistor? I'd think that even with a small one you could use pretty heavy filtration before affecting the response speed.