ADC fluctuations due to voltage instability

[patrick0427]

From page 229 in "Process Control":

"You may note the temperature fluctuating slightly as the heater cycles as shown in Figure
7-5. The heater draws significant current, dropping the supply voltage. This in turns
affects the PWM voltages controlling the ADC. As the heater cycles on and off, the data
from the ADC representing temperature varies slightly. This is another form of noise
caused by cycling of high-current loads."

What is the best practice method to get around this? The options I see are:
1. Use 12 bit ADC and leave offset & span at 0V and 5V respectively, ditching PWM.
2. Use voltage regulators to set the span and offset.
3. Use pot to "hard" set span and offset.

I do not really like options 2 and 3 because I would like to be able to adjust the span and offset dynamically so that the program can keep these values bracketed around the last known temp measurement.

That leaves me with option 1, which would work. But I was hoping for a more elegant solution.

Thanks for any feedback.

Patrick

[PJ Allen]

4. Run the 'Heater' from a separate supply (instead of from "V_in".)

[patrick0427]

4. Run the 'Heater' from a separate supply (instead of from "V_in".)

Ok that may be the best solution. Let me give you a little more detail. The drop on V_in is actually from powering a 12V relay (in my project) that switches a 115 VAC device.

For this project I want to have one cord that plugs into a standard wall outlet which would power the board of education and the AC devices. For Option 4 I would need to pull off 12VDC from the line voltage of 115 VAC before it goes into the BOE I guess. What is the best way to do this as well as supply the DC voltage required to run the BOE without adding an ugly powerstrip?

[PJ Allen]

Ahh, yes... the truth outs, more details.

Place the "ugly powerstrip" behind a veil?
Get an attractive bench supply with outputs sufficient to your needs?
I don't know what you're using at present, but you might source a more capable supply for the BoE, etc., for "V_in" only with more guts.
Maybe you could find a relay that requires less current. Maybe the relay (its coil, it's an electromagnet, remember) is too close to sensitive stuff.

A separate supply should isolate the source of the "tick", if it's related to peak in-rush current demand or other load demand, and then you could proceed with implementing solutions - otherwise we're jumping to conclusions.

Can't say, "do this and do this and do this" till you nail down the "trouble."