I'm working on a temperature controller project that involves a thermistor, a non-inductive resistive type heating element, and a microcontroller. The thermistor and heating element are place about 10 feet away from the electronics (worst case scenario, best case maybe 2 feet). One of my concerns that is beyond my depth is EMI. The heater will be controlled using PWM and frequencies of around 100hz to 1khz. The temperature sensors are sampled around 20hz.
The circuit will be placed inside a metal project box, regardless if it's necessary or not from a shielding point of view. Out of curiosity, is it necessary?
Also, for the cables, I'm assuming I need to use twisted wire for the thermistor cables, but should I also have shielding? We want pretty precise measurements of the thermistors, and they're simply being measure from a voltage divider + 24 bit ADC. We only really need 16-18 bits of resolution. In my "bread board" circuit I was able to get more than 16 and less than 17 bits of resolution when I took the standard deviation of many measurements at a fixed temperature (and better when I used a precision resistor instead of a thermistor).
For the heater wires, they're running at PWM of 100hz-1khz at 24 volts. Since this equipment will be used in a laboratory setting, we think it may be necessary to shield the cables to prevent interference to other equipment. Not sure if this is necessary or not since these are relatively low frequencies. Even if it's not necessary, what's the best practice? I guess shielding wouldn't hurt us, and it's not out of our budget, but I'd like to be able to justify design decisions.
I appreciate your advice/help.
Thanks,
Matt
The circuit will be placed inside a metal project box, regardless if it's necessary or not from a shielding point of view. Out of curiosity, is it necessary?
Also, for the cables, I'm assuming I need to use twisted wire for the thermistor cables, but should I also have shielding? We want pretty precise measurements of the thermistors, and they're simply being measure from a voltage divider + 24 bit ADC. We only really need 16-18 bits of resolution. In my "bread board" circuit I was able to get more than 16 and less than 17 bits of resolution when I took the standard deviation of many measurements at a fixed temperature (and better when I used a precision resistor instead of a thermistor).
For the heater wires, they're running at PWM of 100hz-1khz at 24 volts. Since this equipment will be used in a laboratory setting, we think it may be necessary to shield the cables to prevent interference to other equipment. Not sure if this is necessary or not since these are relatively low frequencies. Even if it's not necessary, what's the best practice? I guess shielding wouldn't hurt us, and it's not out of our budget, but I'd like to be able to justify design decisions.
I appreciate your advice/help.
Thanks,
Matt
