VFDs and DS18B20 Temp Sensors

[Phuz]

Hey folks,

I have been working on a system that is using about 20 one-wire temperature sensors through a Teensy 3.1. The Teesny and temp sensors are powered (5VDC) by a buck converter from the 24VDC supply on the PLC. Things work great with reading temperatures and whatnot until we turn on the VFDs for the water pumps on the system. When we turn on the VFDs, the sensors give garbage readings. The VFDs have their own power supply, but they seem to be the cause of the bad sensor readings. Any of you folks have experience with this? Is it the switching harmonics that are messing with sensors? Is there a way to mitigate that? We can't more forward with our proof-of-concept prototype testing until we figure this out so any help will be greatly appreciated.

Cheers,
~Phuz

 

[Wozzy]

My experience is that high frequency VFDs must use shielded cables to the motor if there is any sensitive instrumentation near by. My best results have been with the shield tied to ground only at one end. Even a slight gap in the shielding can cause EMI leakage. If possible also shield the instrumenation cables. Also check the mains power if it is common with the instrumentation as the VFDs can cause high frequency fluxuations.

 

[joe_prince]

My experience is that high frequency VFDs must use shielded cables to the motor if there is any sensitive instrumentation near by. My best results have been with the shield tied to ground only at one end. Even a slight gap in the shielding can cause EMI leakage. If possible also shield the instrumenation cables. Also check the mains power if it is common with the instrumentation as the VFDs can cause high frequency fluxuations.

I second this. VFDs give off a ton of EMI so shielded cable is a must.

 

[PaulStoffregen]

The VFDs I worked with many years ago (in the late 1990s) also caused massive high frequency noise to be conducted onto the AC lines supplying their power. Not nice!

 

[Constantin]

The VFDs I worked with many years ago (in the late 1990s) also caused massive high frequency noise to be conducted onto the AC lines supplying their power. Not nice!

Not sure about back then but these days there are regulations against that. Most VFDs I take apart feature big dual mode chokes, Y1 caps, etc. to help keep the noise out of the mains. Ditto devices with large inductive loads like microwave ovens. Was quite surprised to see a measurable impact on standby power consumption due to these filtration boards being used on some (but not all) MWOs.

The advances in that field have been incredible. There are clothes washers with IGBTs mounted directly to motherboards without heat sinks on them that are driving washer drums to 300+g. Dishwashers also feature them for the main pump (though much smaller). The biggest VFDs I've encountered were in compressor drive systems. Some of the OEM designed solutions are more complex than the motherboards used in PCs.

I still remember encountering my first one in a fisher Paykel clothes washer, the thing had its heat sinks soldered to the cold water line(!!!). For an emergency stop, they would have to turn on the cold water to keep the drivers for the three channels (still separate components then) from popping like popcorn.