| ??? 08/31/05 09:21 Modified: 08/31/05 09:55 Read: times |
#100324 - Exactly! Responding to: ???'s previous message |
Oliver said:
The pulsed approach of 10msec pulses means you need good high frequency response, which means you are more prone to interference. Exactly!! Think only about the huge harmonics sitting in the edges! You would put high frequency noise into your circuit where no HF would ever need to put its foot in. I would also try filtering the inputs in hardware, as high frequency edges can alias down and cause all sorts of havoc. HF noise often results in strange DC offsets, which are produced by demodulation at non linear acting pn-junctions. This can only be prevented by using a low pass filter at input. As the PT100 does not show any non linear demodulation effects, a simple low pass filtering at the point where the PT100 lines enter the board MUST therefore help to avoid all annoying demodulation effects. No HF noise must reach the first amplifying or mulitplexing stage. Then, no strange DC offsets will be observed! PT100 are deadly simple in removing any HF interference and related effects, because signal bandwidth is nearly DC. I never had any trouble with PT100 lines. Kai |
