Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
Electro magnetic interference, or EMI, is commonly found in industrial environments, and can adversely affect the accuracy of your instrumentation signals. so anyone can help me how to minimize the emi noise?
You also need to consider twisted pair signal cabling, and appropriate grounding of their shielding. Low-impedance Line drivers and receivers may help keep your signal/noise ratio usable. Filtering frequencies and 'spikes' outside your signal range will help, as will nimble software to catch anomalies.
But, there's a 'gotcha': IIRC, raw data from the NASA satellites measuring ozone over the poles was discarded by software as 'too low, so faulty'. Took a couple of guys from BAS with Victorian-era instrumentation to flag the growing seasonal hole in ozone layer. This eventually led to wide-spread ban on CFCs...
So, due care, please ?
Your 'wonky' data may be the truth...
There are lots of way that are commonly used in EMI noise reduction. One of the best technique is to keep the noise level within the EMI that is EMI noise suppression with passive components. EMI noise suppression techniques with passive components include passive filter. Passive components in motor drive systems are bulky and suffer from parasitic effects, resulting in the unsatisfactory performance of passive component suppression techniques. Active filters that are applied to VFD systems for noise reduction sense the noise current or voltage, and then inject a current or voltage with opposite polarity but the same magnitude into the circuit to cancel the EMI noise. This technique reduce CM voltage noise sources by modulating the switching states in a control scheme.
My recent experience: an Arduino connected to (and powered by) PC via USB, and only the (-) lead of a passive power supply connected to Arduino's ground, no other connection. Every time the PSU was plugged or unplugged on the wall, communication was lost. Removing the (-) lead there was no problem.
After going crazy with coiling the USB cable around a ferrite toroid, inserting uH and mH inductors in series with the (-), etc. the only cure was to add copper braid around the USB cable (had to cut one end, transplant the shield from a coax, resolder the connector), soldered to BOTH connectors shells.
I agree with the others.
What I'm missing is, that some EMI can be educed with proper PCB layout.
Often one takes care only on the signal paths, but this is only half of the truth, you need to include the signal_return path.
Often this is a GND plane.
For radiated EMI:
* The length of the signal path determines the lower frequency limit. The shorter the better.
* the enclosed area of signal-and-return-path determines the magnitude of radiation. The smaller, the better.
Thus any cut in a GND plane that crosses the return path will increase EMI, because the area becomes wider.
Also the placement of filter (capacitors) will influence how effective the filter works. Even for conductive EMI.