In my lab we are beginning to do neural recordings which are very sensitive to noise... given the fact that signals are just few microvolts. In addition we are consistently observing nice spikes in the 50Hz signal picked up by the amplifier! They are very periodic in fact (once every 10ms) and each such spike lasts about 1ms and they are phase locked to the 50Hz sine! We realised that it is there in the line voltage but not rectified by the UPS somehow... Is there any solution?? Any front-end equipment (economically priced and easily available) that will rectify the power supply for the entire lab?

 

If the noise is line conducted from your power supply you could look for a
line filters and isolation transformer.

Also you could look at powering your sensor amplifiers using
batteries which would remove all conducted line noise.

(* jcl *)

 

Rite.....This is interesting...I had the same problem a few years ago (but only during the day)....It drove me mad....Even after fitting a dirty geat filter to the input of my 60 Amp supply to the workshop, it was still there. In fact the filter was almost jumping about the place....I contacted the electricity board, who said that it was nothing....So I set about traking down the source of the interfearence....I built a small radio reciever tuned to the frequency of the interfearence, and with the aid of this through a small pair of personal headphones, went for a good long walk in the direction of the underground cables supplying my road......I could pick it up rite to the substation, so I then walked back to the last area fed by the cable.....As I got close to a road junction about 500m away from my house, the signal got louder, until I was standing under a street lamp. There it nearly took my ears off......I looked up and there was the 250 watt sodium lamp flashing gently in time with the signal in my ears.....I called the lighting company, and after some investigations they found that the whole thing was wired up wrong.....They had connected the LDR switch unit in series with the lamp only, and not the whole fitting, so the thing was on all the time but the ignitor was trying to fire the lamp at about 2Kv, this was jumping accross the relay contacts of the switch. (of course that's why it was O.K at night)This was then flying back down the line and causing the problem......Daniel.

 

It also can be due to electromagnetic interference, which normally is created by transformers and ballasts, and may persist even with all the supply filtering cautions. You can easly puck up microvolts with a single wire next to a transformer.

 

Daniel ... nice job of locating the problem.

 

Unless you can identify the source, your efforts may prove fruitless. What is the magnitude of the spikes BTW?