reducing noise in the Rigol by screening the power supply

[Simon]

ok  now some poeple may remember how I winged about the noise on my Rigol scope when i first bought it and did not understand that digital scopes have some inbuilt noise that can't be helped.

I noted that the scopes PSU is totally unscreened and found it and still do find it to be very odd. Now true enough that it seems to have little bearing on the screen readout but I've been playing around.

now initially I had found the noise to be around 125 KHz by using the scopes own digital filter to "tune" it out as best as possible, I installed an aluminium shield around the PSU and beleived it had helped a bit although as I learned aluminium was not the greatest idea and I should have used steel (or better mu-metal), today after running the scope for a bit I took the back off and found the psu to be a bit hot and so decided to remove the screen to help ventilation.

Bingo, the noise did not only increase slightly but I got a negative offset of the trace with no probes connected, ok So I recalibrated the scope and now the trace is back on "0" but well just it's not bang on I'd say it is 1 pixel out on the screen.

now while I had the back off I did some experimenting, I put the probe near the PSU transformer and low and behold i picked up a 125 KHz signal. So i have to say i was right I had initially guessed the "noise" to be around 125 KHz and thats frequency the SMPS works on ! I know it makes very little difference which is why Rigol obviously never bothered but it makes a slight difference.

Essentially yes I know i will never get rid of the digitally generated noise, I looked into the trace with no probes connected: with the screen round the SMPS the noise was 2 div high, with the screen removed from the PSU there is the occasional (because I'm looking at the scopes digital noise is at 100 MHz) spike at another 2 divisions.

I'm going to try another screen (made of steel), this time not right around the SMPS (to allow ventilation) but a single sheet parallel with the main board and connected to the outer case for earthing,

I think rigol should have put more of the circuit in a screened box, I think from the input to the ADC's should have been screened so that the only unscreened signals are digital, but then I suppose it was down to cost and the slight difference it would of made.

[Mark_O]

Good luck, Don Quixote!

- Mark

[Simon]

I'll be getting someone at work hopefully to punch out the shield so may take a while, then I'll let you know the results, I know it won't make a huge difference though

[Ferenczyg]

You always can use modder's mesh, just go to any modder's shop..

[Simon]

yes that would have been a good idea but this time I thought just putting a solid earth sheet under the SMPS running parallel with the main board should do the trick