Analog Oscilloscope advantages?

[GK]

Electronic Analog and hybrid computers, Korn and Korn.


Test results for: Hamlin SPST reed relay, part # HE721A1210. CH1 is the coil voltage, switched to ground by a transistor. CH2 is the SPST contacts switching the +12V to ground via a 1M resistor. This relay would be more than adequate for 50Hz chopper duty. Sounds like a quiet cricket chirping. Coil di/dt would need to be limited though at turn-on to prevent electromagnetic coupling into the high impedance switch circuit, however (note the spike in the "switch-on" scope photo).

EDIT: actually, that "electromagnetically" (or in reality, more likely capacitively) coupled spike might make the relay a dud for this application. A ~2V peak, that's around 2uA in the 1M. The 2V spike would probably integrate out over the 20mS period to a few 10's of millivolts - quite useless when the desired signal is uV! The spike as measured would need to be reduced by ~3 orders of magnitude to make this relay a decent high impedance chopper.
 

[dfmischler]

I didn't see anybody say that the linear power supply in their older analog scope (mine is a Tek 475A) let's them float the scope just by lifting the ground lead with a 2 to 3-prong power adapter (of course, for safety you should never defeat the safety ground in any equipment  ).

[David Hess]

That should work with a modern DSO.  My old DSOs certainly can handle it.  The usual issues with exceeding the breakdown limits between the primary and secondary side of the power supply still exist of course.  I prefer to use an external isolation transformer.

[SeanB]

For lowest bounce you do need to carefully measure the particular reed you are using and apply just enough voltage to close it. That way the contacts close with the least force and the least bounce. You might have to use a constant current drive though as the copper will warm up and cause issues, though using a CC source means switching is a lot easier by just using another transistor to short the coil, no diode needed and less switching noise. Preferably use a shielded relay  and arrange the connections to the coil to connect ground to the inner winding, nearest to the reed, so there is less noise coupled from switching.

[GK]

Some further investigation later and I don't think that any modern commercial reed relay would be suitable for the task. At a minimum, you'd need to drive the coil from low frequency (mains-derived) AC, just like they did back in the day. Then there is still the issue of coil to contact capacitance, which is still way toohigh in any of the reed relay datasheets that I have looked at - even with electrostatic shielding options. The electrostatic shields only appear to have a modest effect, typically reducing the specified coil-contact capacitance by only a half or to one third.