[Test-Equipment] Re: tube input impedance (increasing thereof)

[Barry L. Ornitz]

Randy Zelick wrote:

> I am contemplating modifying a Tek 2A61 high-gain 
> differential pre-amp for a physiology experiment.
> 
> This antique plug-in (for 560-series scopes) has an input 
> impedance of 10 Megohm, and a maximum sensitivity of 10 
> microvolts per division.
> 
> The input circuit is a pair of 5842 triodes (one for + and 
> one for - input lines). Each has a 10 Megohm resistor from 
> the input grid to ground which sets the overall input 
> impedance.
> 
> What I want to do is increase the input impedance to the 
> maximum I can get away with. A gigaohm is not too much, if 
> this can be realized.
> 
> So....
> 
> Does anyone have ideas about this? What is the Zin for such 
> a tube? Can I get away with no grid-to-ground resistor or 
> will the grid float positive?

As others have noted, electrons that hit the grid must have a 
low enough impedance for their current to flow or the tube 
will quickly reach cutoff from the voltage drop across the 
resistor.

There are some tricks with bootstrap circuits where the AC 
input impedance is raised considerably but the DC input 
impedance is increased only slightly.  Essentially you 
capacitively couple the "ground" end of the resistor to a 
replica of the input signal.  But for a physiology experiment, 
I suspect you will need a frequency response down to DC.  Even 
if you do not, the capacitor will have to be an exotic PTFE 
unit or its leakage will cause problems.

Your best bet is to use a solid-state instrumentation 
amplifier.  You can build these from conventional operational 
amplifiers or buy them prefabricated.  Your big concern here 
is what is the common mode voltage they have to withstand.  
Getting a terra-ohm DC input impedance with decent frequency 
response is not that difficult with modern IC parts.

However with an input impedance this high, it takes great care 
to eliminate surface leakage effects.  In a previous job, I 
designed some equipment that made resistance measurements 
routinely up to 10E15 ohms.  Glass is no longer a suitable 
insulator here, and is marginal beginning at the 10E10 ohm 
level.

        Dr. Barry L. Ornitz     WA4VZQ     ornitz@tricon.net