DMM Frequency response

If a DMM has an input impedance of 10 M Ohms and 20 pF, the 3 dB frequency is 1/(2piRC) = 800 Hz. The capacitance in this case is a shunt effect to ground after the 10 Megs of resistance ... so it forms a low-pass filter.

A 1 volt signal at 800 Hz will display as 0.7 volts.

Scopes are better for higher frequencies.

So, the equivalent circuit is something like this?

-----\\/\\/\\/\\/----- Vdmm Vin R | C = | -------------------

If we put DC voltage across the inputs why not the DC voltage shows up (in AC measurement mode)?

How scopes and other more sophisticated devices do? How to represent the these as equivalent circuit?

Thanks very much!

Because it's made to ignore DC potential...

My RadioShank DMM (before it died...sigh) started dropping off at 3kHz, based on my observations.

Tim

-- Deep Fryer: a very philosophical monk. Website:

Yes.

There is a series blocking capacitor.

It's selectable on a scope ... dc coupling (no blocking capacitor) ... or ac coupling (capacitor). dc coupling is also called direct coupling.

couldn't they bypass the 10M with a suitable cap to reduce that ?

wether the DMM can sample the input fast enoug could be another problem..

Bye. Jasen

You are thinking about the distribution of a dc voltage across two capacitors in series ... does not apply here. The capacitor is selected to have a low reactance at the lowest frequency of interest. At dc, 0 Hz, its reactance is infinite; hence the dc blocking function.

"Charles Schuler" wrote in news: snipped-for-privacy@comcast.com:

If we add a series cap (denoted it as C1) at Vin, there should still be DC voltage across C. Maybe C1>>C so that most voltage is on C1? By the way what is the typical capacitance needed for DC blocking? What are the main considerations?

Even we can select dc/ac coupling, it only deals with the DC component, how can it avoid low-pass phenomenen? It should not be simple otherwise all cheap DMMs can do this, I suppose?

Thanks.

Jasen Betts wrote in news: snipped-for-privacy@clunker.homenet:

If they bypass 10M, I think the impedance may be too low.