DC output of a function generator and voltage divider

[mssong]

DC output of a function generator and voltage divider

I set the function generator to 600mV (I set it that way because I didn't match the impedance. I want the output to be 1.2V) and tried to divide the voltage in half with a voltage divider (500k, 500k), but the output of the voltage divider is close to zero.

I measured it with a multimeter and the input resistance is 40Mohm, so it shouldn't change much even if the voltage value changes.

Is it not possible to use a voltage divider in the hundreds of kohm range like above (500k+500k) with the function generator?

 

[danadakk]

Yes. Generally the f() gen is 50 ohm, or at least a selection of low Z.

Your multimeter working on other measurements ? Also you measuring
f() when its outputting sine or a complex waveform ? Asking because
multimeters have limitations, read specs. Investigate your multimeter
crest factor..... 500K and 20 pf stray have a 3 db point of ~ 16 kHz,
sine wave. Then there is crest factor issues for more complex waveforms.

500K seems high, "normally" one would think in low K area. 500 K with
stray C and C loading of meter can affect readings. Again read your
multimeter manual.

What is frequency and kind of your waveform being measured ?

Note your 40M multimeter loading is probably a spec for DC measurements,
and falls for increasing frequency.

Regards, Dana.

 

[KlausST]

Hi,

I´v never used a function generator to generate a DC signal.

there are a lot of informations missing.
What function generator? and what setting?
* what voltmeter did you use? and what setting?
* what frequency?
* what waveform?
* what is the voltage value: (in case it is not pure DC) Vpp, Vp, V_RMS, or anything else?

I see that you wrote "function generator" and "DC output" ... But is this really the case? Did you use all setup and connections to indeed get DC output at all? Maybe there is a series capacitor. Or something else. Maybe a bad connection ...

--> I´d say: there´s something worng. I don´t know what exactly ... thus I asked for complete informations.

The resistance values don´t explain what you describe.

A simple debug method: measure the output voltage without the voltage divider.
And after that: connect the voltage divider ... and measure the FG_Output voltage.

Klaus

 

[dick_freebird]

Are you sure the generator is actually (a) good? I have one
kind of pulse generator which has the features you describe,
but if anyone powered it up and turned on the output
(that would be (b)... ) with the termination not there, the final
amp will blow out and the next guy gets to debug a brick.

Put the 50 ohms it expects, power it up, turn it on, what's
there without all the divider and driven device?

1.2V sounds LVDS-y (or DDRx, same diff) and what you
may want is a common-mode divider (res from the input
that you're feeding to both VDD and VSS) and another
incoming from source to that junction. A 3-body problem
that would easily be solved by some nested SPICE loops
and a finger. Look at LVDS general-purpose RX design guts
(you can find them in older datasheets) for this style to get
"the right" common mode, extend common mode. RS-422
RX are similar but get to a +/-7V common mode (+/-10
tolerant) in 5V technology.

 

[mssong]

Yes. Generally the f() gen is 50 ohm, or at least a selection of low Z.

Your multimeter working on other measurements ? Also you measuring
f() when its outputting sine or a complex waveform ? Asking because
multimeters have limitations, read specs. Investigate your multimeter
crest factor..... 500K and 20 pf stray have a 3 db point of ~ 16 kHz,
sine wave. Then there is crest factor issues for more complex waveforms.

500K seems high, "normally" one would think in low K area. 500 K with
stray C and C loading of meter can affect readings. Again read your
multimeter manual.

What is frequency and kind of your waveform being measured ?

Note your 40M multimeter loading is probably a spec for DC measurements,
and falls for increasing frequency.

Regards, Dana.

I set the DC output of the generator to 600mV because I didn't do a 50 ohm match (intent was 1.2V DC).

There is no AC component, so the multimeter is only measuring DC.

When I measured the output of the voltage divider with a multimeter's DC measurement (input resistance 40Mohm), I got a voltage close to zero.

And if you say “500K seems high, ‘normally’ one would think in low K area. 500 K with
stray C and C loading of meter can affect readings. Again read your
multimeter manual.” What does this mean?

 

[FvM]

You didn't yet report if you see 1.2 V at generator output or not.

 

[mssong]

You didn't yet report if you see 1.2 V at generator output or not.

When I measured the generator output directly with a multimeter, I found that I was getting a good 1.2V.

 

[KlausST]

... then either one of your resistors is wrong (measure their Ohm value),
or your wiring/soldering is wrong

Klaus

 

[danadakk]

Confused here.

Why are you generating DC with what normally thinks of as a AC generator, function generator ?

Can you show a pic of the settings you have on the actual f() gen and a scope shot ?

f() generators do have DC offset controls, most of them So I guess if your generator can turn
off ac generation you would just see DC on the output. from the offset control....?

And if you say “500K seems high, ‘normally’ one would think in low K area. 500 K with
stray C and C loading of meter can affect readings. Again read your
multimeter manual.” What does this mean?

Those comments are largely targeted towards AC issues. High valued R's and stray C make low pass filters.
So if one is doing dividers for AC signals thats "normally" done with values that do not alter signal content/
harmonics, unless thats what you desire.

Again read your
multimeter manual.” What does this mean?

Cant emphasis that enough, manual will tell you of the limitations your multimeter has....

 

[dick_freebird]

When you mention that high-Z voltage divider,
have you incorporated the input impedance
(resistance) of the driven load, to that network
out-facing impedance for the true node Z?

Like maybe the driven-thing has a 50 ohm
input and your 250K divider is out of gas. There
must be a "driven thing", the point of this
exercise can't be to just drive a resistor divider,
right?