HP 33120A Function Generator Amplitude Offset Error

[Shay]

Hello,
My HP 33120A has an issue. The output amplitude is limited to 2.8V AC RMS, on the sine wave option, and the output amplitude is always at a known offset:

Setting the output voltage to 100mV RMS, the output is 80mV RMS

SET 200mV RMS = OUT 160mV RMS

SET 300mV RMS = OUT 240mV RMS

and so on, up to 2.8V RMS maximum. 20mV*range offset.

The AMP_OUT pin at the highest set voltage corresponds to the output voltage (2.8V AC at both), which means it's something before the attenuation.

I'd like to get any insight into fixing this function generator.

Thanks!

[Swake]

How did you measure that?
Input impedance / output impedance matching?
Frequency ?
No other options active such as FM/AM modulation?
What does the waveform looks like on a scope?
Selftest ok?

Specs says: Amplitude (into 50 ohms): 50 mVpp - 10 Vpp
I don't know if that is valid for your specific selection but limited to 2.8V does not seem right indeed.

[Swake]

BTW, how did it happen? All of a sudden or was the generator confronted to some 'signal' on the input?

[Swake]

Might help:

Possible Conflict with Output Amplitude: The output amplitude is
automatically adjusted if you select a function whose maximum
amplitude is less than that of the currently active function. This
conflict may arise when the output units are Vrms or dBm due to the
differences in crest factor for the output functions. For example, if you
output a 5 Vrms square wave (into 50 ohms) and then change the
function to sine wave, the function generator will adjust the output
amplitude to 3.535 Vrms (the upper limit for sine waves in Vrms).
From the front panel, “AMPL LIMIT” is displayed and the amplitude is
adjusted. From the remote interface, a -221, “Settings conflict” error is
generated and the amplitude is adjusted.

[Shay]

How did you measure that?
Input impedance / output impedance matching?
Frequency ?
No other options active such as FM/AM modulation?
What does the waveform looks like on a scope?
Selftest ok?

Specs says: Amplitude (into 50 ohms): 50 mVpp - 10 Vpp
I don't know if that is valid for your specific selection but limited to 2.8V does not seem right indeed.

1. Measured using a scope and also a DMM
2. Tried 50 ohm termination and also HIGH-Z.
3. Frequency does not change much the ampltiude.
4. No other options active
5. Waveform is very clear and crisp, no jitter or anything unusual. The function generator otherwise works great.
6. Self test is OK.

I am not sure how did it happen, the gen used to work great with the correct amplitude. All of a sudden it started doing this issue.

[Swake]

Tried 50 ohm termination and also HIGH-Z.

That is not normal. In high impedance mode you should measure twice the set voltage.

Frequency does not change much the amplitude.

I've never used this generator, but that voltage should remain stable.


The output limit of 2.8Vrms sounds like a setting somewhere. Try resetting everything to factory defaults.

[Swake]

Let me write that in other words: The gen has an output Z of 50 Ohm, your scope must be set to an input impedance of 50 Ohm too else the voltages on the scope and the set voltage in the gen will not match.

Your DMM most likely has an input Z of about 10 MOhm. Therefor it will never show the same voltage as set on the gen unless you have explicitly configured the gen to do so. With standard settings it will show about twice the set voltage.

[Shay]

Let me write that in other words: The gen has an output Z of 50 Ohm, your scope must be set to an input impedance of 50 Ohm too else the voltages on the scope and the set voltage in the gen will not match.

Your DMM most likely has an input Z of about 10 MOhm. Therefor it will never show the same voltage as set on the gen unless you have explicitly configured the gen to do so. With standard settings it will show about twice the set voltage.

You are indeed correct, I wasn't clear enough - the 50 settings does increase the voltage by 2. High Z automtically cuts it off in half. This function works alright.
I noticed the ampltiude error is on all settings except the square wave function.

[macboy]

Let me write that in other words: The gen has an output Z of 50 Ohm, your scope must be set to an input impedance of 50 Ohm too else the voltages on the scope and the set voltage in the gen will not match.

Your DMM most likely has an input Z of about 10 MOhm. Therefor it will never show the same voltage as set on the gen unless you have explicitly configured the gen to do so. With standard settings it will show about twice the set voltage.

You are indeed correct, I wasn't clear enough - the 50 settings does increase the voltage by 2. High Z automtically cuts it off in half. This function works alright.
I noticed the ampltiude error is on all settings except the square wave function.

So amplitude is correct on square wave but not other waveforms? I recall that square wave uses a different low pass reconstruction filter than used for other waveforms, to get fast edges with low overshoot.  Maybe that hints at where the fault is.

[Shay]

Let me write that in other words: The gen has an output Z of 50 Ohm, your scope must be set to an input impedance of 50 Ohm too else the voltages on the scope and the set voltage in the gen will not match.

Your DMM most likely has an input Z of about 10 MOhm. Therefor it will never show the same voltage as set on the gen unless you have explicitly configured the gen to do so. With standard settings it will show about twice the set voltage.

You are indeed correct, I wasn't clear enough - the 50 settings does increase the voltage by 2. High Z automtically cuts it off in half. This function works alright.
I noticed the ampltiude error is on all settings except the square wave function.

So amplitude is correct on square wave but not other waveforms? I recall that square wave uses a different low pass reconstruction filter than used for other waveforms, to get fast edges with low overshoot.  Maybe that hints at where the fault is.

Seems like this is related to the pre-attenuator. ill check this section and report back.