HP 3325a Function Generator: Strange output on square wave at higher frequencie

[_rushin_or_draggin_]

Hello everyone!
I’m going to preface this by saying I’m somewhat new to function generators so I’m not exactly sure whether this is normal or not.
Basically, every waveform (sine, triangle, ramp up and down) looks great at their maximum frequencies; however, whenever I switch to the square wave, it looks good until 1 MHz which is when I start getting overshoot. Whenever it gets up to 10 MHz, it’s maximum, it looks really rounded. I’m not sure whether this is normal or something I need to repair. I’m currently in the process of recapping, just waiting for them to arrive. I’ll attach a photo below of my oscilloscope output at 10 MHz.
Thanks everyone!
Josh

[TheMG]

Are you just connecting a BNC cable straight from output of func generator to the oscilloscope? If so, did you enable 50 ohm termination on the scope (if equipped) or connect an external 50 ohm termination at the scope input?

It becomes important at higher frequencies to have matched impedance and properly terminate into 50 ohms. Keep in mind a square wave has strong harmonic content well above the fundamental, and will appear rounded off if not properly matched/compensated.

[_rushin_or_draggin_]

Are you just connecting a BNC cable straight from output of func generator to the oscilloscope? If so, did you enable 50 ohm termination on the scope (if equipped) or connect an external 50 ohm termination at the scope input?.

Yes, I used a BNC splitter and RG58 coax with a 50 ohm load. I also used a regular scope probe and, while I got a better result, it was still rounded.

[Atomillo]

According to specs, the square wave should have rise and fall times of around 20ns, which seems to fit with which you posted.

Consider also that most function generator specify as "maximum frequency" that of the sine;  that of the square is normally (not always like your HP generator and some "TrueArb" Siglent for example) much smaller and significantly deformed at the highest end due to the high frequency harmonics it contains.

[_rushin_or_draggin_]

According to specs, the square wave should have rise and fall times of around 20ns, which seems to fit with which you posted.

That is exactly what I needed to hear! I tried finding the rise/fall time in the manual, and to be honest I think I did see it and didn’t think anything about it because I was worried about the shape! Thanks for your help!
Only problem I have is when I switch to the high voltage output it’s essentially a sine wave, so I’m going to have to dig through the schematic and figure that out.

Thanks!

[Atomillo]

According to specs, the square wave should have rise and fall times of around 20ns, which seems to fit with which you posted.

That is exactly what I needed to hear! I tried finding the rise/fall time in the manual, and to be honest I think I did see it and didn’t think anything about it because I was worried about the shape! Thanks for your help!
Only problem I have is when I switch to the high voltage output it’s essentially a sine wave, so I’m going to have to dig through the schematic and figure that out.

Thanks!

Glad that I could help!

First of all, regarding the shape, the manual also specifies overshoot. This also fits with the results you have shown.

Secondly, the high voltage amplifiers has a much lower bandwidth! In this thread (https://groups.io/g/HP-Agilent-Keysight-equipment/topic/hp_3325a_help_with_photos/30894537) it is quoted a -3db point of 3.5Mhz (again, that matters for a sine, this means good square waves up to about 100kHz or so).
Also in the eevblog there has been cloning of the board with results presented: https://www.eevblog.com/forum/testgear/cloning-the-hp3325a-option-002-hv-output/

Perhaps you could replicate them as a sanity check.

Good luck and good night!

[David Hess]

... however, whenever I switch to the square wave, it looks good until 1 MHz which is when I start getting overshoot. Whenever it gets up to 10 MHz, it’s maximum, it looks really rounded. I’m not sure whether this is normal or something I need to repair.

That is common for function generators at the higher end of their frequency range.  Square waves are difficult to duplicate with a linear amplifier.  Pulse generators do not use linear output amplifiers so can produce cleaner high frequency square waves.

You could check the high frequency calibration of the amplifier.  It might have drifted over time.  If there are any aluminum electrolytic decoupling capacitors, check or replace them.