HP 33120A square wave burst duty cycle anomaly

[corn11]

Hello everyone,

I've recently discovered a strange behaviour of my HP 33120A fgen while in square wave mode with (single) burst turned on.

When the duty cycle is smaller or equals 51% the output changes to the "low" state after the high period of the single square wave pulse (as expected) but when the duty cycle is greater than 51% the output jumps back up after the low period of the square wave pulse. The other functions of the fgen work fine. I've discovered it while using the following settings:

Mode: Square wave
Frequency: 800KHz (also present at other frequencies)
Amplitude: 5Vpp
Offset: +2,5V
Burst: On, 1 Cycle, 100Hz Period, 0deg Phase
Trigger: Immediate/internal
Output: HighZ

SN: US36.. / built around 1998, no options
FW version: 8.0-5.0-1.0

I've already consulted the user/service manuals, various application notes and several forum threads but I haven't found anything about this strange behaviour. According to the manuals it shouldn't be this way (in case I haven't missed something).

It would be great if somebody could check if this problem also exists on his or her machine?

Thank you in advance!



 

[StefanF]

Hello,

Same behaviour with my 33120A.
Seems to be normal.
Also the same on my 33220A!

[corn11]

Hey StefanF,

thank you for cross checking it on your fgens!

I'm glad that my unit is not faulty.

Although, I'm a little bit supprised that apparently no one ever discovered/reported this bug before and that it also exists on the 33220A.
This is still a current model (till december 2016). I wonder whether they'll still try to fix it on the 33220A if someone reports this bug or if they already know about it and classified it as too minor to roll an update (if it's just a software related problem) ?


 

[CustomEngineerer]

The behavior you show kind of made sense to me at first (at least if the top picture represented <= 50% and the bottom picture would be > 50%, but not necessarily for 51% & 52%). What I was thinking maybe HP was doing, for a single cycle burst of a square wave, for the whole waveform to be high more than 50% of the time, it would have to start high then go low and then back high. The top screenshot (<=50) starts low, goes high, then low. The bottom screenshot (>50) starts high, goes low, then high. Also, on the bottom screenshot, it changes on the back half of the cycle (notice both screenshots waveforms go low at the same point in the overall cycle), while the top screenshot appears to change in the first half of the cycle. But...then I tried it on my Siglent SDG2042X, and that is not the behavior I get at all. Regardless of the duty cycle, it always starts low, goes high, then back low for a single cycle burst of a square wave (at least with the settings the same as yours).

I'm still not entirely sure its a bug though, maybe just different philosophies on how it should be done?