Has anyone else ran into this with these Keysight DSOS oscilloscopes? Warning "Timebase error detected in PLL2"? I've had four of these recently with the same exact error and they had to go back to Keysight's Service Center in Cali. Luckily these have still been under warranty but I'm starting to notice a trend. Wonder if it's a bad acquisition board like older scopes...shouldn't be malfunctioning that quickly though.
Yeah, it was the acquisition board. And after having several of these replaced last year at Keysight they are all failing the 50Ω input impedance from 20mV/division down. Reads 62Ω. WTF.
Out of the (Keysight) scopes I personally calibrated over the past 8 years I have not seen any failures.
It would be a shame if Keysight didn't correct the problem ASAP.
The acquisition failures are common. On is related to the Frac N synthesizer"PLL error" the other is related to a failed input attenuator. Generally that is caused by excessive input in 50 ohm in the low ranges.
The oscilloscopes with the Timebase Errors had the acquisition boards completely replaced. This problem is a known issue by Keysight. They actually made a new model card. I suspected the same thing about the department over ranging them. I went out to the production area and watched what they were doing. They are using 100:1 probes during testing. Where I'm at now is that the last one I worked on passed the input impedance test before I ran the full User Calibration + Timescale. After the User Calibration all the channels fail. I currently have sent in 3 to a Keysight service center and 1 to their R&D lab for evaluation. I suspect the User Calibration routine. But it passes the whole thing with green on every channel + AUX. I'll let you know when the issue is resolved. Frustrating.
After further testing I found out that these S-Series only read 50Ω impedance correctly on the 3458A and 34401A. The Fluke 9500B and 8508A won't read the 50Ω impedance on the lower volts per division correctly. The 9500B was on 50Ω setting for resistance. I even tried introducing a 50Ω feed through termination in several configurations and it cut the resistance in half (as expected). This morning I tried using the Fluke 5520A as well in scope mode and it read the impedance high and out of tolerance too!
Through this process I learned you can make the 8508A read correctly if you select LoI in the resistance configuration menu. It changes the measurement current from 10mA setting to 1mA. No such setting for the 9500B or 5520A. I have never had this issue with any Tektronix or Rhode & Schwarz oscilloscope!! They read fine on the 9500B and 8508A without any special setting.
I wouldn't be surprised if Keysight intentionally designed them to only work with their own test equipment. However I still don't understand why this is the case. Now I have to rewrite that section of my Met/Cal procedure to use the 3458A instead of the 9500B :x
I used to run in to similar issue testing the old Tek PG506. I don't recall if it was 50 or 600 ohm msmt. But it needed to have the dmm resistance range set a range higher than you would tend to use, probably related to the current level being pushed thru it.
Keysight confirmed with both the 9500 and 8508A the impedance issue. I was told that it's part of the design to create minimum noise. The 8508A and 9500 are applying an input voltage that is driving the input amplifier beyond it's dynamic range limits. I'll just rewrite my Met/Cal procedure to use the 3458A and be done with it.
Thanks for keeping us updated on this topic.