| Products > Test Equipment |
| HP/Agilent 54825A Oscilloscope Power Supply Issues |
| (1/3) > >> |
| G0HZU:
Hi In recent months my old HP/Agilent 54825A Infinium scope has been playing up and it has slowly been getting worse and worse. To cut a long story short I think the PSU is generating an alarm flag and this cripples the scope such that it shuts down some of its operation and the trace freezes and often I see a red error message that reads something like "fine interpolator gain cal failed". This error can happen right from the first power up of the day or it can happen after 30 seconds or several hours. Often a reboot restores operation but the scope has been getting worse and worse in that this is becoming a stubborn fault. I've noticed a few symptoms: The AUX OUT signal dies at the same time the scope trace freezes and I think this is a normal reaction for the scope when it receives the alarm flag from the PSU. After a few seconds or minutes the scope then displays the (false?) "fine interpolator gain cal failed" error message. After much wiggling of things and messing with settings and temperature I kept coming back to the power supply and I now think I've found what is causing the scope to go into partial shutdown. By monitoring a test point on the PSU I can tell from the moment the scope is powered on if it is going to be able to run the scope loader program. This program dumps some files into a ramdisk drive and then launches the scope application under Windows 98. However, if the PSU alarm flag is already set then I know the scope launch is doomed to failure. It typically freezes during the loader operation if this flag is set and doesn't even get as far as loading the scope app into the ramdisk. However, if the alarm flag is OK then the scope will launch just fine. Unfortunately, if I monitor the alarm flag then I can tell within a few seconds when the scope trace is going to freeze followed by the "fine interpolator gain cal failed" error message appearing on screen. This happens right after the PSU alarm flag is set by the PSU itself. If I monitor my 5V 12V -12V and -5V rails from the PSU they all look fine except the 12V rail has a small amount of spiky ripple on it. These rails seem to start up correctly and stay at the same voltage regardless of the PSU alarm flag. Has anyone seen this behaviour before? I'm left wondering if one of the supply rails has drifted a bit out of tolerance and the PSU has condemned itself via the alarm flag. If I isolate the alarm flag from the rest of the scope then the scope always seems to work fine but I know this is asking for trouble further down the line. Has anyone worked on these PSUs before? Any info about how to adjust them or test them? Thanks in advance for any replies... |
| tautech:
--- Quote from: G0HZU on July 02, 2022, 10:35:32 pm ---Hi In recent months my old HP/Agilent 54825A Infinium scope has been playing up and it has slowly been getting worse and worse........ I'm left wondering if one of the supply rails has drifted a bit out of tolerance and the PSU has condemned itself via the alarm flag. --- End quote --- Me too. How many runtime hours and boots ? PC based mainboard with SMPS PSU ? Have you tried subbing in another PC PSU to see if it goes away ? Sure smells like dicky PSU or mainboard caps to me. |
| G0HZU:
Hi Tautech This scope is old... probably 25 years old and so the run time will be huge. It's had the official factory upgrade pack so the motherboard/CPU/memory/HDD has been upgraded and it has the LS120 floptical drive. But the PSU looks to be original. The PSU is really complicated and not easy to work on. There's some images of the PSU in this old thread below. This looks to be the same model as my PSU. It would be nice to have a manual for the PSU or a circuit diagram or some idea how the PSU interfaces with the rest of the scope. I've traced out some of the interface circuit but I'm not sure what this alarm/status flag actually means. It might not even be an alarm/status flag but it causes the scope to hang when it gets set. https://www.eevblog.com/forum/repair/agilent-54825a-infiniium-oscilloscope-dead-power-supply/ Presumably the PSU can test itself in some ways. This might include temperature and current. The PSU consumes about 5W in standby and typically 170W when the scope is running. This is measured using a wattmeter on the AC inlet. There might be other reasons the scope needs to communicate with the PSU. There is a separate 'line' feed to the scope board but I think this is just for a line triggering option. Apart from the usual large white 5V/12V/-12V/-5V connectors from the PSU there's also a smaller 11 wire interface from the PSU to the scope mainboard. I suspect the alarm may have something to do with the spiky ripple on the 12V output. But then again the 12V supply probably doesn't have to be that clean so maybe this is normal... |
| G0HZU:
Things are looking hopeful... I'm going round it with an impedance/ESR meter and I've found one totally dead and corroded/leaky electrolytic cap. It's near the 12V section so this might be the cause of the fault. It's hard to get to some of the caps but I'm going to try and access and test them all this afternoon. I'm using an Analog Discovery 2 and "TheStuffMade" (Jaxbird) ESR software/jig to trace out the caps over frequency. Obviously, I've made sure the PSU caps are fully discharged before testing. I keep meaning to buy a DER-5000 but the setup above works really well if set to fast sweep mode over (say) 100Hz to 100kHz. |
| HighVoltage:
That is an interesting Impedance Analyzer for Analog Discovery. Thanks for sharing! |
| Navigation |
| Message Index |
| Next page |