Agilent 54825A Infiniium Oscilloscope - Dead Power Supply [FIXED]

[sean0118]

Hi All,

I've had a 500Mhz Agilent 54825A die on me while it was left on standby

The power supply is completely dead, blown fuse on input and shorted  IRFP450 transistors in the input PFC stage (Q102, Q103). The power supply is a C&D Technologies NX300-6003 (HP PN 0950-3252).

I've already ordered replacements for these, couldn't see any other issues, so hopefully it works with these replaced. The controller for them seems to be a custom black box? It's the one marked MCDI 02-37 3740039 REV D1.


One weird thing I noticed, there's an Airpax 67F090 with one lead not attached, no solder on the lead so looks like it's been that way since the factory? It's a normally open bimetal disc thermostat.

Does anyone have an opinion on whether I should leave the lead disconnected or not? It looks like there's a pad for it, but I'm really only guessing that's what it's meant to be soldered to. There are some factory mods around that area of the PCB too.


Let me know what you think. 

[ttbelectronics]

Sean the AirpaX should be soldered at points E20 and E21 on the PCI.

[sean0118]

Sean the AirpaX should be soldered at points E20 and E21 on the PCI.

Thanks! Btw how do you know this? Do you have a schematic or the same power supply? 



edit: I'm guessing this is your company? Good to know, bookmarked for future reference. 
http://www.ttbelectronics.com/power-supplies

[Dacke]

I'm not familiar with that particular supply but if all three pins on those PFC switchers were shorted (gate being shorted to drain and source) you may want to also check any pre-driver transistors if it has them,  and there could be damage to the oscillator / hybrid as well.  Also check current sense resistors and any snubber or rectifier diodes in the PFC circuit.  Hopefully nothing aside from those two transistors has failed,  but cascade damage in these type of circuits is not uncommon.

[sean0118]

Thanks everyone, just an update this is now fixed. 

Replacement parts were:

1x Fuse, Bussmann / Eaton, BK/ABC-8-R
2x MOSFET, Vishay IRFP450APBF

[ttbelectronics]

I have back engineered many power supplies from Agilent / Keysight, do not supply schematics or any further information.
I'm glad you got your unit running.

[Rana Sajjad Hussain]

I am also facing Agilent 54825A power supply(NX300-6003) failure issue. The power MOSFET was short which is now replaced. But some of the resistors are blown and i don't know their values. Can any one provide me their values or at least pictures of PCB from bottom. Blown resistors are shown in picture attached

[rogerymw]

My one even worst....all disappeared!

anyone can help to measure the voltage of area marked "1" and "2"?
I may need to build another aux power to replace those disappeared.

[sean0118]

I'll see what I can do, remind me if I don't post back in the next week or so.

anyone can help to measure the voltage of area marked "1" and "2"?

Is it just DC voltage between these points you need?

[rogerymw]

Yes, voltage between 2 points at "1" and voltage between 2 points at "2"
I will build a aux power supply to try to power it up.

The PFC controller and both IRF450 dead too, I will try to feed right voltage
to point 1 & 2 and feed 400V after the PFC stage to see it will power up or not.

Thanks for your help

[rogerymw]

Hi Sean,

any update?

Thanks in advance

[sean0118]

Okay, sorry things got really busy, I just took the PSU apart and scanned the bottom of the PCB. I hope it's still useful to someone. 

I still need to measure the requested voltages, I will have that done very soon.

PCB version is 6203120 REV B:

[sean0118]

I went to measure those voltages but Q102 & Q103 are dead again! 

I powered it up with the PCB out of the enclosure, I wonder if that is the cause? Looks like there's no continuity between some earth connections when out of the enclosure, I know it's an issue which can cause some TV power supplies to blow up.

Or maybe it was dead before I took it apart again, it's been sitting on the bench for a few months without use. 

[pandrew]

Hello all,

Sorry for reviving an old post, I am also repairing now the same model power supply, and I created my own post with my progress, please see here: https://www.eevblog.com/forum/repair/repairing-a-agilent-infiniium-54825n-power-supply-(nx300-6003)/
If you have any advice I'd appreciate it.

@Dacke you were correct with your prediction, at least on my board, there was cascade failure. There are no pre-driver transistors, but the PFC controller chip failed, and series resistors going to the gates of the mosfets also failed. And the rectifier bridge also failed. Not sure what you mean by "oscillator/hybrid" maybe this power supply doesn't have it? I also don't see current sense resistors, but I see some components that look like might be current sense transformers, the ones I can read say "HIPOTTED 38 02". Another one says "28675A CHT133001B". Can't find datasheets for these. Anyway, these just have a thick wire going through them, no way they could have failed.

@sean0118 Did you manage to fix yours a second time?
I would be so surprised if running it out of the case caused it to die. I only see some ceramic SMD capacitors, and a pair of Y2 filtering capacitors connecting to the case mounting screw terminals. All the heatsinks have isolating ceramic pads. The only exception is the AIRPAX 67F090 thermostat, but that one doesn't electrically connect its heatsink to it's other pins. Do you have a reference I can read up on of such an example power supplied that blew up because of some missing earth connections?

Btw I had to drill out one of the screw terminals, cause it was rotating freely with the screw, and it was right under the transformer, so I couldn't grab it. Another one was rotating freely but I could grab it with some pliers to get the screw out. I added solder to all of the terminals afterwards, and that fixed the freely rotating one that I didn't have to drill out. I wonder if I should be running an earth wire to that capacitor.

[pandrew]

@sean0118 - After fixing my board I have powered it up without issues out of the enclosure, I don't think that's why your power supply failed. However I did get another fault with a PFC switcher mosfet going shorted, and that happened after I unplugged the power supply and plugged it back in again less than 3 seconds later. I think these power supplies may be sensitive to that. Maybe that's why your power supply died again?

I measured the big capacitor, and when the board is unplugged it discharges from 380V to about 30V "fast", in about 16 seconds. The rest of the way it discharges very slowly over tens of minutes. (this was with the power supply not connected to the oscilloscope)

Since then I did multiple unplugs and replugs, waiting 16 seconds between them, and everything's still working reliably. I don't wish to kill my power supply again, so I won't attempt to collect data to see if shortening the wait time would kill it again. I will keep my oscilloscope unplugged when unused, maybe even use it under a UPS only. Our power distributor is known for both short occasional power outages, and for short occasional pulses too.

@rogerymw - I have the same revision of the board as @sean0118 has. However yours seems slightly different.
I have identified similar structures as you pointed out and measured those, (area 1 is very similar, area 2 is only slightly similar) however the results may not apply to you, anyway, here you go:

When power supply is connected to nothing:
Area 1: 18.015V
Area 2: 17.899V

When power supply has a resistor attached to the control cable as shown in the troubleshooting manual of the scope:
Area 1: 17.950V
Area 2: 17.886V

I did not measure with the power supply connected to the oscilloscope.