HP/Agilent quality. Not a repair but a rant.

[WattsThat]

Bought a very nice HP branded E3631A triple power supply a while back to provide current limiting, something my 6236B did not have. It arrived, looked new, tested, put aside because I didn't have a proper place on the bench. After some juggling, I put it on the bench and attempted to use this past weekend. The -25V supply was goofy, would jump from -12 to -36 volts with a minimal load. Damn, I know i tested that. Find the schematics, open it up and find the -25 supply has a horrendous 10 volt 40 KHz ripple on it. WTF...

Schematic shows 470uf and 100nf caps on the output bus. Hang a 470uf on the output and presto, all is good, no more oscillation. Go looking for caps, cannot find them on the board, pull the front panel and what do I find but wung-hung-low 100uf caps with the -25V terminal in the condition shown. Nice job HP QC!

[TK]

Where did you buy it?  If it was used, how do you know the bad repair was made by HP and not the seller?

[WattsThat]

It came from eBay, not that it matters.

What you’re looking at is what came out of the HP factory in Korea, it’s not a repair. The front panel had never been removed, I had to break the blue thread locker that was on the four front panel screws.

[bdunham7]

You may as well have just bought BK.

[CaptDon]

It is so sad that names once associated with quality products doesn't
stand for anything anymore. Bill and Jack are probably spinning in their
graves after the split-tail C.E.O. sold of the HP test equipment brand to
be absorbed as Agilent and decided the quickest way to kill HP would be
getting into personal computers and printers, like there wasn't already
100 competitors doing a better job!! Didn't she run for public office at
some point??

[DaJMasta]

It came from eBay, not that it matters.

What you’re looking at is what came out of the HP factory in Korea, it’s not a repair. The front panel had never been removed, I had to break the blue thread locker that was on the four front panel screws.

There is not HP factory in Korea making test equipment.  HP became Agilent became Keysight and makes their equipment in Malaysia, typically.


If you bought an HP branded ANYTHING on ebay it has had at least one prior owner and is at least a decade and a half old (the brand may have stuck around on some old stock gear more recently than the Agilent renaming), and I have no idea why an unsoldered, unspecified cap with a manufacturer that doesn't match anything else in the BOM should be considered an official factory installation.

[anotherlin]

It rather looks like it has been modified and/or repaired.
I've not checked the schematics, but I don't see HP/Agilent/Keysight put such a high valued (470uF) capacitor in parallel with the output.
That would not be acceptable in the design of a high "quality" bench supply.

[WattsThat]

It came from eBay, not that it matters.

What you’re looking at is what came out of the HP factory in Korea, it’s not a repair. The front panel had never been removed, I had to break the blue thread locker that was on the four front panel screws.

There is not HP factory in Korea making test equipment.  HP became Agilent became Keysight and makes their equipment in Malaysia, typically.


If you bought an HP branded ANYTHING on ebay it has had at least one prior owner and is at least a decade and a half old (the brand may have stuck around on some old stock gear more recently than the Agilent renaming), and I have no idea why an unsoldered, unspecified cap with a manufacturer that doesn't match anything else in the BOM should be considered an official factory installation.

Wow. So now I guess someone is counterfeiting HP Korea serial number labels. Yeah, that makes sense. Got it.

Some of you guys must doubt what you see with your own two eyes. Sad.

PS: HP Korea opened their doors in 1984. Easy to confirm, not that you care.

[alm]

Changing a 470uF output cap for a lower value sounds like a modification someone might do to improve programming speed or improve behavior as current source.

Unless an instrument is in an unopened factory box, it's hard to know if an instrument was modified or repaired in the past ~20 years.

[pqass]

Wow, the caps are legit! Seems kind of large.    2x470uF and 1000uF; see pages 130,131, C42,C43, and C32.
Service manual: https://dev.xdevs.com/documents/12

Other beauty shots: https://doc.xdevs.com/docs/HP_Agilent_Keysight/E3631A/Unit_2/img/002/

[vk6zgo]

Even back in the "glory days" HP had their moments.
A piece of equipment. (I think it was an N&D set, but too long ago) would every now & then, blow the fuse on the ac power distribution panel feeding it.

It turned out that they used a 3AG fuseholder with a protruding connection on the back, which came perilously close to a metal partitition inside the instrument.
The partition was only secured at the bottom, & could sometimes flex if the device was moved or stressed in some way, shorting the Mains Active to ground.

Quick fix?-----A piece of perspex glued onto the panel with super glue.

I wonder if the normal fuseholder in that spot was a 2AG, & the longer one was the result of a temporary parts shortage.

They were in good company, however, Marconi left a spare 2BA nut loose in one of their instruments, which blew the HT fuse from time to time, till we found it.

[WattsThat]

Even back in the "glory days" HP had their moments.
A piece of equipment. (I think it was an N&D set, but too long ago) would every now & then, blow the fuse on the ac power distribution panel feeding it.

It turned out that they used a 3AG fuseholder with a protruding connection on the back, which came perilously close to a metal partitition inside the instrument.
The partition was only secured at the bottom, & could sometimes flex if the device was moved or stressed in some way, shorting the Mains Active to ground.

Quick fix?-----A piece of perspex glued onto the panel with super glue.

I wonder if the normal fuseholder in that spot was a 2AG, & the longer one was the result of a temporary parts shortage.

They were in good company, however, Marconi left a spare 2BA nut loose in one of their instruments, which blew the HT fuse from time to time, till we found it.

Great story! I never cease to be amazed at what comes of of factories and goes undetected for years. Years ago, I repaired countless HP devices from a specific product line and it got to the point I count tell if the assembler was right or left handed, depending upon the visual clues.

The quality of the US locations was quite good and the original off shore operation in Singapore was on par with the US. Then the cost cutting started in the early nineties and you could see the quality of parts and assembly go down as things moved across Asia multiple times looking for the best labor rates. That’s about the time they spun off Agilent and the rest we know.

[jackthomson43]

Years ago, I repaired countless HP devices from a specific product line and it got to the point I count tell if the assembler was right or left handed, depending upon the visual clues.

Now, that piece of knowledge is going in my  permanent memory storage. I never thought of that before. :O

So, we can check from soldering whether he is right or left handed, I am going to test it out. I ahve to do soldering for Microcontrollers.

[anotherlin]

Wow, the caps are legit! Seems kind of large.    2x470uF and 1000uF; see pages 130,131, C42,C43, and C32.
Service manual: https://dev.xdevs.com/documents/12

I've double checked, yes indeed.
This is not very good for a bench power supply

[WattsThat]

Years ago, I repaired countless HP devices from a specific product line and it got to the point I count tell if the assembler was right or left handed, depending upon the visual clues.

Now, that piece of knowledge is going in my  permanent memory storage. I never thought of that before. :O

So, we can check from soldering whether he is right or left handed, I am going to test it out.

Never said it was the soldering. Your attempt at sarcasm fails. Next contestant please.

[bd139]

I've had an E3630A with similarly shitty soldering and caps. The early 90s and onward line from HP was a bit hit and miss compared to the earlier bits of gear.

The thing that really fucks me off with the E3630 is to calibrate it you have to poke around near the exposed mains switch pins. 

[WattsThat]

I've had an E3630A with similarly shitty soldering and caps. The early 90s and onward line from HP was a bit hit and miss compared to the earlier bits of gear.

The thing that really fucks me off with the E3630 is to calibrate it you have to poke around near the exposed mains switch pins. 

Agreed. Why they didn’t cover the exposed switched contacts is indicative of the cost driven engineering that took over. Crap, a 5 cent piece of scored and formed fish paper would have been better than nothing. A piece of sheet polycarbonate/polyester bent to shape would have been the norm a few years earlier.