¿ªÔÆÌåÓý

  1. HP-Agilent-Keysight-Equipment
  2. Topics

HP 8643A IO board fault

 

Hey y'all,

I turned on my HP 8643A the other day, and it came up with an error code 15.022.116, which implicates the I/O board. This finally motivated me to pull it down from the rack and replace it with the HP 8665A I'd acquired some months ago. The 8665A is heavy, definitely a two person lift to get it to eye level.

Now, I just started tearing into the 8643A, my fear was the backup battery on the CPU board had leaked, but it's just fine.
The instrument checks all point the same way, and I've isolated this as far as the built-in diagnostics will take it.
In initial spelunking the thing I found curious is that the?+5V on the IO board is at ~4.7V. This goes through a R40, a 1Ohm resistor, to local decoupling, and this resistor is dropping about 0.5V. So the IO board is using about 500mA, which seems high. I have the CLIP from Artek, which has full schematics for the alleged culprit.

I suspect I'll find something on the board has started current-hogging off the 5V rail, but I'm damned if I know how to hunt it down. Maybe I can borrow an infrared camera and see what's getting hot, though the hog's only consuming 2.5W in the limit.
Have you all been inside one of those or it's brethren? Any common failures that I should look out for?

Siggi
tmillermdems
 

?
Hi Siggi,
Where are you located?? I?have an IR camera.
?
Regards

----- Original Message -----
From: Siggi
Sent: Sunday, October 28, 2018 11:51 AM
Subject: [HP-Agilent-Keysight-equipment] HP 8643A IO board fault

Hey y'all,

I turned on my HP 8643A the other day, and it came up with an error code 15.022.116, which implicates the I/O board. This finally motivated me to pull it down from the rack and replace it with the HP 8665A I'd acquired some months ago. The 8665A is heavy, definitely a two person lift to get it to eye level.

Now, I just started tearing into the 8643A, my fear was the backup battery on the CPU board had leaked, but it's just fine.
The instrument checks all point the same way, and I've isolated this as far as the built-in diagnostics will take it.
In initial spelunking the thing I found curious is that the?+5V on the IO board is at ~4.7V. This goes through a R40, a 1Ohm resistor, to local decoupling, and this resistor is dropping about 0.5V. So the IO board is using about 500mA, which seems high. I have the CLIP from Artek, which has full schematics for the alleged culprit.

I suspect I'll find something on the board has started current-hogging off the 5V rail, but I'm damned if I know how to hunt it down. Maybe I can borrow an infrared camera and see what's getting hot, though the hog's only consuming 2.5W in the limit.
Have you all been inside one of those or it's brethren? Any common failures that I should look out for?

Siggi
 

Hey Tom,

thanks for your kind offer, but I think I can snarf one from work. To answer your question, though, I'm in the frozen tundra of Montreal, Quebec. Are you nearby?

Siggi

On Sun, 28 Oct 2018 at 12:01 tmillermdems <tmiller11147@...> wrote:
Hi Siggi,
Where are you located?? I?have an IR camera.
?
Regards
 

On Sun, 28 Oct 2018 at 11:51 Sigur?ur ?sgeirsson <siggi@...> wrote:
The instrument checks all point the same way, and I've isolated this as far as the built-in diagnostics will take it.

I forgot to mention that I have two LED error indicators lit on the IO board, looks like they're designated LED_BUS(5) and LED_BUS(6). I haven't seen any mention of those in the service manual, so I don't know what they're trying to tell me - alas.
tmillermdems
 

?
Ok, that is good. I find them to be very useful for these kinds of problems. Also, you might try unplugging modules as you monitor the current. Something is dragging it down.
?
I live in Forest Hill, Maryland, USA. A good bit south of you. I have been to Montreal quite a few years ago. Quite a beautiful place I might say.
?
Good luck, I like the 8643, 4 series generators. it would be nice to know more about reading all the test? points (mux readings) for these units.
?
Later,
Tom

----- Original Message -----
From: Siggi
Sent: Sunday, October 28, 2018 12:49 PM
Subject: Re: [HP-Agilent-Keysight-equipment] HP 8643A IO board fault

Hey Tom,

thanks for your kind offer, but I think I can snarf one from work. To answer your question, though, I'm in the frozen tundra of Montreal, Quebec. Are you nearby?

Siggi

On Sun, 28 Oct 2018 at 12:01 tmillermdems <tmiller11147@...> wrote:
Hi Siggi,
Where are you located?? I?have an IR camera.
?
Regards
 

On Sun, 28 Oct 2018 at 13:12 tmillermdems <tmiller11147@...> wrote:
Ok, that is good. I find them to be very useful for these kinds of problems. Also, you might try unplugging modules as you monitor the current. Something is dragging it down.

Maybe not - I don't find anything getting hot. However, the -15V rail is MIA from the near end of R41, which has gone open. Dead short to ground on the -15V rail, I guess that's my problem :). It's going to be those axial tantalum caps C46/C47 I bet.
All the analog MUXes use the -15 rail, so if it's out that'd throw the diagnostics for a real spin.
 

On Sun, Oct 28, 2018 at 13:51 Sigur?ur ?sgeirsson <siggi@...> wrote:
On Sun, 28 Oct 2018 at 13:12 tmillermdems <tmiller11147@...> wrote:
Ok, that is good. I find them to be very useful for these kinds of problems. Also, you might try unplugging modules as you monitor the current. Something is dragging it down.

Maybe not - I don't find anything getting hot. However, the -15V rail is MIA from the near end of R41, which has gone open. Dead short to ground on the -15V rail, I guess that's my problem :). It's going to be those axial tantalum caps C46/C47 I bet.
All the analog MUXes use the -15 rail, so if it's out that'd throw the diagnostics for a real spin.

It was C29, a .22uF 35V SMD tantalum capacitor. All better now, though I need to order a suitable replacement.

So I guess nothing too unusual here. The original fault was a shorted tantalum in one of the RF modules. After that I needed to recap the PSU, then back to a tantalum...
tmillermdems
 

?
Well played. I repaired a 8644B that was dead above 1030 MHz. It was a bad SPDT microwave Gas switch chip on the doubler module. HP used a lot of these chips (AS002M2-12) that are still available. I see a lot of these switches fail with a lot of loss. The module passes the signal straight through below 1030 MHz then switches in the doubler above that frequency. The same circuit is used in the 8657B generator doubler module.
?
Source here:
?
?
?
?
?

----- Original Message -----
From: Siggi
Sent: Sunday, October 28, 2018 3:30 PM
Subject: Re: [HP-Agilent-Keysight-equipment] HP 8643A IO board fault

On Sun, Oct 28, 2018 at 13:51 Sigur?ur ?sgeirsson <siggi@...> wrote:
On Sun, 28 Oct 2018 at 13:12 tmillermdems <tmiller11147@...> wrote:
Ok, that is good. I find them to be very useful for these kinds of problems. Also, you might try unplugging modules as you monitor the current. Something is dragging it down.

Maybe not - I don't find anything getting hot. However, the -15V rail is MIA from the near end of R41, which has gone open. Dead short to ground on the -15V rail, I guess that's my problem :). It's going to be those axial tantalum caps C46/C47 I bet.
All the analog MUXes use the -15 rail, so if it's out that'd throw the diagnostics for a real spin.

It was C29, a .22uF 35V SMD tantalum capacitor. All better now, though I need to order a suitable replacement.

So I guess nothing too unusual here. The original fault was a shorted tantalum in one of the RF modules. After that I needed to recap the PSU, then back to a tantalum...

1 - 8 of 8