¿ªÔÆÌåÓý

HP8920B Power supply capacitors??


 

I guess I will take a look and see what I can find. Does anyone know if the PS can be voltage tested outside the unit without load?


 

¿ªÔÆÌåÓý

Hi?

From my experience switch mode power supplies require 50% load on the Maine supply? usually the 5 volt supply for them to work properly on the other rails

?


From: [email protected] [mailto:[email protected]] On Behalf Of RFI-EMI-GUY
Sent: 04 October 2019 18:35
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8920B Power supply capacitors??

?

I guess I will take a look and see what I can find. Does anyone know if the PS can be voltage tested outside the unit without load?

No virus found in this message.
Checked by AVG -
Version: 2016.0.8048 / Virus Database: 4793/15886 - Release Date: 08/14/18
Internal Virus Database is out of date.


 

On 10/4/19 5:34 PM, Paul Bicknell wrote:
From my experience switch mode power supplies require 50% load on the
Maine supply? usually the 5 volt supply for them to work properly on the
other rails
Be careful here, some switching power supplies require a load and some
don't. Most very early ones (meaning WAY earlier than the HP 8920) do,
while essentially nothing newer does. My rule of thumb has the dividing
line falling around the early to mid 1980s, but it varies depending on
who made it.

-Dave

--
Dave McGuire, AK4HZ
New Kensington, PA


 

Hi I can only afford Old HP equipment

Paul

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Dave McGuire
Sent: 04 October 2019 22:41
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8920B Power supply
capacitors??

On 10/4/19 5:34 PM, Paul Bicknell wrote:
From my experience switch mode power supplies require 50% load on the
Maine supply? usually the 5 volt supply for them to work properly on the
other rails
Be careful here, some switching power supplies require a load and some
don't. Most very early ones (meaning WAY earlier than the HP 8920) do,
while essentially nothing newer does. My rule of thumb has the dividing
line falling around the early to mid 1980s, but it varies depending on
who made it.

-Dave

--
Dave McGuire, AK4HZ
New Kensington, PA






-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2016.0.8048 / Virus Database: 4793/15886 - Release Date: 08/14/18
Internal Virus Database is out of date.


 

By "old" in the context of switching regulators I mean 1960s and
1970s. Most stuff from the 80s on does not require a load for proper
regulation. There are exceptions, of course.

-Dave

On 10/4/19 5:56 PM, Paul Bicknell wrote:
Hi I can only afford Old HP equipment

Paul

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Dave McGuire
Sent: 04 October 2019 22:41
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8920B Power supply
capacitors??

On 10/4/19 5:34 PM, Paul Bicknell wrote:
From my experience switch mode power supplies require 50% load on the
Maine supply? usually the 5 volt supply for them to work properly on the
other rails
Be careful here, some switching power supplies require a load and some
don't. Most very early ones (meaning WAY earlier than the HP 8920) do,
while essentially nothing newer does. My rule of thumb has the dividing
line falling around the early to mid 1980s, but it varies depending on
who made it.

-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA


 

By old I mean I mean 1960s and 1970s.

OK actually HP 9000 computers

Best regards Paul

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Dave McGuire
Sent: 04 October 2019 22:59
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8920B Power supply
capacitors??

By "old" in the context of switching regulators I mean 1960s and
1970s. Most stuff from the 80s on does not require a load for proper
regulation. There are exceptions, of course.

-Dave

On 10/4/19 5:56 PM, Paul Bicknell wrote:
Hi I can only afford Old HP equipment

Paul

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Dave McGuire
Sent: 04 October 2019 22:41
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8920B Power supply
capacitors??

On 10/4/19 5:34 PM, Paul Bicknell wrote:
From my experience switch mode power supplies require 50% load on the
Maine supply? usually the 5 volt supply for them to work properly on the
other rails
Be careful here, some switching power supplies require a load and some
don't. Most very early ones (meaning WAY earlier than the HP 8920) do,
while essentially nothing newer does. My rule of thumb has the dividing
line falling around the early to mid 1980s, but it varies depending on
who made it.

-Dave

--
Dave McGuire, AK4HZ
New Kensington, PA






-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2016.0.8048 / Virus Database: 4793/15886 - Release Date: 08/14/18
Internal Virus Database is out of date.


 

UPDATE

Well; I took the HP 8920B apart this weekend. It has the same power supply as pictured here (photo from e-bay listing). This also appears to be same as the schematics kindly provided by Ron Wilkinson earlier in this thread.

Weird; I turned the unit on with the covers off and could smell none of the odor I detected before. It was behaving.? I was hoping to see a bit of smoke or a part overheated. I used an UV lamp to see if I could detect any electrolyte leakage, did not see anything, though I don't know if this would work.

With the PS still installed I could observe most of the components in the 120V switcher circuit from the edge (bottom) and I used an IR thermometer to check temperatures after about an hour. This was done at room temp 73F and I found that the hot spots were:

T1 127F - at top of core
Q1 116F - main switcher
C7 113F - a Rifa capacitor .047 630V
J5? 105F - CORCOM input filter. This seems to heat quickly.
L1 106F - Input choke
L2 , C5 and C6 103F - Output of rectifier

Most of the heat is on the side of the transformer closest to mains. Partly because of IR drop of parts and mostly the fan air circulation seems to ignore this side of the board, at least with covers off.

I was hoping to see a part with obvious distress but none was spotted. I remain suspicious of the Rifa capacitor C7 and the bulk filters C5 and C6 . Those two look fine, however the end of the capacitors where the pressure relief exists are covered with a plastic disc so if they vent slightly, that may be unnoticed.

Their is a 10 ohm NTC inrush thermistor RT1 that could be making a puff of heat when it responds. I did see some silicon glue near it (forgot to remove) that may be getting singed. I could not get a temp from this part or the rectifier near it.

I am going to order a crash kit of parts so that next time I can do surgery.

I have read that some capacitors are "self healing" . I don't know if that applies, nor how many times they can heal.

I think my Monitor is a a 1998-1999 vintage I bought used in DEC 2002. Since then its use has been very light a few hundred hour a year at most. When received It had some light screen burn in, no dust inside and had the odor from the beginning.


 

Marc;
Would you by chance have a copy of your parts order so I can order the same set of caps for mine?? Also if you can point out C63 and C65 in the photo above or the schematic, I would be much obliged.

Joe


On Mon, Sep 30, 2019 at 02:23 AM, marcschrdrs wrote:
I had a 8920A recently on the bench with "magic smoke" issues: suspected RIFA X2 issues from the smell.

Turned out there are 3 Rifa's in the PSU circuit proper, one blown and 2 cracked in my case (C7, C63, C65).
Replaced these with Wima mkp10 47n/630V caps (15mm version MKP1J024704C)

There is one more Rifa on the PSU pcb: a 2.2nF Y2, replaced that one for good measure as well.

It is a pain getting the PSU out of the machine, and then the PCB from its case, but for the former problem there is a nice explanation in the hp manual floating around.

good luck,

Marc.


 

Joe,

sure, I ordered this cap (3 EA):
Now, it seems my PS is slightly different from the one you pictured, so yours may very well not have C63/C65.
I found a pic showing (barely) either C63 or C65 (attached) and the circuit seems different (eg 3 coils vs 1 coil)

Marc.


 

Hi,
Sorry I'm lte to this been abroad for work.
I can state that the 8920A/B power supply module will work fine without any load. I've done it more than once. When working on one recently I ran it from the DC input as this is somewhat safer than running it from a mians supply. The DC input runs a converter that provides about 300V DC on the same filter capacitor as the AC mains input.

Robert G8RPI.


 


Thanks, that is good to know as I could use a current limited DC supply while testing unloaded.


On Wed, Oct 16, 2019 at 03:03 PM, Robert G8RPI wrote:
Hi,
Sorry I'm lte to this been abroad for work.
I can state that the 8920A/B power supply module will work fine without any load. I've done it more than once. When working on one recently I ran it from the DC input as this is somewhat safer than running it from a mians supply. The DC input runs a converter that provides about 300V DC on the same filter capacitor as the AC mains input.

Robert G8RPI.


 

have re-capped numerous 892XA/B Supplies ...at this age they all just about need them before the PCB gets damaged.

Yes they are a pain to get out ....be careful on removing the switch "extender" ....the plastic is old and breaks easily. ?? The RF I/O has to come out first.

I usually just re-cap the low voltage caps ....the two high voltage ones never seem to need it though I have replaced? a lot proactively .

Make sure you get the cap on the low voltage input PCB ( mounted on the side )


You can usually lay the power supply on the top to do a final test before installing so it will have the correct load as well.


 

Hi, thanks for a lot of information.

I would like to know more about RT1: you say that is a 10 ohm NTC.But, the ID in one of sides is 0033. Do you now how I can get a similar one?

Thanks in advance.

Gabriel


 

On Sun, Oct 13, 2019 at 08:20 PM, RFI-EMI-GUY wrote:
C7 113F - a Rifa capacitor .047 630V
J5? 105F - CORCOM input filter. This seems to heat quickly.
The Rifa X and Y capacitors from the 80s are replace on sight items - can use new Rifa or other brands as long as they fit, there is not much space in there.

The IEC mains input filter would be on my list, again looks very tight in there and finding a replacement (new not NOS) that fits would be tricky. When these these blow the result is not pretty.

Regards,
John


 

I have acquired a 8920a.? Dead on power up..
Will not power up on 110v or 12v
I have carefully removed the power supply. I dont see anything obvious in the way of damage the supply is completely dead on bench testing fuses don't blow on power up...

Any suggestions ?? I have heard that recapping will solve the issue ???
Does someone offer a complete kit of needed caps ?? ?

Are there some caps that should be changed no matter what ?..not exactly sure where to start on this ...seems as most say it is an incredible Pia to repair...

Thank you

Dave ke0ssm


 

Hello,

Did you probe any power pins, is it outputting any power at all? Be careful though, some of the heatsinks have power through them, it'll wake you up. I just got done rebuilding a PSU that threw flames so hopefully I can help, you can also check for power between the two heatsinks that hold CR23,24,25,26. There are no bleed resistors for the two large caps so be careful even when off.


 

I probed all the pins. Have powered up on 110v AND 12v.? Nothing even comes close to voltages marked on pins.? ?I Have checked for obvious physical damage there is none.? No burnt traces or bulging caps ( however they do look original )I have probed the power switches and all are functioning correctly.? Not exactly where the test points are on this ps or what voltages I should see at those points..

No smell. And no history on the unit as a hole.? It was received to me no power up


 

Check RT1 to make sure that is passing power. That was the first thing to go in mine. It's under the AC fuse next to CR1, the big grey square cap at the corner


 

Ok. On 110v. Big square gray cap.. marked c2. .115 v one side nearest electrolytic and .848 on other side

But this wouldn't affect the 12v side of the power supply as well ?


 

On further looking I found rt1 (varistor? Mov ?)

Ground to each side of it shows 59v.?
?It does show some signs of distress. ( slight discoloration and possible crack.? While still in circuit measures 13.4 ohms

Not sure what a good replacement for this item would be is there a testing option to verify?