Has anyone who has recently rebuilt the HP 8935 power supply compiled a list of the capacitors? Preferably with DigiKey or Mouser part numbers?

I need to rebuild mine and from pictures on the web it seems like there are 20+ on the top board and presumably more on other boards. It would be helpful to have the parts on hand before commencing the rebuild.

I may have asked this some months ago, but I am hopeful that perhaps someone has undertaken a rebuild in the interim.

Just for fun, I priced a used power supply from a commercial source and received a quote of around $800 including shipping! Makes sense to do it myself.

Thanks.

KE6CD

On 02/21/2022 2:30 PM, Jeff Otterson wrote:

Some options are to ship to Amtronix

I'd vote for Amtronix. Rick Bowman at Amtronix knows the 8920/8921/8924/8935 family very well and has many spare parts on-hand. The 8920B that I bought from him developed a problem, so I shipped it back to him. I think shipping with insurance was around $250. Rick diagnosed the problem, replaced a module, recalibrated the unit, and shipped it back to me for about $400.

Bob...

开云体育

Depends entirely on the value you place on your time.

My HP8921A is sick.

I believe there are problems with the input module.? The internal diagnostics tell me there are problems with the "antenna input detector" (it reads high when it should be zero), filter output detector (reads low), and auto-range attenuator (reads low on 20db setting).

It also has the symptom of generating the RF overpower alert if even the smallest rubber duck antenna is attached to the antenna jack.

I have looked at the service manual and the "clips" for 8920, I believe the instruments are similar, I'd feel more comfortable if I knew the schematics matched the device...

I can repair stuff, but I am a little bit leery of taking this on.

Some options are to ship to Amtronix or carry locally to Trescal.? A new (used) one can be had for $3k-$4k.

I'm seeking advice and suggestions.

Thanks,
Jeff n1kdo

Hank,

A few additional comments that may be of interest to others:

Digi-Key was the only major vendor that could supply the parts at this time.

I was able to stick with 105C on all electrolytics, but some were 1000 hour rated life. I would normally buy longer life caps if I had that option. This may be related to ongoing supply chain problems. Longer life caps were listed as out of stock.

The NTE 33 uf caps 6.5 x 6.5 mm fit OK in place of the slightly smaller original parts. The four of those are on the small vertical boards. Those were ordered through Digi-Key Marketplace and shipped directly from NTE Electronics. There was an additional steep shipping charge and it took an extra week.

Only one part required special mounting. The body of the axial tantalum cap was longer than the hole spacing on the board. I mounted it at a 45 degree angle with the positive end up, which resulted in about half inch lead length on that end.

I will do my best to answer any questions.

Paul

Paul,

Great news.? I thought there would be a happy ending to this story.

I've been meaning to ask you what has been going on with the few replacement caps you'd said weren't too good of a fit.

Thanks for the update which is very important to do.? Although quite complete, I do have some questions for you.? Maybe later today or tomorrow for those.

Hank

Just posting an update for the archives...

I replaced 29 capacitors on the power supply board: 28 electrolytic and one tantalum. Also replaced primary and secondary memory backup batteries as preventive maintenance.

On the first boot it came up error code 0040 on the self diagnostics. The service manual says that is Real-Time Clock failure, Memory board. The error appeared just the one time. It has come up all self tests passed on subsequent boots. I will assume that was just some clock confusion due to replacing the memory battery. I didn't even know it had a real time clock but will look into that.

All functions seem to be working normally.

73,
Paul N1BUG

Hank and group,

I am back to it after a day and a half digging out from the blizzard.

I've spent the last several hours working on this. After correcting a couple of mistakes on lead spacing I am stuck on just a couple of caps now.

The infamous 10uF tantalum cap. I cannot read the voltage on the original part but since it is in a 40V circuit I assume 50V. Original is axial, approx 4.5mm D x 9mm L. Closest I can find is 13.46mm L. This will not fit without standing on end with one long lead or (maybe) folding both leads under the part.

My only choice for 33uF 25V, approx 5mm D x 5mm H, 2.5 LS seems to be a slightly larger NTE part offered through Digi-Key Marketplace.

I don't know if lead lengths are critical in this power supply but past experience with switching circuits at 137 kHz tells me it can matter at low frequencies. I am trying to stick with parts that fit as the originals.

This power supply board has had 3 FETs replaced before. Whoever was in here last failed to put thermal paste between the L shaped heat sinks the devices are mounted to and the heat sink plate that mounts to them. Pictures of other E8285As I have seen had thermal paste there.

Paul N1BUG

开云体育

Hello, .

Sunday, January 30, 2022, 0:31:03, Sgc1823 wrote:

I've got an 8921 (analog cellular and lmr) that had similar issues. I noticed it began to take longer to start. Push the switch and nothing happens for a while, then it started up and ran fine. It got so it would take 5 min or so before anything happened. ?I pulled the supply out and replaced all the electrolytics. ?That was my issue.

If you ordered caps from Digikey or some other major distributor - can you share part numbers you ordered from order history? It would be great help to anyone who try to find a set of right caps for their dying 892x.

Regards,
Yuri.


--
Best regards,
Yuri ? ? ? ? ? ? ? ? ? ? ? ? ?mailto:yuri@...

Thanks. If you only fixed 2 of 5 with caps that doesn't sound very promising. I assume the +12V STBY LED should at least be on? Mine isn't, so I'm guessing that means at least one problem with the power supply. I have lost much of my bench capability without this instrument but it's starting to look like I should put it back together and forget it until we have fewer supply issues at the very least. I am not finding replacements for some of these caps that will fit in the available space and have correct lead spacing.

Paul N1BUG

I have read in the group archive some recommendations to replace all the caps. I have already pulled the supply and have the main board here on the bench. I have been able to read the value on all but one cap (oriented such that I can't see it) and am trying to find replacements that have the correct lead spacing and are small enough to fit. With ongoing supply chain problems, now is a bad time to be doing this.

Paul N1BUG

Be aware the fan rpm sensors are used in E8285A. If a fault exists in fan/s, the PSU will not turn on.
Only 2 out of 5 E8285A power supplies I have repaired worked by simple recapping. I never managed to get 2 working. The other had issues with the HV rail - think its 40V? Cant recall - though I do remember the voltage is required for PIN diodes in various PLLs and often the cause of OOL LEDS coming on. Anyway I replaced all parts in the HV supply to get it going. Most modern caps are smaller - make sure you use 105Deg C caps.

I've got an 8921 (analog cellular and lmr) that had similar issues. I noticed it began to take longer to start. Push the switch and nothing happens for a while, then it started up and ran fine. It got so it would take 5 min or so before anything happened.? I pulled the supply out and replaced all the electrolytics.? That was my issue.

I then measured the esr on all the caps and only one was bad, a 33u 50V. As I recall it was on or at least near the small vertical board. If you go to all the effort to pull the supply, I'd replace all the caps. It's not too $$ or time consuming.