Turns out that it's the VFD in my 58503A that is bad, and not the
driver chip.

The good news is that I don't have to do any PCB rework. The bad news
is that I have to find a replacement VFD at a reasonable price.

Any ideas (besides eBay)? And if I do end up buying on eBay, does
anyone have any experience with replacement VFDs from China?

There's a philosophical point where you say:
1) these might be bad, I'll replace then all.
2) I'll only replace the ones that I know are bad.
To me, it depends on how difficult it would be to replace, and also what the probability is that all of these could be bad.
Your pick.? I don't think that there's an absolute "this is the right way".
Harvey
On 11/21/2021 6:06 PM, Paul Bicknell wrote:

Hi Tom

Thank you for looking up the part numbers? please look at the xl document I have put onto the HP 70900 folder in the HP 70000 on this site

as I am using it to compare the values of the capacitors measured in and out of CCT on a number of 70900 units

hopefully this might help speed up repairing 70900 in the future

Vladan I have put a number of Vishay spc sheets in the same file along with a list of their suitability? size and availability

Regards Paul

*From:*[email protected] [mailto:[email protected]] *On Behalf Of *tmillermdems
*Sent:* 21 November 2021 22:41
*To:* [email protected]
*Subject:* Re: [HP-Agilent-Keysight-equipment] I'm old 70+, the 70K system is old; good or bad omen?

From the 70900B CLIP:

C32 0180-3068? 470u/12v

C17, C18? 0180-3441? 270u/25v

C52 0180-3443? 560u/7.5v

C43 0180-3068? 470u/12v

C42? 0180-3442? 1200u/12v

AL-ELCTLT

Regards,

Tom

On 11/21/2021 4:34 PM, pianovt via groups.io wrote:

Paul,

If you (or someone else) can post the HP part numbers for the
capacitors, there is a chance that I may be able to find out what
ESR specs were used when the module was designed.

Vladan

------------------------------------------------------------------------

Avast logo <>



This email has been checked for viruses by Avast antivirus software.
www.avast.com <>

Hi Tom

?

Thank you for looking up the part numbers? please look at the xl document I have put onto the HP 70900 folder in the HP 70000 on this site

as I am using it to compare the values of the capacitors measured in and out of CCT on a number of 70900 units

hopefully this might help speed up repairing 70900 in the future

?

Vladan I have put a number of Vishay spc sheets in the same file along with a list of their suitability? size and availability

?

Regards Paul

?

From: [email protected] [mailto:[email protected]] On Behalf Of tmillermdems
Sent: 21 November 2021 22:41
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] I'm old 70+, the 70K system is old; good or bad omen?

?

From the 70900B CLIP:

C32 0180-3068? 470u/12v

C17, C18? 0180-3441? 270u/25v

C52 0180-3443? 560u/7.5v

C43 0180-3068? 470u/12v

C42? 0180-3442? 1200u/12v

AL-ELCTLT

Regards,

Tom

On 11/21/2021 4:34 PM, pianovt via groups.io wrote:

Paul,

If you (or someone else) can post the HP part numbers for the capacitors, there is a chance that I may be able to find out what ESR specs were used when the module was designed.

Vladan

?

---

This email has been checked for viruses by Avast antivirus software.

?

At least check the 70900A DIP switches on the top are set to something sensible, and check what the front LEDs display when you power (or attempt to power) it up.

?

I have this very instrument. The 70900A should show up in row 0 column 18, and if the HPIB functionality is enabled for this module it will also show up on HPIB address 18. The RF section should show up in row 6 column 18, (70904A in my instrument) the IF section (70902A) should be in row 1 column 18, the display should be in row 0 column 4. If there are no power supply issues (because I notice the "volt/temp" LED is not illuminated) I would also recommend pulling all the modules out and setting them to their default positions

Where are defaults found?

and making sure you have all the necessary SMB and semi-rigid cables routed correctly. CHECKED

Even with dead batteries the unit will still function, but will complain about the confidence test TEST PASSED OK

?

and the 70900A will not keep calibration coefficients after a power cycle.??

Best Regards,

Craig Petersen.

?

From the 70900B CLIP:

C32 0180-3068? 470u/12v

C17, C18? 0180-3441? 270u/25v

C52 0180-3443? 560u/7.5v

C43 0180-3068? 470u/12v

C42? 0180-3442? 1200u/12v

AL-ELCTLT

Regards,

Tom

Paul,

If you (or someone else) can post the HP part numbers for the capacitors, there is a chance that I may be able to find out what ESR specs were used when the module was designed.

Vladan

Hi again folks:

Looking further, I checked the band 1-3 amplifier, and it doesn't look good: the output must be +26 dBm and I only read +14.x, which is the output I measured in the the A16 modulator/splitter (whose outputs look fine). I tried to feed it with an external signal (2.3 to 2.56 gig from my 8663A) and has no gain, instead, it has a 1 db of loss. So the high band amplifier is a thru.
Also, as I said, I am pretty sure that the band 0 amp is dead, I feed it with -10 and +20 (and ground) and looks shorted: drawn almost 1 amp and getting hot quickly. And has 23 dB of loss, instead of 40 dB of gain. Maybe the resistor divider made with discrete component has a shorted one, but no one looks darker or suspicious...

The diagnostic for now is the two amps are failing, allowing me to see some activity on the output of the unit at high bands, but weaked as hell because of insertion loss of the SYTM and other microcircuits at the very end of the signal path. Spare parts are present on eBay but in the States, so shipment and customs multiplies by two the price.

The A8 3.7 GHz phase-locked oscillator reads arround 0 dBm, so looking good.

Best regards,
R.

Thanks Vladan,

That’s a good postulate because I had to replace one of the fuses on
the PS board to get this far in trouble shooting. BTW those fuses are
rare and damned expensive. They are from the early NASA days and
listed as very fast acting fuses. Got a few on order but not wanting
p-off $10 dollar fuses in trouble shooting I improvised by breaking
open standard fast 2A fuses reclaiming the fuse wire. I pinned the
wire in the 2 pins of the socket using the blown fuse. I extended the
wire free of the socket by 3mm to free it from any thermal sink. If I
was going to blow fuses they would be cheap fuses. Before refusing I
could not get all the keys on the 702005A to respond, i.e. row column
selection on A-Map. Also none of the front status LEDs on the L.O.
were illuminated. Best guess was with that fuse out, inboard fuse,
some mutual bus was being inhibited. Refusing cleared 70205A
functionality and LEDs were lit, albeit maybe dimmer than the RF
section. Of note, now the IF LEDs were out where prior to fuse replacement at least one, I think both, were lit.

I think next step is to pull the PS board and…

1. Visual on full board top and bottom 2. Run C and ESR readings on
the electrolytics 3. Verify what the bad fuse fused then verify all
actives in that path

I still don’t know if the 70905A/L.O. should be seen in the Address
Map

*From:* [email protected]
<[email protected]> *On Behalf Of *pianovt via
groups.io
*Sent:* Saturday, November 20, 2021 5:36 PM
*To:* [email protected]
*Subject:* Re: [HP-Agilent-Keysight-equipment] I'm old 70+, the 70K
system is old; good or bad omen?

Gene,

I don't know the answers to your configuration question, but one
possibility is that a power supply in the LO module is down. After
years of service, the electrolytic capacitors in that module tend to
fail. You may be able to see what's going on by removing the mainframe
cover and the LO module cover.

Vladan

Hello Bruce

I found the first (and hope the only) culprit of this issue: the band 0 power amp. Tried inside the 8341A and outside, feeding it with my lab power supply and sig gen and no amplification at all. The service manual talks about 40 dB of amplification, but a few mV measured with scope is what I found, with correct frecuency (100 Mhz). The output of the mixer is about -17 dBm, I think this is normal, because manual says 9 dB atten thru mixer.
The amp is clearly dead, I search for info in the manual, but I don't think I can do much to fix it, so I'll buy spare from eBay.

Regards,
R.

Thanks Vladan,
That’s a good postulate because I had to replace one of the fuses on the PS board to get this far in trouble shooting. BTW those fuses are rare and damned expensive. They are from the early NASA days and listed as very fast acting fuses. Got a few on order but not wanting p-off $10 dollar fuses in trouble shooting I improvised by breaking open standard fast 2A fuses reclaiming the fuse wire. I pinned the wire in the 2 pins of the socket using the blown fuse. I extended the wire free of the socket by 3mm to free it from any thermal sink. ?If I was going to blow fuses they would be cheap fuses. Before refusing I could not get all the keys on the 702005A to respond, i.e. row column selection on A-Map. Also none of the front status LEDs on the L.O. were illuminated. Best guess was with that fuse out, inboard fuse, some mutual bus was being inhibited. Refusing cleared 70205A functionality and LEDs were lit, albeit maybe dimmer than the RF section. Of note, now the IF LEDs were out where prior to fuse replacement at least one, I think both, were lit.
I think next step is to pull the PS board and…
1. Visual on full board top and bottom
2. Run C and ESR readings on the electrolytics
3. Verify what the bad fuse fused then verify all actives in that path
I still don’t know if the 70905A/L.O. should be seen in the Address Map
*From:* [email protected] <[email protected]> *On Behalf Of *pianovt via groups.io
*Sent:* Saturday, November 20, 2021 5:36 PM
*To:* [email protected]
*Subject:* Re: [HP-Agilent-Keysight-equipment] I'm old 70+, the 70K system is old; good or bad omen?
Gene,
I don't know the answers to your configuration question, but one possibility is that a power supply in the LO module is down. After years of service, the electrolytic capacitors in that module tend to fail. You may be able to see what's going on by removing the mainframe cover and the LO module cover.
Vladan