¿ªÔÆÌåÓý

One (1) is closed/on

Since I can never tell what is 1 and what is 0 on these switches, here is a picture of the "Normal" setting for S400:



I can't find any information on what positions 1-7 do. I do remember they have something to do with how the instrument boots.

While you have the thing apart, check the battery on the processor board. It may be near the end of its life.

Vladan

I think you have option 089, or at least some of the required hardware. I see an R-channel switch in your picture. It's that gold plated microcircuit between the wide rainbow-colored ribbon cable and the lone step attenuator.



This is a great analyzer you got, the only problem I see with it is that someone tinkered with it and now you have to retrace what they did. I don't remember off-hand what all the other switch positions on the processor board do. I will look later and let you know.

What do you see on the screen when you first turn on the analyzer (make sure the analyzer is set to boot up in "Preset: Factory" and not "User")? Which options are listed on the screen? Does the serial number on the screen match the number on the rear panel label?

Vladan

Hi Vladan,

Thanks for the thoughts on the options. How can I know if option 89 is available (hardware wise)? Being able to do mixer measurements (easily) would be really nice. I have no problem using external bias-t's, it's what I'm used to at the university. I'll look more at the signal path diagrams for Option 85 vs 400. I guess they're mutually exclusive? I'm certainly looking a gift horse very far down the mouth at this point. Having a VNA like this is already a fantasy come alive.

I have yet to re-assemble the VNA (lots of dusting to do and maybe the fan replacements) so I'll make sure that everything is well seated before closing it up.

I did think it odd that the J400 position 8 being switched to off would have caused an issue, but I put it down to the VNA having gotten into a bad state in the past. Do you have more info about the switch settings? I think the all on and #8 off is the Alter state for doing service actions? Is there any reason to look at the other switch states or am I good as is?

Thanks,
Derek

Derek, to me those pictures look like it's an option 007 + option 085 configuration (high power measurements).This means no bias-Ts, and it will be a bit of work to convert to 4-sampler. This option also includes direct sampler access (opt. 012). You may or may not have option 89 (mixer / frq.offset measuremnts), I can't tell from the pictures.

The service switch setting should not stop the instrument from booting up. It's possible that you "fixed it" while moving things around, possily that pinched cable.

Vladan

On Tue, 2 Apr 2019 07:41:18 -0400, you wrote:

The number one tool of anyone fixing modern
electronics should be an ESR meter of the
Dick Smith variety.

Arguments about about ESR vs capacitive reactance
are numerous, but immaterial. When a capacitor
goes bad, its ESR will usually be 2 to 100 times
higher than normal.

Shorted capacitors will also be obvious, as the
ESR meter will show their ESR to be 1/4 (or less)
the normal value for a capacitor of that value.

Plugged in a Tektronix 7000 series counter (7D15 to be exact), on an
extender. It wasn't working. I had the room lights dim for whatever
reason, hopefully to see anything that was unhappy.

Everything was fine, but the scope wasn't happy. Even the little LED
on the main board of the counter was glowing a nice cheerful red.

Wait a minute, *what* LED?

I don't remember an LED on that board....

TUrned up the lights a bit, and saw one of the dipped tantalums
glowing a nice cheerful red.

Not quite the way you want to find that out. No smoke, no smell, just
rather gray and glowing.

Harvey

Capacitors are the least reliable part made today.

Also, use your eyes, and scan the field. If almost
all of the capacitors used are, say, Nichicon, and
you find one or two that are YooSun, or Samson, or
Jokcon, or something else weird, it means the supply
line for Nichicons of that value was blocked up for
some reason, and purchasing bought what they could
find.

-Chuck Harris

Michael Williams wrote:

Found the problem!

C448 had failed short.

Many, many thanks to all the guys who contributed to this thread ¨C especially Mark and AmirB. You guys pulled my head out of the rabbit hole with that MUX nonsense. I¡¯m so glad it was something as simple and easy to fix as a bad cap. I had fears of bad ASICs, and other such unobtanium floating through my head.

If anyone is interested in my fault-finding approach this evening, read on. Otherwise, please accept my thanks ¨C this group has been very helpful to me.

My probing the other day had shown no signal at R525, and when I was unable to see anything resembling an 800MHz source signal (due to bad probing) I started checking every voltage I could find, which led me down the rabbit hole with 2ns_VCO_Vcc and the MUX nonsense.

After listening to Mark B and Amirb (and the many others who contributed to this thread) I had a more organized plan of attack for tonight. After double-checking the supply voltages, I started working from the other end of the 50MHz signal path (the 1/16 divider). I confirmed the presence of the 800 MHz 2nd VCO by making an ad-hoc RF probe with a loop of wire connected to my SDR Play, I then walked all the way through the 50Mhz path. After seeing signal on one side of R523, and no signal on the other side, I began to suspect C448. A quick power-down and check with a multimer confirmed my suspicion. I may not know one end of a MUX from another, but I do know that capacitors should not act like low-value resistors. I pulled the cap, and the analyzer started to behave itself. Little did I realize I was so close to the problem when I probed R525 the other day!

I have a replacement cap on order. I assume it¡¯s purpose is part of a filter along with L55 and R523 to clean up the 50MHz signal. The analyzer would probably function reasonably well without it, but I may as well replace it while I have the RF module open.

Cheers,

Mike Williams
AG5QX

Hello,

I've just acquired an HP 8720D which was "working until one day it stopped". I tried turning it on and the machine started continuously rebooting every ~6 seconds. Nothing is ever displayed to the LCD or out via VGA.

[EDIT: Writing this email was good in forcing me to be methodical, I've made serious progress. It's actually fully running now and I'll be getting some test devices to do an approximate calibration. This post is this more to document my findings and hopefully be of use to someone in the future.]

I've had a quick look through and can confirm that the A15 Power Supply pre-regulator is in working order and that the A8 post-regulator is also producing all the correct voltages. When A8 is disconnected from the rest of the system, but still powered, the rebooting does not occur.

The next step in troubleshooting is supposed to be identifying what assembly is causing the power fault, but I was only able to check the A14 Frac-N assembly. The others I was unable to find wiring labels to know what to unplug. I'll take more of an investigative look over the next days, but wanted to ask if anyone knows where the W89 cable is for the Testset Interface? (Page 188 in the service guide).?
https://doc.xdevs.com/doc/HP_Agilent_Keysight/HP%208719D,%2020D,%2022D%20Service%20guide.pdf

I then looked at the CPU LEDs and can't make much of them at the moment other than they aren't saying "A'OK". The service manual doesn't contain a lot of information about the codes. I believe (based on having 8 LEDS) that I have the newer 32 bit CPU with firmware 7.xx+, but don't know how to check that. I've made a video of the LEDS hoping that someone here can shed some light on their meaning. When I hold the Preset key the VNA does not power cycle, it stays minimally on.

I then removed the rear panel and found a squished, possibly cut!, ribbon cable (W86). I need to find some 0.1 inch headers to do connectivity tests, but maybe this is it! When the A7 CPU board is detached (W87 disconnected) the boot loop does not occur either.

I then noticed that position 8 of dip switch J400 on the A7 CPU board was in the off position, not what the manual calls "Normal" mode. So I set it back to ON and tried booting. It works! I'm able to boot into full operation and will try an approximate calibration very soon with what I have. Firmware 7.74 is installed which is the most recent I can find any reference to, so good there too!

Some photos of the VNA, cable, and switch are here:

Thanks to anyone who has comments or advice about this VNA. I'm pretty familiar with VNA operation and design, but each design has it's own peculiarities. I'm really hoping that this VNA can be brought back so I can use it for my personal projects outside of my school studies. The CPU fan is very rattly, I'll have to try and find a replacement though unfortunately it looks specialized. I'll probably also replace the main fan if it's available as these are the main moving parts (other than the 20 GHz switch!).

If anyone is familiar with the hardware options of the 8720D, could you please take a look at the photo album and let me know if you can spot any of them being installed? The VNA only lists option 010 during boot, but I'm reasonably certain that the attenuators that are installed are options.

Many thanks,
Derek

MW0LNA, hi to the "Amateurs" on the list!

More about this common type of failure:



I had one go off in a HP3457A. Took out 16 A line breaker, created a jet of hot tar out of the back of the unit in the process. Scary. The filter is still sitting on my terrace because I don¡¯t want to touch it again :-). Replaced with new item in all my 3457As, FN9222-3-06; do NOT go for NOS.

The rivets with which these are mounted are a major pain; I finally gnawed away the outside plastic of the old filter and snapped off the now protruding rivets using a ¡°Rabitzzange¡± (pincers, but not your garden variety ones).

Gr¨¹?e, Carsten

I would suggest *not* replacing with the original.

Schafner filters are fine in the USA, but double the mains
voltage, and they are stink bombs.

I would not use a used or old filter as a replacement. You
can get new in the same configuration.

-Chuck Harris

NigelP wrote:

¿ªÔÆÌåÓý

Yes, totally normal behaviour! Problem is the filter caps are pre the power-switch/voltage selector and they just give up eventually.

A lot of my equipment was bought in the USA so those caps have spent most of their life feeding off 115VAC and the shock of 240VAC here seems to be too much for them and they lose their presence of mind - often within a few minutes of power on - so I have taken to treating the Shaffner filters the same as Rifa filter caps and just changing them anyway. FEC or RS usually have them in stock.

All of this trouble due to a shorted 390 pF capacitor!

¿ªÔÆÌåÓý

Had a disaster with the mains socket/filter/voltage selector on my 8657B yesterday. Wasn't even switched on and smoke suddenly poured out of the back; I initially assumed it was an electrolytic but quickly discovered that the filter components in the power socket complex had gone S/C L-N and taken out a 5A fuse in the UK mains plug (thank goodness we have fuses in the plugs in the UK). The unit was off but connected to the live bench system and of course the filter is directly connected 100% of the time that power is applied to the end of the mains lead.

So, I can buy another socket here?in the UK but just in case anyone has a box of these laying around looking for a home :-), I 'm looking for a Shaffner FN370-2 Filtered Power Entry module...... can anybody (in the UK) oblige, preferably NOS rather than a parted out device with questionable lifetime??

Hi folks,

Thanks for all your tips. I have just been in that counter. My main concern was that there was something bad going on causing that humming, like shorted diodes or caps. But there isn't, all voltages are perfectly in spec. It's really just the transformer being noisy.
Modding the existing power switch would need some serious rework on that nice gold plated PCB. I don't want to do that, since the unit is in perfect shape outside and inside and I want to keep it that way.
So an external mains power switch is the way to go here....

Cheers,
Wolfgang

Caesar

Is there any SW you can recommend that will work on my machine? My very old copy of Acronis Enterprise?didn't seem to work. I have used this SW many times on XP without an issue but it was always PATA to PATA. My only option now is to use the USB-PATA adapter I have.

¿ªÔÆÌåÓý

Good job Mike! ?

Excellent! Good work Michael.