|
|
Re: How to remove the output cable of the isolator in the 8672A
Had another good look.
Removing the side panels to get access to the output SMA of the isolator is a lot of work,
For just checking the input level of the YTM it seems simpler to disconnect the YTM as these SMA's are accessible from the top.
Is there a risk in disconnecting the YTM? Breaking pins of the DIP connector? Breaking YTM SMA connectors?
Is there any risk in running the 8672A with disconnected YTM?
|
Thanks to you for your attention!
This activity gives me an unreasonable joy. Yes, I spent my life in electronics, but in the last 15 years I was more a manager than a technician, and however the worries of a company on your shoulders much ¡°dilute¡± the pleasure.
Now that I am retired, restoring and documenting this forgot equipment, I found again the enthusiasm of when I was fifteen, amplified by the fact that what I do is absolutely useless¡ ?(this is a fundamental component of pleasure¡)
????? Thanks!
|
Re: Any takers for semi-working 85060C?
Hi!
If Peter doesn't take it, I'm interested in taking it.
Best regards, Denys
On Sun, Jul 14, 2024 at 9:12?PM Peter Gottlieb via <hpnpilot= [email protected]> wrote: I'd be interested, I'm in Natick.
Peter
On 7/14/2024 8:37 PM, bagojfalvibagoj via wrote:
>
> This is mostly to calm my conscience. Does any of you want a repair project
> 8506C ECal controller? I know it is basically trash, but considering the two
> pieces available on ebay now are $1k, this might be something someone want in
> lieu of a better alternative.
> I have already replaced the floppy drive with USB and the 24V PSU in it but
> didn't help.
> During longer cals it just hangs mid process. It is essentially a 486 PC and
> I'm guessing a cap is giving out or alike but I don't have the drive to fix it.
>
> *Free *if you want to pick it up in the Boston area, or if you're willing to
> pay for shipping+packaging service. (I'm willing to take it to the UPS Store,
> but that's about the amount of effort I have for it.)
>
> No matter how broken I hate throwing away once expensive equipment.
>
>
|
Re: VNA calibration kit to 50 GHz
One could borrow from ultra high vacuum technology.
It would be possible to fabricate copper gaskets exactly matching the waveguide size, and either provide a bevel on the gasket to concentrate force at the inner edge or make flanges that pinch gaskets similar to the way a conflat flange functions.
Neither are exactly the sort of thing that¡¯s easy to fabricate and it¡¯s easy enough to damage surfaces. Unlike UHV use though a hermetic seal isn¡¯t necessary to be better than conventional methods.
All things being equal, beveled copper gaskets would be cheaper but less durable. They could be surface ground and then lapped on a fixture. The conflat type flange would be somewhat more damage resistant being recessed but harder to fabricate.
|
Re: VNA calibration kit to 50 GHz
Of course, its this simple for coax, a short/ an open and a termination in the characteristic impedance of the TEM line being used.
We obtained excellent results in the infancy of the ANA art.
Of course waveguide is not easy: The impedance is changing over frequency, the phase velocity is changing, and there is no "open" as the waveguide radiates.
But wasnt the discussion about how a person could come up with a waveguide cal kit? Not a basic primer on Network Analysis coax calibration basics? And didnt the guy want to do relatively narrowband stuff around 10 GHz. Of course, wideband cal is difficult using ANY transmission medium, coax, waveguide, slabline fiber optics, etc.
I didnt say to use "inappropriate" lengths. Instead I said you could make ANY length offset short and should experiment, after some calculation of course to get "close", then build a few of various lengths. Maybe I should have used more simple English. Some people strive to misunderstand apparently. JK
?
An S11 calibration requires 3 standards. But they can't be random lengths. Ideally the phase difference between two standards should be as close to 180 degrees as possible, as that gives you best stability. In coax, that's relatively easy to achieve. Ideally you want the offset length of the short a little longer than the open, but equal lengths is quite okay. But in waveguide it is not so easy to achieve. The change in wavelength vs frequency means that even with optimal shorts, the phase difference between the shorts becomes around 30 degrees at the ends of the frequency range of the waveguide. If the shorts are of inappropriate length, the phase difference between them will become 0 at some point, and the calibration impossible, as then you are trying to solve 3 simultaneous equations with 2 unknowns. When the phase difference is close to 0, then theoretically calibration is possible, but it is not stable.
|
Re: Any takers for semi-working 85060C?
I'd be interested, I'm in Natick.
Peter
toggle quoted message
Show quoted text
On 7/14/2024 8:37 PM, bagojfalvibagoj via groups.io wrote:
This is mostly to calm my conscience. Does any of you want a repair project 8506C ECal controller? I know it is basically trash, but considering the two pieces available on ebay now are $1k, this might be something someone want in lieu of a better alternative.
I have already replaced the floppy drive with USB and the 24V PSU in it but didn't help.
During longer cals it just hangs mid process. It is essentially a 486 PC and I'm guessing a cap is giving out or alike but I don't have the drive to fix it.
*Free *if you want to pick it up in the Boston area, or if you're willing to pay for shipping+packaging service. (I'm willing to take it to the UPS Store, but that's about the amount of effort I have for it.)
No matter how broken I hate throwing away once expensive equipment.
|
Any takers for semi-working 85060C?
This is mostly to calm my conscience. Does any of you want a repair project 8506C ECal controller? I know it is basically trash, but considering the two pieces available on ebay now are $1k, this might be something someone want in lieu of a better alternative.
I have already replaced the floppy drive with USB and the 24V PSU in it but didn't help.
During longer cals it just hangs mid process. It is essentially a 486 PC and I'm guessing a cap is giving out or alike but I don't have the drive to fix it.
Free if you want to pick it up in the Boston area, or if you're willing to pay for shipping+packaging service. (I'm willing to take it to the UPS Store, but that's about the amount of effort I have for it.)
?
No matter how broken I hate throwing away once expensive equipment.
|
The white ¡°knobs¡± on the top of the flanges for the circuit cards are specifically for pulling the cards from the motherboard/card cage. Some of the cards have gold plated side edges that mate to brass or gold plated tension plates in the card cage - all done to reduce RFI.
Occasionally you will find spurious RF signals on the display that relate to frequencies found in the 85660A/B RF section. The first thing to do is check all of the screws in the A6 section for tightness while watching the displayed spurious signal of interest. If not solved then closely inspect the card cage for a whitish gray corrosion around any of the A6 assemblies, if found remove that assembly and clean both the card flange and card cage mating surfaces with an eraser until shiny, that should solve the spurious emission. Very rarely you may find corrosion on the bottom of the card cage where it mates with the motherboard. If that happens, then remove all the cards in the cage remove the cage, clean the mating surfaces as above and reassemble.
You should be able to isolate the specific assembly radiating the spurious signal based on the signal frequency and the RF section/A6 theory of operation and troubleshooting.
Don Bitters
|
Hi All,
I have a pair of 8702 Lightwave component analsyers, does anyone know if there is a downloadable copy of the service manual available?
Can they be modified for an external monitor?
Thanks in advance
Wayne eckert
VK2WDE
|
Re: VNA calibration kit to 50 GHz
OK, with further thought, if you did do some software mods for a waveguide cal system, you wouldn't actually need to make the the precision offset short, you just need to know how long they actually are and use that mechanical distance in the software to calculate from. Really anything close would work, because, as Ed (Hi Ed!) pointed out, really only 1/8 wave at one frequency, you just have to know THAT freq!.
An S11 calibration requires 3 standards. But they can't be random lengths. Ideally the phase difference between two standards should be as close to 180 degrees as possible, as that gives you best stability. In coax, that's relatively easy to achieve. Ideally you want the offset length of the short a little longer than the open, but equal lengths is quite okay. But in waveguide it is not so easy to achieve. The change in wavelength vs frequency means that even with optimal shorts, the phase difference between the shorts becomes around 30 degrees at the ends of the frequency range of the waveguide. If the shorts are of inappropriate length, the phase difference between them will become 0 at some point, and the calibration impossible, as then you are trying to solve 3 simultaneous equations with 2 unknowns. When the phase difference is close to 0, then theoretically calibration is possible, but it is not stable.
|
|
|
Jinxie,
Yes, they can be used to wiggle the modules loose and pull them from their sockets. The same things are used in other instruments such as the 8510C. Do be gentle as they may crack from old age!
toggle quoted message
Show quoted text
On Jul 14, 2024, at 4:34?PM, Jinxie via groups.io <paul666@...> wrote:
?HI all, I would like to pull out one or more of the A10 modules in my 8566B for examination and testing. These are the ones indicated by the arrow in the attached photo. HP have incorporated white plastic bell-shaped stand-offs which I assume are there to make the base of the case more rigid, but can they also be used to pull the modules out of their sockets or is some special tool required? There's no obvious other way to get a grip on them. They've thoughtfully provided 'ring-pulls' for the boards in the A19-A21 slots so it would be odd if they hadn't done something similar for all the A10 ones. Many thanks, J.
<ASSY Block.jpg>
|
HI all,
I would like to pull out one or more of the A10 modules in my 8566B for examination and testing. These are the ones indicated by the arrow in the attached photo. HP have incorporated white plastic bell-shaped stand-offs which I assume are there to make the base of the case more rigid, but can they also be used to pull the modules out of their sockets or is some special tool required? There's no obvious other way to get a grip on them. They've thoughtfully provided 'ring-pulls' for the boards in the A19-A21 slots so it would be odd if they hadn't done something similar for all the A10 ones.
Many thanks,
J.
|
Re: HP 8566B - Can my high voltage transformer be faulty?
Well I found the box of parts. I think they are most of a CRT retrofit kit for the 85662A, as the assemblies in the box mostly match those listed for the S/N 3004A and up variants. In these models, the HV multiplier is integral to the A1A3 HV board. Sounds like this is not what is needed for saevar's unit.
Dave Casey
Hi Don,
Wouldn't a shrunken display usually indicate that the HV was
too high, speeding up the electrons, and reducing how much the
electrostatic deflection system can deflect them?
That's how it works on conventional scopes...
-Chuck Harris
On Sat, 13 Jul 2024 00:31:43 -0500 "Don Bitters via "
<donbitters=[email protected]> wrote:
> I have replaced several in the 85662A many, many years ago. There are
> 2 types depending whether you have a long CRT (PN 5083-5791, IIRC),
> or a shorter CRT newer type dot raster scan CRT (Panasonic). The HV
> transformer is inside the metal box at the rear, left of the CRT
> (A1A6 circuit).? They usually fail open.? You should see a shrunk
> image on the CRT when the HV transformer fails, both horizontally and
> vertically, or a dim dot. You will still have a good display if you
> check the video and horizontal outputs on an oscilloscope.
>
> Don Bitters
>
>
>
>
>
|
Enormously excellent pieces of work. Thanks for making them available to all.
WT
|
Re: VNA calibration kit to 50 GHz
True, ?¦Ë0 (free space) is ~ 29.118 mm, but in WR90, ?¦ËG (wavelength in the waveguide) is something like 37.722 mm, using ??¦ËG = ?¦Ë0/sqrt(1 - ?(¦Ë0/ ¦ËC)**2), where ?¦ËC is the cut-off wavelength = 2a, where a = the longer internal waveguide dimension.
|
FS (US/CAN) : HMC365S8 RF prescaler (replacement for Agilent 1GC1-4207)
Cross-posted to EEVblog forums too, https://www.eevblog.com/forum/buysellwanted/fs-(uscan)-hmc365s8-rf-prescaler-(replacement-for-agilent-1gc1-4207)/
***
This is an equivalent of infamous 1GC1-4207 , used in early-mid-2000's HP/Agilent ESA spectrum analyzers (E44xx etc) and ESG signal generators, as part of the PLL chain.
Qty 2, in unopened digi-key bag . These sell for 35$ each. Would like to get 50$ (CAD) for both, buyer pays shipping. US/CAN/Europe.
Selling because my unit has a later revision RF deck with a different prescaler MMIC.
Chris
|
Re: Made one out of workshop scrap to enable the correct tightening of SMA 's
Shameless plug: I make these and sell them on eBay "SMA wrench" , eBay handle "kmec-at-aol".
Doesn't make a lot of money (some time involved), AND you can do it yourself, but I have some fun and enjoy metalwork. (You have to, if you like making microwave stuff). I also have gotten hollow shaft 5/16 Xcelite nutdrivers (yellow handle), and ground away 1/2 of the nut&barrel, makes a MOST handy tool for tightening sma connections, especially on sma relay (hard to get to the center connector with hardlines in the way).
J.Kruth
In a message dated 7/12/2024 2:42:30 PM Eastern Standard Time, james.ford@... writes:
?
Regarding 5/16 inch wrenches with the slot to slip over coax, Mini-Circuits part number HT-4-SMA is about $37.? Nice knurled handle for good grip, too.? Found one in the lab this morning.?
|
Re: VNA calibration kit to 50 GHz
OK, I'll stop after this, I promise!
?
I said " So to that end, short sections of flanged waveguide with shorting plates would work."
?
Lets make it simpler: I am sure we either own (I do) or have a friend that bought one of those $200 simple CNC milling machines (I know, not great but not bad either). No sweat to put a piece of 1/8 inch thick brass plate in one and mill out some WR-90 endcaps. And mill a little grove beside them to make it easy to cut out with a manual hacksaw or bandsaw.
Then scrounge some nice flanged sections of WR-90 (can cut off some junk) and soft or hard solder the end caps on (after you clean up the cut on the guide of course).
Soft solder was used by WE (Western Electric, sorry, didnt spell out in last message) for the Bell System Long Lines hardware and it works fine but doesnt silver plate as well as silver solder (of course).
Now you can make as many custom length shorts as you want for TRL type cal setups. Hell , you might even be able to copy the length to match a set of cal constants you might find tabulated somewhere!
Thought of this while shaving just now (great time to think!).
Jeff Kruth
|
Jinxie
Dr. David Kirkby, Kirkby Microwave Ltd
fenugrec