Sorry to hear that. Any chance it can be glued? It is good to know there is a part number on the switch itself. Best of luck in finding a replacement. There must be other equipment out there that used the same switch. Now the detective work begins!
toggle quoted message
Show quoted text
On Apr 28, 2020, at 10:58 PM, Michael Yellin <michaelhq54@...> wrote:
? Hello,
Not so well.? During the surgery, i did something that caused a plastic piece to snap off, rendering the switch inoperative.? No worries though, it has a part number, an ALPS SDS3P.? I don't think its?made anymore, but there appear to be some alternatives.
Thanks, Michael Go for it, and let us know how it turns out! ? Was considering trying to get it open and see.? It's no good, so I won't lose anything.? Speaks a thousand words! Hopefully someone’s got one to spare. Or, is it possible to open up the defective one and see if it’s repairable? I’ve had to do that on unobtainium switches a time or two.? ? Hopefully this will help.
?? On Mon, Apr 27, 2020 at 4:39 PM Paul Bicknell < paul@...> wrote:
Hi Michael
?
A picture specks a thousand words
?
?
You said you wanted a
power button, not a cable assembly with switch. According to Keysight the 5062
number you quoted is a cable assembly - line switch, not a cable assembly with
line switch. Switch part numbers start with 3101. It may be buried deeper in
the parts list.?
?
According to the service manual, it is a?CABLE ASSY WITH LINE
SWITCH, A1 TO A16J9
?
That’s a cable
assembly...
?
From what I can gather, the part number is?5062-4809.
?
Michael,
A part number would help
those of us who don’t have that particular model of spectrum ?analyzer
?but might have some parts.?
?Hello,
I am looking for the power button for HP 859x and 856x series of Spectrum
Analyzers.??The switch may also be used in other HP
units, just not sure.? Let me know what you have, and the cost.
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Internal Virus Database is out of date.
|
Re: HP8562A Errors 304,335,317,333
Yes! Been there done that, I had that experience having owned the SA just a couple of weeks. Replaced Resistor, and all rather large and 1 blown RIFA caps. ? Did allot of research and found these to be best.. Still manufactured by RIFA though. See how long they last ? Any suggestions how to load this board up before going all out into the SA.. Feel allot better knowing the +5volts didn’t become +xVolts across all relative boards.. ? Original PSU is 08562-60094 ? Benchmark PSU ?Below has no noticeable markings or Board ID but similar! ? Below: 
? ? 
? Anyone recognise it or have a CLIP schematic so I can do a like for Like comparison..? ? Presume a large high wattage Resistor say 47ohm..? Stu. ?
toggle quoted message
Show quoted text
From: [email protected] <[email protected]> On Behalf Of Tom Gardner
Sent: 29 April 2020 10:41
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8562A Errors 304,335,317,333? A suggestion: before finally replacing the PSU, consider replacing the "delayed action smoke generator" RIFA capacitors.
I've had one blow in a Tek 2465. The smoke cloud came from the series resistor which had detonated, removing a 1cm diameter patch of the PCB prepreg. I don't want to repeat that, thank you.
I'm just about to start selectively recapping my 8562b, partly because the RIFAs are already cracked, and partly because there is a -70dBc spur at +-31/63kHz.
Before going into the breach, I'm going to experiment with several techniques, including preheating the board, Chip-quik low temperature bismuth/lead solder, and a cheap and nasty solder-sucker-cum-iron, controlling its temperature with a variac.
On 29/04/20 06:59, stur_mm via groups.io wrote: Peter, ? Did have that thought, but not sure! With the outputs unregulated to a degree not sure connecting the board up now may kill the SA. Thoughts anyone? ? Just wanted to be sure that the PSU was good to put back in the SA. ? Both boards Tested Offload, but both are different design. Next step check the schematics for the old board. ? Still Not sure what +5V FL Refers too (Floating?) ?
?
|
Re: 04285-61001 resistor (100 ohm) for calibration of 4285A (75 kHz to 30 MHz) precision LCR meter
Never desolder neither a resistor nor a short because, with the heat, the value specialy of the short becomes out of specs
I doubt it. The specification is a resistance of the short is less than 1 ohm. For the open, the specification is a resistance of > 1 M ohm. The specifications are very wide.?
When it comes to 04285-61001 resistor, Keysight can’t calibrate it. What they offered to do was to determine if it met the specifications of the 100 ohm resistor in their megga expensive resistance standard kit. They have no idea whether it would meet that specification, but it was the best they?could do.?
I am reasonably convinced that the value of the resistor is not that important, which would explain why there’s no calibration procedure for it. I might make one up to 95 ohms and see what happens if I attempt a calibration.?
During the calibration stage of the 4285A, two things need to be done, in addition to putting opens and shorts on cables.?
1) The resistor is connected directly to the meter and measured?
2) The extension cables are added, the resistor put on the end of the cables, and the resistor measured again.?
What I suspect happens is that the meter applies a correction for the lead lengths, based on the fact that the resistor should measure the same whether on the meter or on the end of the leads. As such, I don’t think the value is critical.?
-- Dr. David Kirkby, Kirkby Microwave Ltd, drkirkby@...Telephone 01621-680100./ +44 1621 680100 Registered in England & Wales, company number 08914892. Registered office: Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom
|
Re: 04285-61001 resistor (100 ohm) for calibration of 4285A (75 kHz to 30 MHz) precision LCR meter
Never desolder neither a resistor nor a short because, with the heat, the value specialy of the short becomes out of specs On Wednesday, April 29, 2020, 10:49:52 AM GMT+2, Jean Red via groups.io <tarburner@...> wrote:
This is the? 04285-61001 teardown.
I can't further disassemble because of the welded BNCs.
|
Re: HP8562A Errors 304,335,317,333
A suggestion: before finally replacing
the PSU, consider replacing the "delayed action smoke generator"
RIFA capacitors.
I've had one blow in a Tek 2465. The smoke cloud came from the
series resistor which had detonated, removing a 1cm diameter patch
of the PCB prepreg. I don't want to repeat that, thank you.
I'm just about to start selectively recapping my 8562b, partly
because the RIFAs are already cracked, and partly because there is
a -70dBc spur at +-31/63kHz.
Before going into the breach, I'm going to experiment with several
techniques, including preheating the board, Chip-quik low
temperature bismuth/lead solder, and a cheap and nasty
solder-sucker-cum-iron, controlling its temperature with a variac.
On 29/04/20 06:59, stur_mm via groups.io wrote:
toggle quoted message
Show quoted text
Peter,
?
Did
have that thought, but not sure! With the outputs
unregulated to a degree not sure connecting the board up now
may kill the SA.
Thoughts
anyone?
?
Just
wanted to be sure that the PSU was good to put back in the
SA.
?
Both
boards Tested Offload, but both are different design. Next
step check the schematics for the old board.
?
Still
Not sure what +5V FL Refers too (Floating?)
?
|
Re: HP8562A Errors 304,335,317,333
Hi
Probably FL means Full Load.
George G6HIG On Wednesday, 29 April 2020, 07:00:06 BST, stur_mm via groups.io <stuart.ian.richardson@...> wrote:
Peter, ? Did have that thought, but not sure! With the outputs unregulated to a degree not sure connecting the board up now may kill the SA. Thoughts anyone? ? Just wanted to be sure that the PSU was good to put back in the SA. ? Both boards Tested Offload, but both are different design. Next step check the schematics for the old board. ? Still Not sure what +5V FL Refers too (Floating?) ? 
?
From: [email protected] <[email protected]> On Behalf Of Peter Hansen
Sent: 28 April 2020 22:34
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8562A Errors 304,335,317,333 ? Hello Stu did you measure both psu unconnected ?. I might be the regulator does not like no load ?
Peter et al, ? Plot thickens as testing all relevant test points shows that the PSU seems to be “Hunting” as the 28.0volt is stable all other voltages are swinging up and down as much as 3volts really not sure what is going on. 28v stable as sourced from a supply voltage of 40Volts ? ? Peter, ? Hi, manage to find the time to get stuck in to replacing them . Before I take the plunge just done a quick offline check of voltages and found that all OK including the 28.3Volt supply rail but TP308 shows a very wobbly 8-9volts DC, should be +5Volts. Tried My other PSU board and rock solid at +5.0volt. Not sure on the TP308 +5V FL what does FL mean ? ? 
? L301 is a Toroid so not sure why there would be a differential across it? ? Both TP307 and 308 show same volts not +5v though.(Is this not some simple filtering? Somethings a miss somewhere. Any Guidance would be appreciative. ? Stu. ? ? Hello Stu I use what I can get my hands on. That is often panasonic or nippon. but in fact all NON china will do ok. The reason why I said 220uF is because it is hard to find 180uF at least in my junk box ?. the caps you shoe is really bad. change them and you have your analyzer back ?
Peter, etal ? Decided they all were suspicious and with great care removed. Spot the suspect cap for 28.8volt supply? ? What’s the best manufacturer for Electrolytics, 180/220uF(Know you suggested 220uF) going to order some today from RS components.. ? That was the suspicious odour that came from the analyser I guess.. ? Remove them with great care, including the 12v 470uF cap, decision after seeing the darkened appearance and bulging bottom of the cap. No vias were removed at the same time just lots of Heat and Flux, braid and they came out easily. (Pictures may look like via’s still attached to legs, but it’s solder! ? Cheers, Stu. ? 
? ? ? ? ? ? ? Hi,
I used Chipquick to remove a broken ADC from my TDS540B acquisition board and from a donor board.
Worked great.
I added Chipquick to all the joints and heated the whole chip using hot air to remove it.
Afterwards I cleaned the soldering pads using a normal iron and flux.
Putting the donor part back was done using a normal, small flat soldering iron.
Pictures can be found here:
/g/TekScopes/album?id=21745
|
Re: 04285-61001 resistor (100 ohm) for calibration of 4285A (75 kHz to 30 MHz) precision LCR meter
This is the? 04285-61001 teardown.
I can't further disassemble because of the welded BNCs.
|
Wanted frequency counter (preferably HP, Agilent or Keysight)
I am looking for a low frequency reciprocal-counting frequency counter. Typically these are 225 or 350 MHz maximum, although some have an optional higher?frequency Input. Something like an Agilent 53132A would fit the bill, ?but I will consider something order or newer.?
I would prefer something in the EU or UK, but will consider ?purchasing from elsewhere.?
Please email me pictures, price and list of option(s) if any are fitted.?
Dave, G8WRB.?
-- Dr. David Kirkby, Kirkby Microwave Ltd, drkirkby@...Telephone 01621-680100./ +44 1621 680100 Registered in England & Wales, company number 08914892. Registered office: Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom
|
Re: N-Connectors at uW Frequencies
Remember as others have said.? There are 'N' connectors, and there are
"Precision-N" connectors.? Though they will inter-mate without*
"Damage", they have very different specifications above 12GHz, and
subtly different construction.? Most if not all of the "modern" (built
since 2000) HP RF kit with what look like 'N' type's are usually the
better precision types.
* Unless a non-captive plug with an out of spec' "proud" pin position,
is mated with a precision socket, that is virtually guaranteed to damage
the socket, precision type or otherwise!
Many instruments that go to 20GHz use the Precision-N types, for a reason.
Remember too, that there are also different types of SMA connector, for
much the same reasons.? Though, connect different types the wrong way,
and you may (likely) compromise the high spec' part for life.
Regards.
Dave B (G0WBX)
--
Created on and sent from a Unix like PC running and using free and open source software:
|
Re: HP8562A Errors 304,335,317,333
Peter, ? Did have that thought, but not sure! With the outputs unregulated to a degree not sure connecting the board up now may kill the SA. Thoughts anyone? ? Just wanted to be sure that the PSU was good to put back in the SA. ? Both boards Tested Offload, but both are different design. Next step check the schematics for the old board. ? Still Not sure what +5V FL Refers too (Floating?) ? 
?
toggle quoted message
Show quoted text
From: [email protected] <[email protected]> On Behalf Of Peter Hansen
Sent: 28 April 2020 22:34
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8562A Errors 304,335,317,333? Hello Stu did you measure both psu unconnected ?. I might be the regulator does not like no load ?
Peter et al, ? Plot thickens as testing all relevant test points shows that the PSU seems to be “Hunting” as the 28.0volt is stable all other voltages are swinging up and down as much as 3volts really not sure what is going on. 28v stable as sourced from a supply voltage of 40Volts ? ? Peter, ? Hi, manage to find the time to get stuck in to replacing them . Before I take the plunge just done a quick offline check of voltages and found that all OK including the 28.3Volt supply rail but TP308 shows a very wobbly 8-9volts DC, should be +5Volts. Tried My other PSU board and rock solid at +5.0volt. Not sure on the TP308 +5V FL what does FL mean ? ? 
? L301 is a Toroid so not sure why there would be a differential across it? ? Both TP307 and 308 show same volts not +5v though.(Is this not some simple filtering? Somethings a miss somewhere. Any Guidance would be appreciative. ? Stu. ? ? Hello Stu I use what I can get my hands on. That is often panasonic or nippon. but in fact all NON china will do ok. The reason why I said 220uF is because it is hard to find 180uF at least in my junk box ?. the caps you shoe is really bad. change them and you have your analyzer back ?
Peter, etal ? Decided they all were suspicious and with great care removed. Spot the suspect cap for 28.8volt supply? ? What’s the best manufacturer for Electrolytics, 180/220uF(Know you suggested 220uF) going to order some today from RS components.. ? That was the suspicious odour that came from the analyser I guess.. ? Remove them with great care, including the 12v 470uF cap, decision after seeing the darkened appearance and bulging bottom of the cap. No vias were removed at the same time just lots of Heat and Flux, braid and they came out easily. (Pictures may look like via’s still attached to legs, but it’s solder! ? Cheers, Stu. ? 
? ? ? ? ? ? ? Hi,
I used Chipquick to remove a broken ADC from my TDS540B acquisition board and from a donor board.
Worked great.
I added Chipquick to all the joints and heated the whole chip using hot air to remove it.
Afterwards I cleaned the soldering pads using a normal iron and flux.
Putting the donor part back was done using a normal, small flat soldering iron.
Pictures can be found here:
/g/TekScopes/album?id=21745
|
Hello,
Not so well.? During the surgery, i did something that caused a plastic piece to snap off, rendering the switch inoperative.? No worries though, it has a part number, an ALPS SDS3P.? I don't think its?made anymore, but there appear to be some alternatives.
Thanks, Michael
Go for it, and let us know how it turns out! ? Was considering trying to get it open and see.? It's no good, so I won't lose anything.? Speaks a thousand words! Hopefully someone’s got one to spare. Or, is it possible to open up the defective one and see if it’s repairable? I’ve had to do that on unobtainium switches a time or two.? ? Hopefully this will help.
?? On Mon, Apr 27, 2020 at 4:39 PM Paul Bicknell < paul@...> wrote:
Hi Michael
?
A picture specks a thousand words
?
?
You said you wanted a
power button, not a cable assembly with switch. According to Keysight the 5062
number you quoted is a cable assembly - line switch, not a cable assembly with
line switch. Switch part numbers start with 3101. It may be buried deeper in
the parts list.?
?
According to the service manual, it is a?CABLE ASSY WITH LINE
SWITCH, A1 TO A16J9
?
That’s a cable
assembly...
?
From what I can gather, the part number is?5062-4809.
?
Michael,
A part number would help
those of us who don’t have that particular model of spectrum ?analyzer
?but might have some parts.?
?Hello,
I am looking for the power button for HP 859x and 856x series of Spectrum
Analyzers.??The switch may also be used in other HP
units, just not sure.? Let me know what you have, and the cost.
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.
|
Re: N-Connectors at uW Frequencies
Well, the outside diameter is large, but the actually microwave part is fairly small, not more than a few mm.
I like type N connectors on test equipment because they are easy to use.? Panel-mount SMAs, or 3.5 mm, 2.92 mm, etc. are kind of a pain to get your fingers around.? Much easier to work with a type N and an adaptor.? Of course, if you really need 2.92 mm or smaller due to the high frequencies, type N is not going to work, and you have to live with the mechanical inconvenience.
Jim Ford
------ Original Message ------
Sent: 4/28/2020 6:55:54 PM
Subject: Re: [HP-Agilent-Keysight-equipment] N-Connectors at uW Frequencies
Thanks for posting
that....answers my concerns perfectly. Who would have guessed
that such a large connector would function fine up that high
in frequency.
Gedas, W8BYA EN70JT
Gallery at
Light travels faster than sound....
This is why some people appear bright until you hear them speak.
On 4/28/2020 5:26 PM, David Feldman via
groups.io wrote:
I ran into this question some time back, and found the following:
The posting has seemed to move around within the keysight web pages from time to time, but a search for "A Signal Analyzer Connector Puzzler" should get to the document.
It appears the trade-off is related to mechanical robustness of N connectors.
|
Does any one know how to successfully transfer a CalKit file from Windows 10 to 8720D via floppy?
I have created a custom Calkit file with VNA Cal Kit manager from the notes from the good Dr.
I can copy the file to a disk from Windows 10 and view it on the 8720D if disk is formatted in DOS format by the 8720D but it fails to load in the 7820D. It gives me some error.
If I format the disk in 8720D as LEF windows 10 is unable to write to the disk.
Next step if I cant transfer via disk is to hook GPIB up which will take seom fooling around no doubt.
Thanks
Dave
VE7HR?
|
Re: N-Connectors at uW Frequencies
Thanks for posting
that....answers my concerns perfectly. Who would have guessed
that such a large connector would function fine up that high
in frequency.
Gedas, W8BYA EN70JT
Gallery at
Light travels faster than sound....
This is why some people appear bright until you hear them speak.
On 4/28/2020 5:26 PM, David Feldman via
groups.io wrote:
toggle quoted message
Show quoted text
I ran into this question some time back, and found the following:
The posting has seemed to move around within the keysight web pages from time to time, but a search for "A Signal Analyzer Connector Puzzler" should get to the document.
It appears the trade-off is related to mechanical robustness of N connectors.
|
Re: N-Connectors at uW Frequencies
Hi
I think a sweep of 10 to 50 Ghz for a 3.5mm and a 2.92 compared to a 2.4 mm
would be moor useful to all and it would probably require 3 teats
10 to 20 ghz 22 to 40 ghz and 33 to 50 ghz
But who is equipped and up for it ?
suggest you use a 50 ghz power detector for all tests
verification of the test setup might be a problem
Paul
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected][mailto: [email protected]] On Behalf Of Reginald Beardsley via groups.io Sent: 29 April 2020 00:20 To: [email protected]Subject: Re: [HP-Agilent-Keysight-equipment] N-Connectors at uW Frequencies I'd like to suggest sweeping a 3.5 or 2.48 mm to N-F adapter up to 26 GHz or above with a known good cable on each side, zero padding the data out 500 GHz and doing a transform to time. That will give ps resolution of the reflections at the cable to adapter interfaces. You can get femtosecond resolution by adding more zeros. Though at some point your sampling in frequency and accuracy will enter the picture if you push it too far. If someone posts data I'll be happy to kibbitz on the FFT. In the testing I've been doing using a Tek 11801 & SD-24, I've been pleasantly surprised at how little difference there was between Weinschel APC-7/3.5mm F and Chinese versions. I got very lucky and picked up 4 of the Weinschels new for less per piece than the 2 Chinese I bought just to test. The Weinschels are better, especially the APC-7 side. But they actually differ between assemblies. I had similar experience with some unmarked, but obviously high quality SS, APC-7/N-F bulkhead adapters. The TDR plots are here: msg3012344/#msg3012344 Note that this only goes to about 12 GHz because of the Chinese RG402 jumpers I used. But it should give a good idea of what your transform to time should look like. At this point I'm beginning to be more concerned about cable losses than connector quality. But I do test and verify. I should also note that I am not taking into account the rise time of the SD-24 pulse and head and cable BW as would be needed for precise work. I just want to get a general idea of how good or bad the cheap stuff is. I'll get to nit picking when my 8510C is installed and running in my screen room. After a bunch of really bad Chinese BNCs I've been quite happy with the rest. I think all of the devices I've culled via TDR have been BNCs including a bunch of Pomona attenuators which were simply worn out to the point of not being reliable. I would like to find cheap RG402 jumpers with less reflection at the SMA-M ends. These were the cheapest I could find on ebay. Gedas had asked me to try some and owed him more than one favor. They're not great, but not that bad either. The SD-24 step is 250 mV so you can calculate the RC & SWR from the photos. Have Fun! Reg ----- 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.
|
Re: Velocity of propagation (so I can determine insertion delay) for a 3.5mm coax adapter?
Watch for a Harbor Freight sale flyer offering digital calipers for $10 and get one. Carefully file down the depth gauge tip so there is a 4 mm long segment at the end.
It has 0.05 mm resolution. I've not tested one against my grade B (50 millionths) gauge block set, but 0.5 mm seems a reasonable expectation of accuracy *if and only if* you take into account the coefficient of expansion of steel. Touching the caliper makes it expand much more than most people think. And temperature gets really gnarly and nasty if the materials are different.
Just don't leave the battery in it. Take it out when you're done as otherwise it will be dead the next time you try to use it. A ps one way is ~0.3 mm so you should be able to determine insertion length and velocity factor to better than needed accuracy.
I have the good fortune at almost 67 of being able to buy almost any T&M toy I want. Yesterday was an HP 4145B & 16058A test set. But I have very painful memories of when that was not the case. So my primary interest is T&M on a budget. I'm trying to show how to do more with less. My latest adventure is a $28 ebay noise source.
I'm quite impressed by the performance. Marketing, no. It's a 1 GHz source not a 2 GHz source and needs 10 V, not 12 V. But to go farther I'll have to learn how to calibrate and use my 8970B & 346B. That's awkward in my current lab space constraints.
Bug me directly if I happen to drop off the list for a while to get real work like my lab done. PMs via the mailing list seem to work best if I forget to check the SPAM bin.
Have Fun!
Reg
|
Re: N-Connectors at uW Frequencies
Here is an Agilent reply to a previous question on the same topic:
"Excellent question. N-type industry connectors generally work only up to 18 GHz, but still E4440A that goes up to 26.5 GHz has an N-type connector.
The concern is that a Type-N connector would introduce “modes” in the Spectrum Analyzer frequency response. A “mode” occurs when the Type-N connector structure becomes a weakly resonant cavity. This resonance would appear as a very narrow band “suckout” in the connector insertion loss. This mode would presumably not be corrected by the Spectrum Analyzer flatness calibration.
Behind the Spectrum Analyzer front panel the internal cabling is 3.5 mm. The Type-N front panel connector is simply a Type-N (f) to 3.5 mm (f) adapter. So, any “moding” would be attributable to the Type-N input adapter. Modes do not occur in Agilent Spectrum Anayzers for several reasons.
Advances in precision machining have allowed Type-N connectors to operate mode-free to 26.5 GHz. The PSA Type-N adapter shares many characteristics with the Agilent 1250-1745, Type-N (f) to 3.5 mm (f) adapter. Take for instance the Agilent 1250-1745 3.5 mm (f) to Type-N (f) adapter. The inner shield of the Type-N connector is solid. General purpose Type-N adapters (the Agilent 1250-1778 for instance) have a slotted inner shield. The slots allow for mechanical tolerance variations, however they create a non-uniform ground plane on the inner shield. The slots can lead to fringing of the E-field. These fringing effects impact connector performance.
Another area where Agilent has contributed to connector performance is in the center conductor support bead. Older Type-N adapters have Delron, or Teflon supporting the inner conductor. Looking into a general purpose Type-N adapter one can see a white dielectric surrounding the center conductor. The 1250-17XX family of Type-N adapters have a black support structure. This black support bead is an Agilent patented material that suppresses modes.
One simple test to look for modes can be done by mating an Agilent 1250-1744 (Type-N (m) to 3.5 mm (f)) and a 1250-1745 (Type-N (f) to 3.5 mm (f)). The insertion loss of the assembly can be measured with a network analyzer. The measurements should be made over a very narrow bandwidth stepping from 18 GHz to 26.5 GHz. Any modes will appear as a notch in the S21 response. Extensive testing on a PNA Network Analyzer with many connector combinations has shown the 1250-17XX family to be mode free."
|
Re: N-Connectors at uW Frequencies
I'd like to suggest sweeping a 3.5 or 2.48 mm to N-F adapter up to 26 GHz or above with a known good cable on each side, zero padding the data out 500 GHz and doing a transform to time. That will give ps resolution of the reflections at the cable to adapter interfaces. You can get femtosecond resolution by adding more zeros. Though at some point your sampling in frequency and accuracy will enter the picture if you push it too far.
If someone posts data I'll be happy to kibbitz on the FFT.
In the testing I've been doing using a Tek 11801 & SD-24, I've been pleasantly surprised at how little difference there was between Weinschel APC-7/3.5mm F and Chinese versions. I got very lucky and picked up 4 of the Weinschels new for less per piece than the 2 Chinese I bought just to test. The Weinschels are better, especially the APC-7 side. But they actually differ between assemblies. I had similar experience with some unmarked, but obviously high quality SS, APC-7/N-F bulkhead adapters.
The TDR plots are here:
Note that this only goes to about 12 GHz because of the Chinese RG402 jumpers I used. But it should give a good idea of what your transform to time should look like. At this point I'm beginning to be more concerned about cable losses than connector quality. But I do test and verify.
I should also note that I am not taking into account the rise time of the SD-24 pulse and head and cable BW as would be needed for precise work. I just want to get a general idea of how good or bad the cheap stuff is. I'll get to nit picking when my 8510C is installed and running in my screen room. After a bunch of really bad Chinese BNCs I've been quite happy with the rest. I think all of the devices I've culled via TDR have been BNCs including a bunch of Pomona attenuators which were simply worn out to the point of not being reliable.
I would like to find cheap RG402 jumpers with less reflection at the SMA-M ends. These were the cheapest I could find on ebay. Gedas had asked me to try some and owed him more than one favor. They're not great, but not that bad either. The SD-24 step is 250 mV so you can calculate the RC & SWR from the photos.
Have Fun!
Reg
|
Re: HP8562A Errors 304,335,317,333
Hi Stuart
ALL HP instrument switch mode power supplies need a minimum load to operate correctly, most of the time, but not always, it is the +5VDC supply that needs loading. The +5VDC supply is rated at 4.5A so a minimum load would be say 3A. However you might find that additional loads on the other supply rails are needed.
Be very careful HP manufactured a number of power supplies based on the same basic design, but they are not always interchangeable.
Regards
George G6HIG On Tuesday, 28 April 2020, 23:13:21 BST, stur_mm via groups.io <stuart.ian.richardson@...> wrote:
Dave, ? J1 is open circuit and not under load, board is running standalone so not connected to anything. C322 replaced with new and not suspect. ? I have a donor comparison board although it’s not identical from another SA not sure but need to check . May take a picture maybe someone can identify it so I can use as a straight swap. ? Maybe it needs to be connected to load not sure but checking voltages all are there but swinging about the rails. ? So: 5Volts – 9-10volts both sides of toroid. 28.3volts – Fixed but swinging between 42-46Volts on the non reg side. -15+15Volts – 14-15volts ? Very odd. Is there some regulated or control that needs to be in place as a result of un connected cable, not sure.. ? Cheers, Stuart. ?
From: [email protected] <[email protected]> On Behalf Of David Speck
Sent: 28 April 2020 18:35
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8562A Errors 304,335,317,333 ? Stu, A significant voltage drop across L301 suggests that there is a significant current drain downstream. This could be C322 gone leaky, or perhaps L 301 has partly failed and is showing a higher than normal resistance to a normal current drain.? How big is the voltage drop across L301? Can you measure the DC resistance of L301?? It should be fairly close to zero ohms.? Dave On 4/28/2020 1:18 PM, stur_mm via groups.io wrote: Peter et al, ? Plot thickens as testing all relevant test points shows that the PSU seems to be “Hunting” as the 28.0volt is stable all other voltages are swinging up and down as much as 3volts really not sure what is going on. 28v stable as sourced from a supply voltage of 40Volts ? ? Peter, ? Hi, manage to find the time to get stuck in to replacing them . Before I take the plunge just done a quick offline check of voltages and found that all OK including the 28.3Volt supply rail but TP308 shows a very wobbly 8-9volts DC, should be +5Volts. Tried My other PSU board and rock solid at +5.0volt. Not sure on the TP308 +5V FL what does FL mean ? ? 
? L301 is a Toroid so not sure why there would be a differential across it? ? Both TP307 and 308 show same volts not +5v though.(Is this not some simple filtering? Somethings a miss somewhere. Any Guidance would be appreciative. ? Stu. ?
|
Re: HP8562A Errors 304,335,317,333
Dave, ? J1 is open circuit and not under load, board is running standalone so not connected to anything. C322 replaced with new and not suspect. ? I have a donor comparison board although it’s not identical from another SA not sure but need to check . May take a picture maybe someone can identify it so I can use as a straight swap. ? Maybe it needs to be connected to load not sure but checking voltages all are there but swinging about the rails. ? So: 5Volts – 9-10volts both sides of toroid. 28.3volts – Fixed but swinging between 42-46Volts on the non reg side. -15+15Volts – 14-15volts ? Very odd. Is there some regulated or control that needs to be in place as a result of un connected cable, not sure.. ? Cheers, Stuart. ?
toggle quoted message
Show quoted text
From: [email protected] <[email protected]> On Behalf Of David Speck
Sent: 28 April 2020 18:35
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8562A Errors 304,335,317,333? Stu, A significant voltage drop across L301 suggests that there is a significant current drain downstream. This could be C322 gone leaky, or perhaps L 301 has partly failed and is showing a higher than normal resistance to a normal current drain.? How big is the voltage drop across L301? Can you measure the DC resistance of L301?? It should be fairly close to zero ohms.? Dave On 4/28/2020 1:18 PM, stur_mm via groups.io wrote: Peter et al, ? Plot thickens as testing all relevant test points shows that the PSU seems to be “Hunting” as the 28.0volt is stable all other voltages are swinging up and down as much as 3volts really not sure what is going on. 28v stable as sourced from a supply voltage of 40Volts ? ? Peter, ? Hi, manage to find the time to get stuck in to replacing them . Before I take the plunge just done a quick offline check of voltages and found that all OK including the 28.3Volt supply rail but TP308 shows a very wobbly 8-9volts DC, should be +5Volts. Tried My other PSU board and rock solid at +5.0volt. Not sure on the TP308 +5V FL what does FL mean ? ? 
? L301 is a Toroid so not sure why there would be a differential across it? ? Both TP307 and 308 show same volts not +5v though.(Is this not some simple filtering? Somethings a miss somewhere. Any Guidance would be appreciative. ? Stu. ?
|
Dr. David Kirkby, Kirkby Microwave Ltd
Paul Bicknell