|
Hi Guys,
I have a slightly OT question.
We use at work a 40 GHz VNA and have microwave test cables with 2.92mm connectors from a well-known cable manufacturer whose name starts with G. Now, a colleague managed to break off one of the contact fingers of the female conductor, so this connector is actually damaged and useless. To me, the connector looks like it can be dismantled and the contacts can be exchanged; so I quoted the manufacturer for a repair of the connector or spare parts so we can do it by ourselves. However, they told me that it cannot be repaired and they don't sell spare parts; if something is damaged, I shall buy a new cable. Well, one of those microwave test cables costs around 2200$, so I am a bit pissed that I should buy a new cable when only a contact finger of a female conductor is broken.
So I wonder:
a) is this normal? what do you do if something from such a connector breaks? I know it should not happen and they should be treated with extreme care, but accidents happen sometimes...
b) or is it only for the "cheap" cables that they don't repair them? I think not repairing such an expensive cable is quite dubious and if I need to buy a new cable I will for sure evaluate another manufacturer which is more trustworthy, like Huber+Suhner.
Have you ever seen that a connector on a microwave VNA test cable got damaged, and what did you do?
Best
Tobias
|
Hi Tobias,
I do not know how they mount the pin to the cable, but I could imagine the whole assembly has a limited number of mate/demates (maybe 1?) So, they will not guarantee performance up to 40 GHz, and because of "commercial reasons" will not do it -> "I got it repaired, but now it's not perfect anymore."
One could always try a "hack-a-fix", like I did with a 2,92 mm Attenuator:
(Sorry, German only, but pictures should be self-explanatory) In your case by salvaging another 2,92 mm connector.
All I experienced so far was: Cable broken -> To the bin. Never seen somebody actually repairing a pre-assembled test-port-cable in a professional context.
Best Regards, Martin
Hi Guys,
I have a slightly OT question.
We use at work a 40 GHz VNA and have microwave test cables with 2.92mm connectors from a well-known cable manufacturer whose name starts with G. Now, a colleague managed to break off one of the contact fingers of the female conductor, so this connector is actually damaged and useless. To me, the connector looks like it can be dismantled and the contacts can be exchanged; so I quoted the manufacturer for a repair of the connector or spare parts so we can do it by ourselves. However, they told me that it cannot be repaired and they don't sell spare parts; if something is damaged, I shall buy a new cable. Well, one of those microwave test cables costs around 2200$, so I am a bit pissed that I should buy a new cable when only a contact finger of a female conductor is broken.
So I wonder:
a) is this normal? what do you do if something from such a connector breaks? I know it should not happen and they should be treated with extreme care, but accidents happen sometimes...
b) or is it only for the "cheap" cables that they don't repair them? I think not repairing such an expensive cable is quite dubious and if I need to buy a new cable I will for sure evaluate another manufacturer which is more trustworthy, like Huber+Suhner.
Have you ever seen that a connector on a microwave VNA test cable got damaged, and what did you do?
Best
Tobias
|
So on the face of it it seems unreasonable but from the manufacturer's point of view it may not be the unhelpful rip-off it seems. Remember the cost to them of processing a single returned cable will be high, especially if they are not set up to do it routinely.
You will expect a fixed price for the repair and, presumably, the repaired item to meet the full specification and be returned to you with all the appropriate certification?
So it will arrive at their facility, it will need logging in to the 'system' someone will most likely have to be pulled from either their lab or production to assess the damaged item, issue a re-work plan and then attempt to repair it. They have to assume (I would on a simple cable) that in the worst case they cannot actually salvage it by re-making the damaged end and that makes it scrap and would then have to supply a new one anyway, plus they've had the additional pain of trying to fix it, and feeding it through their QA system and so on.
From the above you may assume I've been there! However, what we did in that situation was to simply offer the customer a replacement item but at a discounted cost, the amount of the discount did, I admit, vary with how sweet we wished to keep the customer concerned and was typically -30% (sweetish) to 'at cost' (very sweet!)
Adrian
toggle quoted message
Show quoted text
On 7/18/2019 8:40 AM, Tobias Pluess wrote: So I wonder:
a) is this normal? what do you do if something from such a connector breaks? I know it should not happen and they should be treated with extreme care, but accidents happen sometimes...
b) or is it only for the "cheap" cables that they don't repair them? I think not repairing such an expensive cable is quite dubious and if I need to buy a new cable I will for sure evaluate another manufacturer which is more trustworthy, like Huber+Suhner.
|
On 7/18/19 4:00 AM, Martin Rickes wrote: One could always try a "hack-a-fix", like I did with a 2,92 mm Attenuator:
Hey, totally not related to this thread, but I like that adjustable stand in kstecker006.jpg on that page. It looks to be 3D-printed; is that design on Thingiverse? If so, can you send a link? Thanks, -Dave -- Dave McGuire, AK4HZ New Kensington, PA
|
Hi Dave
Here: Do not expect any wonders in load-capacity, it's struggling with a HP Power-Head already.
If you have some detailed questions, I guess we should move off list :)
73 Martin
Am Do., 18. Juli 2019 um 18:23?Uhr schrieb Dave McGuire < mcguire@...>:
On 7/18/19 4:00 AM, Martin Rickes wrote:
> One could always try a "hack-a-fix", like I did with a 2,92 mm Attenuator:
>
? Hey, totally not related to this thread, but? I like that adjustable
stand in kstecker006.jpg on that page.? It looks to be 3D-printed; is
that design on Thingiverse?? If so, can you send a link?
? ? ? ? ? ? ?Thanks,
? ? ? ? ? ? ?-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
|
It still looks very handy. I will print one soon. Thanks!
-Dave
toggle quoted message
Show quoted text
On 7/18/19 1:22 PM, Martin Rickes wrote: Hi Dave
Here:
Do not expect any wonders in load-capacity, it's struggling with a HP
Power-Head already.
If you have some detailed questions, I guess we should move off list :)
73
Martin
Am Do., 18. Juli 2019 um 18:23?Uhr schrieb Dave McGuire
<mcguire@... <mailto:mcguire@...>>:
On 7/18/19 4:00 AM, Martin Rickes wrote:
> One could always try a "hack-a-fix", like I did with a 2,92 mm
Attenuator:
>
? Hey, totally not related to this thread, but? I like that adjustable
stand in kstecker006.jpg on that page.? It looks to be 3D-printed; is
that design on Thingiverse?? If so, can you send a link?
? ? ? ? ? ? ?Thanks,
? ? ? ? ? ? ?-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
--
Dave McGuire, AK4HZ
New Kensington, PA
|
In metal, stronger, and easier...there are cheaper ones too :-)
Stainless Steel Adjustable Drill Lift Laboratory Lifting Platform Table Bench Lifter Router Shank Height Woodworking Lab Jack
https://s.click.aliexpress.com/e/3iTOFq0
|
Nice! I've made a note of that, and will probably get a couple. The
only drawback is that I can't have one of those *tonight*. ;)
(my brain's firmware was compiled without -DENABLE_PATIENCE")
-Dave
toggle quoted message
Show quoted text
On 7/18/19 1:46 PM, Leo Kroonenburg wrote: In metal, stronger, and easier...there are cheaper ones too :-)
Stainless Steel Adjustable Drill Lift Laboratory Lifting Platform Table
Bench Lifter Router Shank Height Woodworking Lab Jack
--
Dave McGuire, AK4HZ
New Kensington, PA
|
Haha, -DENABLE_PATIENCE :-)? my firmware is missing this flag also!
Concerning those adjustable stands, I have several of these here in different sizes. There is a company here in Switzerland which makes them of anodised aluminium. I think they call those stands 'lab boy' or something like that. They are pretty powerful
in terms of the maximum weight they can lift, but don't expect wonders in stability. When in the uppermost position, they become a bit flaky, but otherwise they are really good. I wouldn't use a 3D printed one; I cannot believe it would be robust enough.
Concerning the VNA test cable: well, I didn't expect the manufacturer to repair them, but I certainly would expect them to make the spare parts available, so everybody who has the tools can repair the connectors by himself. And since we do have a VNA we
can even test the cables whether they are good. But for sure I will not buy a new 2000$ cable because a contact finger is broken, this is close to burglary, in my opinion.
For example, you can also buy mechanical wrist watches in a similar price region, and if something breaks, it is absolutely normal to get the mechanics repaired and checked whether everything is in spec, so I really don't see the point concerning the connectors.
But of course I learned something: I will buy the next cable from another manufacturer and also order a LOT of connector savers. Sadly these also cost a fortune for 2.92mm :-(
Tobias
HB9FSX
Sent from my Samsung Galaxy smartphone.
toggle quoted message
Show quoted text
-------- Original message --------
From: Dave McGuire <mcguire@...>
Date: 7/18/19 19:52 (GMT+01:00)
Subject: Re: [HP-Agilent-Keysight-equipment] Slightly OT: VNA test cables
? Nice!? I've made a note of that, and will probably get a couple.? The
only drawback is that I can't have one of those *tonight*. ;)
? (my brain's firmware was compiled without -DENABLE_PATIENCE")
??????????? -Dave
On 7/18/19 1:46 PM, Leo Kroonenburg wrote:
> In metal, stronger, and easier...there are cheaper ones too :-)
>
> Stainless Steel Adjustable Drill Lift Laboratory Lifting Platform Table
> Bench Lifter Router Shank Height Woodworking Lab Jack
>
>
>
--
Dave McGuire, AK4HZ
New Kensington, PA
|
I'd be surprised if any vendor repairs their cables.
As others have pointed out, it isn't really economical in today's environment, #of people who will complain about a 'bad' repair >> number of people who will complain about lack of ability to repair, and I suspect that repairing the cable may significantly impact the integrity of the cable. Probably parts of this are crimped together, which does not lend itself to disassembly and re-assembly at the same length. By the time you are done taking a connector off, replacing the part that's busted, and putting it back on, you are well over 50% of the cost of outright replacing the cable. Remember, the cost isn't in the materials, it's in the knowledge of how to put it together and the labor to do so. And that's just for a connector fix, not the one for the cable that got run over by the cart with 200lbs of gear on it.
What I learned working in the computer repair business for a bunch of years is that 98% of your customers are great. The remainder make you sure everyone in the world is a jerk. Even more so in today's world of instant negative feedback.
Haha, -DENABLE_PATIENCE :-)? my firmware is missing this flag also!
Concerning those adjustable stands, I have several of these here in different sizes. There is a company here in Switzerland which makes them of anodised aluminium. I think they call those stands 'lab boy' or something like that. They are pretty powerful
in terms of the maximum weight they can lift, but don't expect wonders in stability. When in the uppermost position, they become a bit flaky, but otherwise they are really good. I wouldn't use a 3D printed one; I cannot believe it would be robust enough.
Concerning the VNA test cable: well, I didn't expect the manufacturer to repair them, but I certainly would expect them to make the spare parts available, so everybody who has the tools can repair the connectors by himself. And since we do have a VNA we
can even test the cables whether they are good. But for sure I will not buy a new 2000$ cable because a contact finger is broken, this is close to burglary, in my opinion.
For example, you can also buy mechanical wrist watches in a similar price region, and if something breaks, it is absolutely normal to get the mechanics repaired and checked whether everything is in spec, so I really don't see the point concerning the connectors.
But of course I learned something: I will buy the next cable from another manufacturer and also order a LOT of connector savers. Sadly these also cost a fortune for 2.92mm :-(
Tobias
HB9FSX
Sent from my Samsung Galaxy smartphone.
-------- Original message --------
Date: 7/18/19 19:52 (GMT+01:00)
Subject: Re: [HP-Agilent-Keysight-equipment] Slightly OT: VNA test cables
? Nice!? I've made a note of that, and will probably get a couple.? The
only drawback is that I can't have one of those *tonight*. ;)
? (my brain's firmware was compiled without -DENABLE_PATIENCE")
??????????? -Dave
On 7/18/19 1:46 PM, Leo Kroonenburg wrote:
> In metal, stronger, and easier...there are cheaper ones too :-)
>
> Stainless Steel Adjustable Drill Lift Laboratory Lifting Platform Table
> Bench Lifter Router Shank Height Woodworking Lab Jack
>
>
>
--
Dave McGuire, AK4HZ
New Kensington, PA
|
-DENABLE_PATIENCE is deprecated, try -DENABLE-PATIENCE, there
was a standardization of options formats a few years back.
-Chuck Harris
Tobias Pluess wrote:
toggle quoted message
Show quoted text
Haha, -DENABLE_PATIENCE :-) my firmware is missing this flag also!
Concerning those adjustable stands, I have several of these here in different sizes. There is a company here in Switzerland which makes them of anodised aluminium. I think they call those stands 'lab boy' or something like that. They are pretty powerful in terms of the maximum weight they can lift, but don't expect wonders in stability. When in the uppermost position, they become a bit flaky, but otherwise they are really good. I wouldn't use a 3D printed one; I cannot believe it would be robust enough.
Concerning the VNA test cable: well, I didn't expect the manufacturer to repair them, but I certainly would expect them to make the spare parts available, so everybody who has the tools can repair the connectors by himself. And since we do have a VNA we can even test the cables whether they are good. But for sure I will not buy a new 2000$ cable because a contact finger is broken, this is close to burglary, in my opinion.
For example, you can also buy mechanical wrist watches in a similar price region, and if something breaks, it is absolutely normal to get the mechanics repaired and checked whether everything is in spec, so I really don't see the point concerning the connectors. But of course I learned something: I will buy the next cable from another manufacturer and also order a LOT of connector savers. Sadly these also cost a fortune for 2.92mm :-(
Tobias
HB9FSX
Sent from my Samsung Galaxy smartphone.
-------- Original message --------
From: Dave McGuire <mcguire@...>
Date: 7/18/19 19:52 (GMT+01:00)
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] Slightly OT: VNA test cables
Nice! I've made a note of that, and will probably get a couple. The
only drawback is that I can't have one of those *tonight*. ;)
(my brain's firmware was compiled without -DENABLE_PATIENCE")
-Dave
On 7/18/19 1:46 PM, Leo Kroonenburg wrote:
In metal, stronger, and easier...there are cheaper ones too :-)
Stainless Steel Adjustable Drill Lift Laboratory Lifting Platform Table
Bench Lifter Router Shank Height Woodworking Lab Jack
--
Dave McGuire, AK4HZ
New Kensington, PA
|
Dammit, that's what happened!
-Dave
toggle quoted message
Show quoted text
On 7/19/19 8:36 AM, Chuck Harris wrote: -DENABLE_PATIENCE is deprecated, try -DENABLE-PATIENCE, there
was a standardization of options formats a few years back.
-Chuck Harris
Tobias Pluess wrote:
Haha, -DENABLE_PATIENCE :-) my firmware is missing this flag also!
Concerning those adjustable stands, I have several of these here in different sizes. There is a company here in Switzerland which makes them of anodised aluminium. I think they call those stands 'lab boy' or something like that. They are pretty powerful in terms of the maximum weight they can lift, but don't expect wonders in stability. When in the uppermost position, they become a bit flaky, but otherwise they are really good. I wouldn't use a 3D printed one; I cannot believe it would be robust enough.
Concerning the VNA test cable: well, I didn't expect the manufacturer to repair them, but I certainly would expect them to make the spare parts available, so everybody who has the tools can repair the connectors by himself. And since we do have a VNA we can even test the cables whether they are good. But for sure I will not buy a new 2000$ cable because a contact finger is broken, this is close to burglary, in my opinion.
For example, you can also buy mechanical wrist watches in a similar price region, and if something breaks, it is absolutely normal to get the mechanics repaired and checked whether everything is in spec, so I really don't see the point concerning the connectors. But of course I learned something: I will buy the next cable from another manufacturer and also order a LOT of connector savers. Sadly these also cost a fortune for 2.92mm :-(
Tobias
HB9FSX
Sent from my Samsung Galaxy smartphone.
-------- Original message --------
From: Dave McGuire <mcguire@...>
Date: 7/18/19 19:52 (GMT+01:00)
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] Slightly OT: VNA test cables
Nice! I've made a note of that, and will probably get a couple. The
only drawback is that I can't have one of those *tonight*. ;)
(my brain's firmware was compiled without -DENABLE_PATIENCE")
-Dave
On 7/18/19 1:46 PM, Leo Kroonenburg wrote:
In metal, stronger, and easier...there are cheaper ones too :-)
Stainless Steel Adjustable Drill Lift Laboratory Lifting Platform Table
Bench Lifter Router Shank Height Woodworking Lab Jack
--
Dave McGuire, AK4HZ
New Kensington, PA
--
Dave McGuire, AK4HZ
New Kensington, PA
|
VNA cables are very touchy about repairs, however I have repaired some of the APC 3.5mm cables - successfully. I had to do comparison measurements to a few known good cables to check the quality of my repair. My parts source was the opposite end of a damaged cable. I do have a pair of HP 85131F cables that repair was attempted by the factory ($1500) but was still OOT, >0.02dB amplitude change with cable movement. So we bought a new set at >$5000in the 1980¡¯s. The ¡°GTX¡± cables seem to be less repairable than the ¡°H&S¡± cables. I have repaired about 10 of the ¡°H&S¡± cables. I have not had an opportunity to repair a 40GHz, 2.92mm cable yet.
Don Bitters
|
My HP VNA tops out at only 1.3Ghz, but even that frequency, I personally found, makes it next to impossible to make up your own phase-matched cables for. To account for cable stretch during the process; to get the lengths of the core and shield correct for each cable was just too much to ask. By the time you get to 40Ghz, then it's practically black magic and well beyond an amateur repair. So they really have you by the balls when you break one, I'm afraid.
|
Dr. David Kirkby from Kirkby Microwave Ltd
Hi Guys,
I have a slightly OT question.
We use at work a 40 GHz VNA and have microwave test cables with 2.92mm connectors from a well-known cable manufacturer whose name starts with G. Now, a colleague managed to break off one of the contact fingers of the female conductor, so this connector is actually damaged and useless. To me, the connector looks like it can be dismantled and the contacts can be exchanged; so I quoted the manufacturer for a repair of the connector or spare parts so we can do it by ourselves. However, they told me that it cannot be repaired and they don't sell spare parts; if something is damaged, I shall buy a new cable. Well, one of those microwave test cables costs around 2200$, so I am a bit pissed that I should buy a new cable when only a contact finger of a female conductor is broken.
So I wonder:
a) is this normal? what do you do if something from such a connector breaks? I know it should not happen and they should be treated with extreme care, but accidents happen sometimes...
Best
Tobias
If it was an HP/Agilent/Keysight cable, the, at least in 3.5 mm, the centre contacts can be replaced by the user. You just need to buy the contact (not too pricy) and the tool for the job (expensive).?
I don¡¯t know why you have to say a company beginning with G. Why not just say Gore?
The obvious question to ask is whether one finger is going to make an differences. I asked about a 3.5 mm connector (26.5 GHz) max, as I could not see any fingers. Dr. Joel Dunsmore wrote that the fingers were probably there, but even if they were all missing, the male centre pin is unlikely to be right in the middle, so sufficient contact would be made.?
So the question to ask, is whether a single missing finger is going to make any practical difference. I believe that your assumption that the cable is useless is probably not true.?
Do you have a very a VNA verification kit? If so, does it pass the verification??
I know there¡¯s a lot of things written about missing fingers, but I do question the validity of some of the information, when there¡¯s a post on the Keysight website, from Dr. Joel Dunsmore, that the issue is not as serious as one might believe.?
I am not impressed with Gore. I have some leaky feeder made by them, which is incredibly expensive, having be cleared for aeronautical use. The quality of the N connectors is not that great considering the cost of the cables, which make VNA cables seem very cheap. ?
Dave -- 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
|
Dear all
?
I have the following leads
?
85134-60003? sn 01094
?
85134-60004? sn 01086
?
Is there a service manual for them as from
this thread it appears I might be able to replace the worn connectors and
return it to its original specification?
?
Regards Paul B
?
?
toggle quoted message
Show quoted text
On Thu, 18 Jul 2019 at 08:40, Tobias Pluess <tobias.pluess@...> wrote:
Hi Guys,
I have a slightly OT question.
We use at work a 40 GHz VNA and have microwave test cables with 2.92mm
connectors from a well-known cable manufacturer whose name starts with G. Now,
a colleague managed to break off one of the contact fingers of the female
conductor, so this connector is actually damaged and useless. To me, the
connector looks like it can be dismantled and the contacts can be exchanged; so
I quoted the manufacturer for a repair of the connector or spare parts so we
can do it by ourselves. However, they told me that it cannot be repaired and
they don't sell spare parts; if something is damaged, I shall buy a new cable.
Well, one of those microwave test cables costs around 2200$, so I am a bit
pissed that I should buy a new cable when only a contact finger of a female
conductor is broken.
So I wonder:
a) is this normal? what do you do if something from such a connector breaks? I
know it should not happen and they should be treated with extreme care, but
accidents happen sometimes...
Best
Tobias
If it was an HP/Agilent/Keysight cable, the, at least in 3.5 mm, the
centre contacts can be replaced by
the user. You just need to buy the contact (not too pricy) and the tool for the
job (expensive).?
I don¡¯t know why you have to say a company beginning with G. Why
not just say Gore?
The obvious question to ask is whether one finger is going to make an
differences. I asked about a 3.5 mm connector (26.5 GHz) max, as I could not
see any fingers. Dr. Joel Dunsmore wrote that the fingers were probably there,
but even if they were all missing, the male centre pin is unlikely to be right
in the middle, so sufficient contact would be made.?
So the question to ask, is whether a single missing finger is going to
make any practical difference. I believe that your assumption that the cable is
useless is probably not true.?
Do you have a very a VNA verification kit? If so, does it pass the
verification??
I know there¡¯s a lot of things written about missing fingers, but
I do question the validity of some of the information, when there¡¯s a
post on the Keysight website, from Dr. Joel Dunsmore, that the issue is not as
serious as one might believe.?
I am not impressed with Gore. I have some leaky feeder made by them,
which is incredibly expensive, having be cleared for aeronautical use. The
quality of the N connectors is not that great considering the cost of the
cables, which make VNA cables seem very cheap. ?
--
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
No virus found in this message.
Checked by AVG -
Version: 2016.0.8048 / Virus Database: 4793/15884 - Release Date: 08/14/18
Internal Virus Database is out of date.
|
A few comments:
1) Put the suspect connector in series with another connector and examine the time domain response to see how large the reflection coefficient is. Window off the reflection and examine the reflection in the frequency domain. That will tell you the magnitude of the damage and what frequencies are affected.
This is a DSP 101 level task, but not everyone studies that and even if you do, if you haven't done it since school it's not at all obvious how to do it. It is also painfully easy to get it wrong. I'll be happy to write a detailed explanation if there is interest and Dave Kirkby can add it to his site for future reference.
Dunsmore discusses the rudiments of this in his book, but there is far more to it as his treatment assumes someone like me has written the software for you and you have paid Keysight for it.
2) If you have a cable and the connector is clearly damaged, *as determined by actual measurement of the reflection*, beyond use, cut the cable an inch or so from the connector, dissect it *very* carefully and study how it was made. For very high BW cables, I'd expect that spot welding is the preferred technique. However, I have *NO* information. That's just a guess from a guy who has bought and poured over a great many books on manufacturing processes. The K-Weld thread on EEVblog discusses constructing a spot welder of sufficient sophistication. At the price of 2.48 mm connectors you don't want to experiment to find the right settings. Instead, you calculate exactly how much energy is required to melt the metal for the weld. And take the speed with which heat is conducted away from the weld into account.
As the goal would be spot welding copper and copper alloys I'd expect it would require tungsten electrodes which were very precisely made. This is not trivial to do, but well within the abilities of anyone sufficiently determined. $2200 for a single cable would make me *very* determined.
Cable companies do not buy them ready made from God.
Have Fun!
Reg
BTW *Please* trim trailing text when replying to posts to no more than the preceding post.
|
Reg, I'd love to see that process explained/documented.
Thanks,
John
----
toggle quoted message
Show quoted text
On 8/22/19 2:44 PM, Reginald Beardsley via Groups.Io wrote: A few comments:
1) Put the suspect connector in series with another connector and examine the time domain response to see how large the reflection coefficient is. Window off the reflection and examine the reflection in the frequency domain. That will tell you the magnitude of the damage and what frequencies are affected.
This is a DSP 101 level task, but not everyone studies that and even if you do, if you haven't done it since school it's not at all obvious how to do it. It is also painfully easy to get it wrong. I'll be happy to write a detailed explanation if there is interest and Dave Kirkby can add it to his site for future reference.
Dunsmore discusses the rudiments of this in his book, but there is far more to it as his treatment assumes someone like me has written the software for you and you have paid Keysight for it.
2) If you have a cable and the connector is clearly damaged, *as determined by actual measurement of the reflection*, beyond use, cut the cable an inch or so from the connector, dissect it *very* carefully and study how it was made. For very high BW cables, I'd expect that spot welding is the preferred technique. However, I have *NO* information. That's just a guess from a guy who has bought and poured over a great many books on manufacturing processes. The K-Weld thread on EEVblog discusses constructing a spot welder of sufficient sophistication. At the price of 2.48 mm connectors you don't want to experiment to find the right settings. Instead, you calculate exactly how much energy is required to melt the metal for the weld. And take the speed with which heat is conducted away from the weld into account.
As the goal would be spot welding copper and copper alloys I'd expect it would require tungsten electrodes which were very precisely made. This is not trivial to do, but well within the abilities of anyone sufficiently determined. $2200 for a single cable would make me *very* determined.
Cable companies do not buy them ready made from God.
Have Fun!
Reg
BTW *Please* trim trailing text when replying to posts to no more than the preceding post.
|
John et al,
I just finished testing a bunch of Chinese SMA-F to N-F adapters to 20 GHz with a Tek 11801 & SD-24. The SD-24 came today via ebay, from Tektronix! That's the easy way, no math required. Just connect a good 50 ohm load and look at the reflection.
The attached pictures should help:
500.png Chinese SMA-F to N-F adapters from ebay
503.png unmarked but obviously very high quality
508.png the SA normalized to the high quality adapter
509.png the Chinese adapter spectrum
There is a lot more cabling and stuff on the SA setup, so it is not as clean as the 11801 TDR which is just a piece of RG402 feeding the adapters and an Inmet N load. So I really was forced to normalize.
Unfortunately, getting to 40 GHz in the time domain is tough. Even with a 50 GHz SD-32 head, you need a 5-6 ps rise time step which is about 1/4 the 11801 calibrator and SD-24 rise time. Keysight sells a 3.9 ps rise time pulser, but I don't want to know the price.
Let me reread Dunsmore and think about how to do this so I don't bugger up the explanation. Basically, one is going to collect data in the frequency domain, transform to time and measure the amplitude of the reflection.
The issue of concern is that there are constraints on the sampling in the frequency domain. I'm a seismologist, so I'm used to doing things in the opposite direction, from time to frequency. Further complicated by my having never used a VNA despite owning 3. Been too busy playing with my 11801 ;-)
But how you collect the data doesn't affect the mathematics. And in reflection seismology you switch domains constantly.
Have Fun!
Reg
|
Dr. David Kirkby from Kirkby Microwave Ltd
IOn Thu, 22 Aug 2019 at 19:44, Reginald Beardsley via Groups.Io <pulaskite= [email protected]> wrote: A few comments:
1) Put the suspect connector in series with another connector and examine the time domain response to see how large the reflection coefficient is.? Window off the reflection and examine the reflection in the frequency domain.? That will tell you the magnitude of the damage and what frequencies are affected.
It will be hard to know if the reflection is due to the adapter or the cable. A make-female adapter with a slotless connector is I believe going to be needed. A male-female adapter of the same quality as the VNA cables is not going to be cheap (probably > $1000). I checked the Keysight 3.5 mm male-female adapter (85052-60013) and that¡¯s ?1021, which is about $1250.?
Keysight do not to my knowledge sell 2.92 mm connectors, but they do 3.5 mm and 2.4 mm. But I think the 3.5 mm prices give you some idea of the sort of cost of adapters with slotless connectors. If you stick a 2.92 mm or 3.5 mm connector on, that¡¯s not slotless, then I am not convinced one is achieve much.?
FWIW, I just found the 3.5 mm & N slotless repair kits on the Keysight website 85052-60049 (3.5 mm) and 85054-60056 (N). The kits contain 10 contact. If I wanted a Keysight device repaired I believe that I know of a company able to do it. I would not want to but a complete kit to do one contact - I would rather pay someone else to do it. You could possibly argue that the males are never slotless, so the adapter doesn¡¯t need to be slotless, but you still need to connect something to the female end.?
This is a DSP 101 level task, but not everyone studies that and even if you do, if you haven't? done? it since school it's not at all obvious how to do it.? ?It is also painfully easy to get it wrong.? I'll be happy to write a detailed explanation if there is interest and Dave Kirkby can add it to his site for future reference.
Yes, I am willing to add it to?
Is there any chance of some open-source GPL software to do this? A bit of C code that can read a Touchstone file and output the time domain code would be very useful. If it¡¯s in standard C, it could be compiled on any platform. No need for any GUI - people can plot the data with Excel, gnuplot or similar.?
Dunsmore discusses the rudiments of this in his book, but there is far more to it as his treatment assumes someone like me has written the software for you and you have paid Keysight for it.
That¡¯s the problem with his book, but it does have ¡°advanced¡± in the title, and it is clearly aimed at those with access to the sort of software that costs $$$$¡¯s.?
Cable companies do not buy them ready made from God.
I assume you mean Gore, not God! I am not impressed with Gore, although they are better than Pastercrap
Have Fun!
Reg
-- 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
|
Martin Rickes
Dave McGuire
bownes
Jinxie
Paul Bicknell
