Advantest R3961B S Parameter Test Set N Connector conical inner conductor broken

[The_Praiodanish]

Hi folks,

it's my first time writing here as I hope you can help me out (like many times before as a reader of this forum ) ).
I'm a development engineer for rf electronics and software and an electronics and rf hobbyist.

A few days ago I've bought an Advantest R3961B S-Parameter test set which should be added to my 8753C with some glue logic, it's a test set specified from 300kHz up to 3.6 GHz but a quick hookup to our campany's VNA showed it works quite well up to 6 GHz with roughly 24 dB directivity.

Now I have a big problem with one of the N connector test ports on one of the directional bridges, the previous owner crapped the inner conductor .
I've managed to disassemble the bridge and remove the connector but the conically shaped center conductor is broken.
Does anyone of you know which manufacturer builds similar tapered N connectors or where I can get a replacement?
Thanks a lot in advance!







Here some pics of the opened bridge, only taken with my mobile but I will add some better ones:

[Fraser]

Sadly that is a custom Advantest part.

Your best chance of repair is to buy a high quality N connector, such as used by HP and transplant its gold plated segmented centre to your damaged connector.

I would approach this by grafting the new segmented centre assembly onto the stub of your damaged assembly after removing the damaged part. If the new section is hollow, file the old section to form a pin that inserts into the new section. This provides physical stability.

You will lose the continuous gold plated skin, but if you use silver solder to join the old and new parts, you will still have a good quality connection.

I have had to carry out this sort of repair many times in the past when dealing with unique connectors.

The other alternative is to modify the case of the bridge to take a quality SMA connector. This should be threaded and screwed into the case with just its centre pin connecting to the original connection point for the N connector. Adapters may then be used if an N connector is essential. This option is likely to change the RF characteristics of the input however.

Aurora

[The_Praiodanish]

That's the same solution I've thought about.
While searching the web I've stumbled across the HP 86205A directional bridge which is 300kHz to 6 GHz and looks quite similar.
The service manual shows the following part:





So perhaps Advantest just bought the bridge from HP and did some label engineering?
Anyone of you wants to offer me a 86205-60002 pin & bead assembly? 

[The_Praiodanish]

Hi,

I've done like Aurora suggested, here's the result:



And with this fix I achieve a directivity > 20dB up to 6 GHz, so it will do the job until I may find the correct replacement! Thanks a lot!

[Fraser]

Nice work. Exactly as I did my repairs. Did you use a lathe ? I suggested a file to shape teh end as lathes are less common. I bought a lathe for this sort of small parts work and have never regretted it.

Very interested to see the HP diagram. The centre pin certainly looks very similar. I have seen such tapered centre pins on HP equipment so even if the whole assembly is not HP, the connector could well be. I suspect it would be an expensive spare part though.

I am pleased that the repair went so well and it looks a very professional job I must say 

Aurora

[The_Praiodanish]

I've used the mechanical workshop in our company to machine the part 
I shortened the tapered inner conductor on a lathe and drilled a 1mm centering hole in it, the replacement " inner n connector receptable" is from a new n jack, also shortened on the lathe with a short 1mm pin in the middle to place it centered on the tapered piece (uh hard to describe mechanical stuff properly ^^)

btw I found this rather interesting patent drawing which reminds me extremely on my bridge:
http://patentimages.storage.googleapis.com/pdfs/US4962359.pdf
Even the bond wires added for matching are in the same position!

Thanks again for your advice to repair it this way, I would have struggeled in fear of destroying it completely this way!

[Rupunzell]

Make sure to check the pin depth once this N connector is all back together. If the pin depth is wrong it could damage the mating connector by over stressing the bead. In correct pin depth can cause poor return loss performance beyond damaged connectors.

Good microwave center pins are often made of Beryllium copper, silver-gold plated after the collet end has been heat treated with they collet fingers held close in a heat treating fixture. These are HIGHLY precision devices with tolerances in the 0.01mm range.



Bernice

[The_Praiodanish]

Hi,

the overall length is 0.03 shorter as the original one, I wasn't sure how much height the solder would add so I've shortened the two pieces that the overall length was 0.05 shorter as before.
A too short connector would just cause less matching while a too long one would damage my connectors ;-)
I forgot to scan the two return loss and directivity plots, I will do this next week, now it's time to design some glue logic to interface the test set to my 8753C. It's a great thing that companies like HP put a lot of afford in some very detailed description on how their S-Parameter test set communicates with the VNA. At least I haven't seen companies doing that nowadays...