Anritsu S820D - damaged RF connector

[TheMG]

Got an Anritsu S820D as "non functional" for a very good price. Didn't take long to find the problem: someone had obviously shoved something into the RF connector (type K, similar to SMA) and spread open all the fingers of the center female contact like a flower.

I was able to bend two of the four fingers back into shape, the other two broke off. It works now, but I'm sure what the two missing fingers is going to do to the impedance of the connector especially at the higher GHz end of things. Anyone know just how much of an impact this might have?

Also, anyone had any luck getting parts (such as just the connector in this case) from Anritsu rather than the complete board/module? Only the board/module part numbers are listed in their service/maintenance manual. I have little doubt that this connector is a custom Anritsu part, since it has the additional threads near the bottom to accommodate a special N-type adapter (comes with the unit) for added mechanical strength.

[geggi1]

to me its looks like the damaged connector is sitting on a N-type connector. Have you checked if the connector is removable and you can use the N-type?

[TheMG]

Thread is smaller than N type.

[TheSteve]

That is an Anritsu proprietary connector as far as I know. You'll probably have to find another device to steal one from. In the short term what I would recommend is installing the proper N adapter and never taking it off or buying a SMA/3.5mm connector saver, installing it and never removing it. The missing petals will not be great for impedance performance but if you perform a calibration before doing any analysis it should compensate for it. The most important thing is to avoid mating cycles. Also you don't want to damage other cables when connecting them as the petals won't be sitting perfect.

[edpalmer42]

Have you looked inside?  I see that model is also available with an ordinary N connector.  Maybe replacing the connector won't be that hard.  The flats on that threaded portion of the connector are just begging to be wrenched on!   

[wraper]

to me its looks like the damaged connector is sitting on a N-type connector. Have you checked if the connector is removable and you can use the N-type?

Yeah, it looks like an adaptor screwed on top of original connector.

[TheMG]

The flats on that threaded portion of the connector are just begging to be wrenched on!   

I've learned years ago that blindly wrenching on an RF connector is not a good idea. In fact, some test equipment manufacturers even include strongly worded warnings about not doing that.

Usually the center conductor needs to be desoldered from the PCB before attempting to unscrew the connector out of the assembly or you'll end up breaking the trace.

[wraper]

Isn't there a hole or two un the side of part without thread? It looked like there is on some not too clear pictures I googled, as if there should be some tool to grab it and unscrew.

[TheMG]

Well I decided to take a leap of faith and try to unscrew the top portion of the connection from the main body of the connector.

Came apart pretty easy and most importantly without tearing anything up inside the microwave module, as the top part of the connector has a long male pin that mates with a female socket deep down inside.

Still seems like a very much custom part though. Given how easy it is to remove, I'm really surprised Anritsu doesn't list that part of the connector as a replaceable part. Maybe I'll try contacting them anyways, and cross my fingers.

Also although I did sort of bend two of the fingers back into shape, it's still not a very reliable contact the results are not always repeatable.

[TheMG]

I had an idea... What if I were to obtain another 2.92mm (K type) female connector, steal the center pin out of it, and graft it onto the center pin of the proprietary connector?

What I have in mind is cutting the damaged end off square a few mm below the base of the fingers/petals, drilling a small hole into the end. Then with the donor pin, cut it to length and machine the end of it so it will insert snugly into the hole previously drilled. Then, a touch of solder in the hole to hold it firmly in place.

With this method, hopefully any impedance bump will be minimal and can easily be calibrated out, but most importantly I end up with a reliable, repeatable connection.

[TheSteve]

Sounds like a fun idea - but can you get it straight enough, or the exact length? If you think you can make it work then I'd go for it. If not I'd make sure the two remaining petals are crimped a little more(or add a small teflon sleeve around them) and tighten a 3.5mm (or 2.92) connector saver on it a little tighter then the recommended 8 in-lbs or use a small amount of thread lock. That will then be your new connector and should never move so there won't be any variable impedances from the original connector damage. It will stick out a little further of course.

I've attached a picture of a 3.5mm connector saver from my collection.

[TheMG]

Well I pretty much had to go with the pin grafting idea. I tried bending the petals back just a little bit more in order to get a reliable connection but another one broke off, leaving only one intact. No good.

I did contact Anritsu but just as I feared, they will not sell a replacement connector piece. They want the unit sent in for evaluation and they to send a repair/cal quote. Forget it. Pretty much guaranteed that would be over $1000!

I did as I had planned except I scrapped the idea of drilling a hole down the middle of the pin, that didn't work. Whatever material the pin is made of, is darn near impossible to drill. It also put a nice gouge in my nice set of flush cutters when I was trimming the pin to length.

So anyways, only the solder is holding the two pieces together, but I was able to get the length to within 0.1mm and the centering is not absolutely perfect but close enough for the connectors to mate without any trouble. Most importantly, I get a reliable electrical connection now.

I have no way to tell how much of an impedance bump this has created (obviously it's not as good as the original intact connector would have been). I could only compare with another instrument going to 3.3GHz and the results appear the same. Beyond 3.3GHz up to the max of 20GHz... no idea. While any imperfections in the connector can theoretically be calibrated out, wouldn't this potentially degrade the available dynamic range of the instrument?

Anyways, I'm not willing to dump $1000+ to getting this properly repaired, so it will have to do.

[Miti]

I did as I had planned except I scrapped the idea of drilling a hole down the middle of the pin, that didn't work. Whatever material the pin is made of, is darn near impossible to drill. It also put a nice gouge in my nice set of flush cutters when I was trimming the pin to length.

It is not a good idea to drill, cut or solder those pins. Good quality ones, and coming from an Anritsu instrument I bet this is one, are made of beryllium copper. Beryllium is not something that you want to breathe or even come into direct contact with. The cheaper ones are made of bronze phosphorus, but you can’t tell.