How to use these SMA connectors?

[Andrey_irk]

Hello everyone!
I've come across these connectors (attached) and I don't understand what are the use cases for them.
So, it is supposed to be mounted on a panel, thats is ok.  Then, I suspect, there should be a 7mm-thick wall with a 4.1mm hole, but what's next? Is it supposed to be connecter to a PCB? But the center contact is too thick and expands too far away, I guess. Or maybe there are some kind of adaptors of sliding contacts that connect to it?
Moreover, there are modifications of this connector that have even longer coaxial end! Like 15mm, 20mm which makes it even more weird (20mm-thick walls?).
This connector has the maximum frequency of 12GHz, but there are 18GHz ones with the same dimensions.
The dielectric is PTFE.

[Andrey_irk]

Yes, but how the other end is supposed to be connected?

[xrunner]

Yes, but how the other end is supposed to be connected?

Well it's not SMA isn't it SMB? One end snaps on the SMB  cable and the other end is soldered to the coax cable - center conductor to center pin and braid to ground. 

[Andrey_irk]

Yes, but how the other end is supposed to be connected?

Well it's not SMA isn't it SMB? One end snaps on the SMB  cable and the other end is soldered to the coax cable - center conductor to center pin and braid to ground. 

Let it be SMB (it is a standard interface anyway) that is not the point. But how would you solder a coax cable to it? I mean, for a solder joint to work there has to be a hole in the center contact, in which you put the center conductor of the cable. Because otherwise you'll have to solder them one on another and it will ruin your VSWR. And it is even worse with the dielectric and the shield. Remember, we need to get a good VSWR up to 12 or even 18GHz!

[Bud]

It is soldered to PCB track behind the panel. A lot of boxed RF modules such as Minicircuits ones use them.

Edit: it may also be used as a transition from coax cable to waveguide and back.

[Andrey_irk]

It is soldered to PCB track behind the panel. A lot of boxed RF modules such as Minicircuits ones use them.

I suspected that. Isn't it a problem that the center contact expands 4mm away and is 1.27mm in diameter? There should be a very wide track for it not to create a capacitive discontinuity.

EDIT: Twe coax to waveguide transition thing makes more sense. I've never thought of that, thanks!

[chrisl]

Not exactly the same connector but used like this one....

[Andrey_irk]

Not exactly the same connector but used like this one....

Yeah, these aren't the same.

I've used the ones that looked like the ones on the file attached.
These work up to 18GHz and, as you can see, the center pin is much smaller. But it still creates some discontinuity (unless you get a microstrip with wide track). Also, there is no PTFE around the pin.
The whole coax to PCB transition consists of the connector  (attached), an airline (when the center pin goes through the wall) and the actual PCB.

As for the connector from the topic - the dimentions don't make sense for me (if it is a coax to microstrip transition).

The reason I started this topic is that I see a lot of connector similar to this one, which means they must be commonly used, but how? Can someone show an example?

So far the only use I see is coax to waveguide transition.

[hagster]

I sometimes use these connectors when making custom one-off antennas etc. E.g feeding the base of a helix, or on a feed on a air dielectric patch antenna. Generally I need  to cut down the PTFE and conductor to the desired length.

I agree that there is no standard way to make a connection from these to any form of standard transmission line (microstrip of coax), but they are still useful.

[Wolfgang]

It is supposed to go into a milled enclosure with a wall thickness equal to the length of the PTFE insulator. The center conductor lands directly on a 50 Ohm PCB stripline.
If everything is made precisely, this could have an VSWR of 1.2 or less up to 18GHz.

[Andrey_irk]

It is supposed to go into a milled enclosure with a wall thickness equal to the length of the PTFE insulator. The center conductor lands directly on a 50 Ohm PCB stripline.
If everything is made precisely, this could have an VSWR of 1.2 or less up to 18GHz.

I've just modeled the transition in CST just to illustrate my point better (some of the people don't get it by the looks of it).
There is a rear part of the connector with the dimentions from the drawing (first post), then the center pin goes onto a pcb. There are two simulations with two different PCBs. The substrate material is Rogers RT5880 (it is basically PTFE) in both cases. The thickness of the first board is 0.254mm, of the second board - 0.508mm. Microstrip widths are 0.77mm and 1.56mm accordingly.
The results show significant capacitive discontinuity (as to be expected, see TDR plots). In the second simulation it is significantly smaller due to wide microstrip.  Although it is possible to further compensate the second transition by narrowing down the microstrip under the pin, it is not much one can do to get even 1.3 VSWR in the first case. If, however, we choose a substrate with higher dielectric constant (in case of, say, RO4003 or 4350) then the track width will be even more narrow thus making the problem even bigger. Imagine what happens if the substrate E is as high as 10.

The background material of the model is PEC, there is some air over the PCB (there is a hidden cube made of vacuum) and the upper boundary is open, the others are conductive.

Am I missing something? How would you compensate the transition in the first case?

[ejeffrey]

I would guess it was designed to be used with a thicker substrate.  I have used similar connectors with a 1 mm RO4350.  The much wider center trace then reduces the mismatch.  I also trimmed the dielectric back to match the box wall, and trimmed the center pin as well.

But this brings up something that completely pisses me off about connector manufacturers.  They will happily quote a VSWR of a connector like this without specifying how it is mounted.  I realize that there might be more than one way to use a connector but without showing the configuration you use specs like VSWR are undefined and meaningless.

[hagster]

I would guess it was designed to be used with a thicker substrate.  I have used similar connectors with a 1 mm RO4350.  The much wider center trace then reduces the mismatch.  I also trimmed the dielectric back to match the box wall, and trimmed the center pin as well.

But this brings up something that completely pisses me off about connector manufacturers.  They will happily quote a VSWR of a connector like this without specifying how it is mounted.  I realize that there might be more than one way to use a connector but without showing the configuration you use specs like VSWR are undefined and meaningless.

I totally agree. When you buy am IC it comes with a full data sheet that details how it performs under different conditions. These connectors largely come without anything more than basic dimensions.

That would perhaps be ok for a very cheap itrms, but when the big brands charge a small fortune for them they really should step up the quality of the documentation.

[chrisl]

It is the same kind of type of connector...    a view from a different angle. 
Widely used by Minicircuits up to 12GHz in their modules.

In old days I used the ones shown in your original post a lot for module designs.
1. choose the dielectric thickness so the transmission line width is wider than the center conductor.
2. trim the PTFE with an exacto knife so it will flush with the inner wall of the enclosure.
3. trim the center conductor as short as possible.

In multilayer PCBs I sometimes drop the reference to layer3 by having a GND copper void right under the PAD as well.

[TheUnnamedNewbie]

I would guess it was designed to be used with a thicker substrate.  I have used similar connectors with a 1 mm RO4350.  The much wider center trace then reduces the mismatch.  I also trimmed the dielectric back to match the box wall, and trimmed the center pin as well.

But this brings up something that completely pisses me off about connector manufacturers.  They will happily quote a VSWR of a connector like this without specifying how it is mounted.  I realize that there might be more than one way to use a connector but without showing the configuration you use specs like VSWR are undefined and meaningless.

I totally agree. When you buy am IC it comes with a full data sheet that details how it performs under different conditions. These connectors largely come without anything more than basic dimensions.

I have more than once gotten STEP and HFSS files of the connector to include them in simulation. I think the reason they don't give you simple plots of the connector performance is because it depends so much on how it is built into the circuit.

[ConKbot]

If you really want to be frustrated, try to find the PCB mount socket for the center conductor that lets you use that style of connector as a field replaceable connector on RF modules. I.e. Undo the 4 screws, and you can pull the mangled SMA out of the the amplifier module or whatever, plug a new one in, and re-do the screws.

Also, re:vswr, I think the key is that the housing is connected with low impedance to the PCB, so you can get a better VSWR in the minicircuits style module.  The discontinuity would certainly be pretty bad in any other case. In your CST model, assuming the bottom face is PEC, throw a slug of copper or PEC between the bottom face and the bottom of the dielectric.

[Andrey_irk]

Also, re:vswr, I think the key is that the housing is connected with low impedance to the PCB, so you can get a better VSWR in the minicircuits style module.  The discontinuity would certainly be pretty bad in any other case. In your CST model, assuming the bottom face is PEC, throw a slug of copper or PEC between the bottom face and the bottom of the dielectric.

It is already there. Otherwise the VSWR would be much worse. The background material is PEC (as I said in my post), I just added a brick of air over the pcb, but it is hidden.

If you really want to be frustrated, try to find the PCB mount socket for the center conductor that lets you use that style of connector as a field replaceable connector on RF modules. I.e. Undo the 4 screws, and you can pull the mangled SMA out of the the amplifier module or whatever, plug a new one in, and re-do the screws.

I saw field replaceable sockets that are used with sealed plugs - didn't see any problems at all, especially given the fact that the center pins of the plugs are much smaller, which makes them usable with thin substrates.
As for the sockets - I thought they were used mostly to relief the strain which occures with the change of temperature and usually in conjunction with glass sealed plugs.
Yeah, you can use them to make the connector replaceable, but I imagine the diameter of the thing...

[Andrey_irk]

I would guess it was designed to be used with a thicker substrate.  I have used similar connectors with a 1 mm RO4350.  The much wider center trace then reduces the mismatch.  I also trimmed the dielectric back to match the box wall, and trimmed the center pin as well.

But this brings up something that completely pisses me off about connector manufacturers.  They will happily quote a VSWR of a connector like this without specifying how it is mounted.  I realize that there might be more than one way to use a connector but without showing the configuration you use specs like VSWR are undefined and meaningless.

I totally agree. When you buy am IC it comes with a full data sheet that details how it performs under different conditions. These connectors largely come without anything more than basic dimensions.

I have more than once gotten STEP and HFSS files of the connector to include them in simulation. I think the reason they don't give you simple plots of the connector performance is because it depends so much on how it is built into the circuit.

I wonder, who is that generous manufacturer and in what quantities do you buy connectors?
The only thing I've seen is just 3D models without internal details - for mechanical design.

[coppercone2]

I had an isolator that someone sat on and destroyed the input connector, I managed to fix it by dabbing one of those things you got with some ultra high conductivity silver epoxy and replacing it. Soldering it in was way impossible. Maybe shiva could do it with 4 hands, if time was slowed down and you had praying mantis vision.

[Andrey_irk]

It is the same kind of type of connector...    a view from a different angle. 
Widely used by Minicircuits up to 12GHz in their modules.

In old days I used the ones shown in your original post a lot for module designs.
1. choose the dielectric thickness so the transmission line width is wider than the center conductor.
2. trim the PTFE with an exacto knife so it will flush with the inner wall of the enclosure.
3. trim the center conductor as short as possible.

In multilayer PCBs I sometimes drop the reference to layer3 by having a GND copper void right under the PAD as well.

Ah, I see. It wasn't clear from the first photo. What is the frequency band of the device? A couple of GHz? That solder blob on the pin makes me nervous  .

So, it is hard to follow this procedure if you want to produce any reasonable quantity of devices. Trimming every single connector by hand... Do Minicircuits do it this way?
And then in case of a multilayer PCB you'll have to have several layers of expensive material instead of one just to make that transition. Not to mention designing a good transition from one layer to another (not that it is particularly hard, but not everyone has access to a 3D EM simulation package).

Thanks again for your insightful response!

[TheUnnamedNewbie]

I would guess it was designed to be used with a thicker substrate.  I have used similar connectors with a 1 mm RO4350.  The much wider center trace then reduces the mismatch.  I also trimmed the dielectric back to match the box wall, and trimmed the center pin as well.

But this brings up something that completely pisses me off about connector manufacturers.  They will happily quote a VSWR of a connector like this without specifying how it is mounted.  I realize that there might be more than one way to use a connector but without showing the configuration you use specs like VSWR are undefined and meaningless.

I totally agree. When you buy am IC it comes with a full data sheet that details how it performs under different conditions. These connectors largely come without anything more than basic dimensions.

I have more than once gotten STEP and HFSS files of the connector to include them in simulation. I think the reason they don't give you simple plots of the connector performance is because it depends so much on how it is built into the circuit.

I wonder, who is that generous manufacturer and in what quantities do you buy connectors?
The only thing I've seen is just 3D models without internal details - for mechanical design.

Amphenol RF is one of them. For this specific connector I used recently (in low-volume, ordered like 8 connectors from digi-key), the HFSS model is online, and you can just download it after setting up an account for free.

[eb4fbz]

Coaxial to microstrip transitions should allways be carefully simulated. Sometimes you place a small gap between the pin dielectric and PCB edge for stray capacitance compensation.

Is not something the connector manufacturer can do for you, as there are too many variables.

[ejeffrey]

Coaxial to microstrip transitions should allways be carefully simulated. Sometimes you place a small gap between the pin dielectric and PCB edge for stray capacitance compensation.

Is not something the connector manufacturer can do for you, as there are too many variables.

I'm not asking them to enumerate every possible configuration or detail.  I just want them to tell me the test conditions under which their performance numbers were measured.  If they don't want to do that I think they shouldn't put VSWR in the datasheet at all -- I think that is dishonest.

[hendorog]

It is supposed to go into a milled enclosure with a wall thickness equal to the length of the PTFE insulator. The center conductor lands directly on a 50 Ohm PCB stripline.
If everything is made precisely, this could have an VSWR of 1.2 or less up to 18GHz.

I've just modeled the transition in CST just to illustrate my point better (some of the people don't get it by the looks of it).
There is a rear part of the connector with the dimentions from the drawing (first post), then the center pin goes onto a pcb. There are two simulations with two different PCBs. The substrate material is Rogers RT5880 (it is basically PTFE) in both cases. The thickness of the first board is 0.254mm, of the second board - 0.508mm. Microstrip widths are 0.77mm and 1.56mm accordingly.
The results show significant capacitive discontinuity (as to be expected, see TDR plots). In the second simulation it is significantly smaller due to wide microstrip.  Although it is possible to further compensate the second transition by narrowing down the microstrip under the pin, it is not much one can do to get even 1.3 VSWR in the first case. If, however, we choose a substrate with higher dielectric constant (in case of, say, RO4003 or 4350) then the track width will be even more narrow thus making the problem even bigger. Imagine what happens if the substrate E is as high as 10.

The background material of the model is PEC, there is some air over the PCB (there is a hidden cube made of vacuum) and the upper boundary is open, the others are conductive.

Am I missing something? How would you compensate the transition in the first case?

Have you tried removing some of the ground plane under the transition/pin? I haven't read the thread fully but if I understand correctly that you are trying to optimise the coax to microstrip transition then that might help.

[eb4fbz]

Coaxial to microstrip transitions should allways be carefully simulated. Sometimes you place a small gap between the pin dielectric and PCB edge for stray capacitance compensation.

Is not something the connector manufacturer can do for you, as there are too many variables.

I'm not asking them to enumerate every possible configuration or detail.  I just want them to tell me the test conditions under which their performance numbers were measured.  If they don't want to do that I think they shouldn't put VSWR in the datasheet at all -- I think that is dishonest.

I bet they dont take the coaxial to microstrip transition into account in their specs. It's probably just the VSWR and loss of the coaxial structure.

[Andrey_irk]

It is supposed to go into a milled enclosure with a wall thickness equal to the length of the PTFE insulator. The center conductor lands directly on a 50 Ohm PCB stripline.
If everything is made precisely, this could have an VSWR of 1.2 or less up to 18GHz.

I've just modeled the transition in CST just to illustrate my point better (some of the people don't get it by the looks of it).
There is a rear part of the connector with the dimentions from the drawing (first post), then the center pin goes onto a pcb. There are two simulations with two different PCBs. The substrate material is Rogers RT5880 (it is basically PTFE) in both cases. The thickness of the first board is 0.254mm, of the second board - 0.508mm. Microstrip widths are 0.77mm and 1.56mm accordingly.
The results show significant capacitive discontinuity (as to be expected, see TDR plots). In the second simulation it is significantly smaller due to wide microstrip.  Although it is possible to further compensate the second transition by narrowing down the microstrip under the pin, it is not much one can do to get even 1.3 VSWR in the first case. If, however, we choose a substrate with higher dielectric constant (in case of, say, RO4003 or 4350) then the track width will be even more narrow thus making the problem even bigger. Imagine what happens if the substrate E is as high as 10.

The background material of the model is PEC, there is some air over the PCB (there is a hidden cube made of vacuum) and the upper boundary is open, the others are conductive.

Am I missing something? How would you compensate the transition in the first case?

Have you tried removing some of the ground plane under the transition/pin? I haven't read the thread fully but if I understand correctly that you are trying to optimise the coax to microstrip transition then that might help.

Just removing the gound plane won't help much as the PCB is supposed to be soldered on a metal surface. But if I remove the ground and create a cavity under it  - this might work. I'll give it a try!
Although this approach may add some problems during the assembly process (you need a good contact between the PCB and the enclosure and it is hard to make sure the solder doesn't get into the cavity).

[Andrey_irk]

Coaxial to microstrip transitions should allways be carefully simulated. Sometimes you place a small gap between the pin dielectric and PCB edge for stray capacitance compensation.

Is not something the connector manufacturer can do for you, as there are too many variables.

I'm not asking them to enumerate every possible configuration or detail.  I just want them to tell me the test conditions under which their performance numbers were measured.  If they don't want to do that I think they shouldn't put VSWR in the datasheet at all -- I think that is dishonest.

I bet they dont take the coaxial to microstrip transition into account in their specs. It's probably just the VSWR and loss of the coaxial structure.

I'm aware of this actually. Nevertheless, how would you connect together the ones from my first post achieving good VSWR? Tricky, huh? And wil it have anything to do with reality?

[ejeffrey]

Coaxial to microstrip transitions should allways be carefully simulated. Sometimes you place a small gap between the pin dielectric and PCB edge for stray capacitance compensation.

Is not something the connector manufacturer can do for you, as there are too many variables.

I'm not asking them to enumerate every possible configuration or detail.  I just want them to tell me the test conditions under which their performance numbers were measured.  If they don't want to do that I think they shouldn't put VSWR in the datasheet at all -- I think that is dishonest.

I bet they dont take the coaxial to microstrip transition into account in their specs. It's probably just the VSWR and loss of the coaxial structure.

How would they measure that?  And why would I care about that?