Help identifying an RF connector.

[Fflint]

I have these 460MHz old CDMA amplifiers that I retuned to 435MHz (70cm ham band). They have these weird connectors that look exactly like MCX but the male part has to be 3.5mm rather than 3.75mm that normal MCX is.

This is the board (female) side. Looks exactly like MCX but it is 3.5mm inside.


And this is the modified mcx (outside diameter filed down to 3.5mm) that fits that connector:


I was hoping this is mmcx, but mmcx is definitely too small.

So what on earth is it? I have 4 of them and I'd rather not have to file them all.

[ahbushnell]

Like this? MMCX

https://www.digikey.com/en/products/detail/samtec-inc/MMCX-J-P-H-ST-TH1/6561611

[Fflint]

Like this? MMCX

https://www.digikey.com/en/products/detail/samtec-inc/MMCX-J-P-H-ST-TH1/6561611

No, mmcx is too small (as I wrote in my post).



It is too small. Where mmcx is 2.7mm it should be 3.5mm. it doesn't mate

[tooki]

Doesn’t look “exactly like MCX”, insofar as there’s no spring basket inside.

Could it be a “1.0/2.3” slide-on connector?

See pages 420ff (PDF pages 87ff) of the attached Pasternack RF connector catalog.

[Fflint]

Doesn’t look “exactly like MCX”, insofar as there’s no spring basket inside.

What spring basket? We're looking at a female part. Here is a picture of both male and female MCX:



On the right there is the female connector. Looks very similar to what I have on the pcb, but the ID is 3.5mm. On the left is the male. I haven't got one. The connector I show that I managed to fit is a ground down MCX.

Could it be a “1.0/2.3” slide-on connector?

See pages 420ff (PDF pages 87ff) of the attached Pasternack RF connector catalog.

They don't give any dimensions other then external ones. But measuring pixels on the image the male part seems like it would fit. However the drawing is missing the white plastic isolator.

Is this a kind of connector one would use for 450mhz?

[tooki]

Doesn’t look “exactly like MCX”, insofar as there’s no spring basket inside.

What spring basket? We're looking at a female part. Here is a picture of both male and female MCX:



On the right there is the female connector. Looks very similar to what I have on the pcb, but the ID is 3.5mm. On the left is the male. I haven't got one. The connector I show that I managed to fit is a ground down MCX.

Oops sorry, never mind — I was thinking MCX but looking at the diagram of SMB… 

With that said, are you absolutely, positively certain it’s not just MCX? The 0.2mm difference between 3.5 and 3.7mm is easily within measurement error for measuring an inner diameter with standard calipers.

Could it be a “1.0/2.3” slide-on connector?

See pages 420ff (PDF pages 87ff) of the attached Pasternack RF connector catalog.

They don't give any dimensions other then external ones.

No, but you could google to find them. And you could use calipers to measure additional dimensions of your connector. Then use that guide to narrow down to candidates and then look for the full dimensions of those candidates.

FYI, a huge number of RF connectors are defined in US military standard MIL-STD-348B: https://quicksearch.dla.mil/qsDocDetails.aspx?ident_number=35726

You can also use the Huber+Suhner RF coax connectors catalog PDF, which has high-quality drawings of practically every RF connector on the intro page to each connector’s section.

(Also, though it’s not a complete guide to connector types, I also came across H+S’s RF Connector Guide, which goes into all the theory of RF connections, materials, measurement, etc. Looks to be an amazing resource! At around the middle, it does have some guidance on which connector types are used for what applications.)


There’s also the Amphenol RF website, e.g. the 1.0/2.3 page here: https://www.amphenolrf.com/rf-connectors/1-0-2-3-connectors.html

But measuring pixels on the image the male part seems like it would fit. However the drawing is missing the white plastic isolator.

Yeah, it is. :/ And  the H+S catalog section on them makes me think it’s not that after all, given that it’s designed to be an insert in big multipin connectors, apparently.

Is this a kind of connector one would use for 450mhz?

I’m sure you could. 450MHz* is not a particularly demanding frequency, and those connectors are rated from DC to 10GHz.

*I assume you mean MHz (megahertz) and not mHz (millihertz), and not milli-hecto-zepto-something.

[Fflint]

Doesn’t look “exactly like MCX”, insofar as there’s no spring basket inside.

What spring basket? We're looking at a female part. Here is a picture of both male and female MCX:



On the right there is the female connector. Looks very similar to what I have on the pcb, but the ID is 3.5mm. On the left is the male. I haven't got one. The connector I show that I managed to fit is a ground down MCX.

Oops sorry, never mind — I was thinking MCX but looking at the diagram of SMB… 

With that said, are you absolutely, positively certain it’s not just MCX? The 0.2mm difference between 3.5 and 3.7mm is easily within measurement error for measuring an inner diameter with standard calipers.

Well. I do have MCX pigtails and I had to file down 0.2mm from the diameter. I only used calipers, but I trust myself with them to 50 microns or so. (12 thou for the people thinking in inches).

If I was to guess if say it is some variant of mcx perhaps? But what? Who knows. That's why I started this thread.

I have 4 such boards and 2 mcx pigtails. 0.2mm on a diameter is a lot when something barely fits at 3.5mm. I'm pretty sure 8t is no "normal MCX", but maybe a variant?

This stuff was made in 1990s BTW.

Could it be a “1.0/2.3” slide-on connector?

See pages 420ff (PDF pages 87ff) of the attached Pasternack RF connector catalog.

They don't give any dimensions other then external ones.

No, but you could google to find them. And you could use calipers to measure additional dimensions of your connector. Then use that guide to narrow down to candidates and then look for the full dimensions of those candidates.

FYI, a huge number of RF connectors are defined in US military standard MIL-STD-348B: https://quicksearch.dla.mil/qsDocDetails.aspx?ident_number=35726

You can also use the Huber+Suhner RF coax connectors catalog PDF, which has high-quality drawings of practically every RF connector on the intro page to each connector’s section.

(Also, though it’s not a complete guide to connector types, I also came across H+S’s RF Connector Guide, which goes into all the theory of RF connections, materials, measurement, etc. Looks to be an amazing resource! At around the middle, it does have some guidance on which connector types are used for what applications.)


There’s also the Amphenol RF website, e.g. the 1.0/2.3 page here: https://www.amphenolrf.com/rf-connectors/1-0-2-3-connectors.html

But measuring pixels on the image the male part seems like it would fit. However the drawing is missing the white plastic isolator.

Yeah, it is. :/ And  the H+S catalog section on them makes me think it’s not that after all, given that it’s designed to be an insert in big multipin connectors, apparently.

Is this a kind of connector one would use for 450mhz?

I’m sure you could. 450MHz* is not a particularly demanding frequency, and those connectors are rated from DC to 10GHz.

*I assume you mean MHz (megahertz) and not mHz (millihertz), and not milli-hecto-zepto-something.

Yes Mhz...  I'll check the catalogs. I was hoping someone would just know.

[W4PJB]

Who is the manufacturer and model of the amplifier? Could it be a Proxim connector? I cannot find any dimensions on this, but seems to be common in certain cell hardware.

[Fflint]

Who is the manufacturer and model of the amplifier? Could it be a Proxim connector? I cannot find any dimensions on this, but seems to be common in certain cell hardware.

No idea of the model, these are just some PCBs pulled from an old CDMA cell tower made by Ericsson. There are various markings including "TVA 117 2906 R4" on one of the older boards with Ericsson name. And "TVK 119 3266 R2" on newer boards (they are exact same circuit) marked "Radiosystems Sweden AB".

One also has a paper sticker with:
ROA 1193151/1
R1K  A6195W11
A610095H0K

That is all I know..

[Fraser]

I do not know how many of these connectors are involved in your situation, but could you not remove the mystery connector and fit a more common connector type. Trying to find the original connector can be challenging and expensive. Just because the manufacturer used that weird RF connector does not mean you have to 

Fraser

[Fflint]

I do not know how many of these connectors are involved in your situation, but could you not remove the mystery connector and fit a more common connector type. Trying to find the original connector can be challenging and expensive. Just because the manufacturer used that weird RF connector does not mean you have to 

Fraser

There is 4. I was hoping this is some variation of mmcx or mcx that I can buy. There are so many rf adapters available. It is not the case of some 50 year old soviet hardware where they designed incompatible connectors on purpose (I have quite a few of these devices, but I was lucky enough to be able to buy mating connectors everywhere I need it). It is Ericsson, so if expect them to use something standard.

Can I replace them? Possibly, although I expect the hole pattern to also not match anything standard exactly. Instead I'll probably just machine down 3 more MCX male connectors to fit as its just the outside diameter that needs to have 0.2mm taken off the diameter.

It's quite funny that it's easier to buy some 50 year old Soviet RF connector, or the pcb mounted RF connector used in Kenwood ham gear made in 1970s and nowhere else since -that was a subject of a similar thread I posted some time ago). Than this 1990s made mass manufactured RF hardware made by Ericsson.

[tooki]

Who is the manufacturer and model of the amplifier? Could it be a Proxim connector? I cannot find any dimensions on this, but seems to be common in certain cell hardware.

Interestingly, the second Google search result for "Proxim connector" shows the picture you posted, and right in the search result text preview, quotes from the page: "The coaxial connector (SMB type) is housed in a protected screw cap." But it sure doesn't look anything like SMB to me.

[tooki]

I have these 460MHz old CDMA amplifiers that I retuned to 435MHz (70cm ham band). They have these weird connectors that look exactly like MCX but the male part has to be 3.5mm rather than 3.75mm that normal MCX is.

This is the board (female) side. Looks exactly like MCX but it is 3.5mm inside.
...

OK, so I decided to actually look at the MCX interface dimensions, as opposed to taking your word for it, and I'm now 99.9999% certain that what you have is standard MCX.

The MCX interface specifies an internal diameter for the jack of 3.42-3.48 mm. The outer diameter of the plug is specified as 3.72-3.80 mm. This agrees 100% with your own measurements, allowing for 0.02 mm measurement error on the jack.

See the attached screenshot from the Huber+Suhner catalog I linked the other day. The jack internal diameter is dimension H, the plug external diameter is dimension A.

The problem, then, is one of two things:
1) Your plugs are insufficiently slotted to allow compression, have a really poor surface finish that makes them bind, or are otherwise noncompliant
OR
2) You are afraid to use sufficient force when mating them

[tooki]

And here's Radiall's drawing, from their 2024 full catalog (first download link on https://www.radiall.com/document-library/rf-coaxial-connectors.html )

They don't state a minimum outer diameter for the plug, otherwise the same as H+S.

[Fflint]

So the conclusion seems to be MCX plugs commonly available that do mate with pretty much any modern made "consumer grade" MCX I have (mostly DVB sticks, LTE modems etc) are a "Chinesium" standard that doesn't actually mate with any MCX made by a reputable company like Hubner-Suhner, etc? Hilarious! Every device I've ever owned with a MCX jack up to this point was made in China so I never realised :-)

The thing with these drawings I just noticed is that they show the plug outer diameter dimension on the petals and not on the cylindrical part of the diameter. The plugs I have are all 3.7mm min on the cylindrical part and even bigger on the petals (I haven't measured there, I will later).

The Randall page shown on the post above shows something like a thinning of the cylindrical section and this is marked as diameter H at 3.6mm max. The plugs I have do not have this thinned out section and are all 3.7 to 3.75 there. The slots are absolutely tiny. I haven't measured but by the eye they look like 0.05mm at most. Sadly no drawing shows what the width of slots should be or how many there are.

If we allow the petals to flex in and say 3.6mm cylindrical section has to compress to 3.48mm that means in the original plug the slots have to be 0.4mm collectively (roughly). So 4 slots of 0.1mm would make sense probably.

Now if we take a Chinesium plug with cylindrical diameter of 3.7mm (sometimes even more) that would gave to have slots that are collectively 0.7mm wide to compress to 3.48mm. It has 3 slots, so 0.23mm each? They are not even close to being that wide.

I'll be ordering more MCX plugs to see if any made in China are different, but I'm not holding my breath. I'm not spending $20 on an MCX plug (my assumption) by hubner-suhner etc, but I'm happy to modify my Chinesium brand plugs if I can't get in-spec ones.

Measuring such a tiny internal diameter precisely is very difficult, but I'd love to measure the Chinese MCX jacks to see what they are. I have some measuring rods. I should definitely have 3.5mm. If it fits loosely into a Chinese made MCX jack we have our answer.

Thank you for your help in getting to the bottom of this.

Now I can imagine a guy who spent "good money" on quality MCX plugs to make an antenna extension for his LTE modem and his plugs fall out because the Jack's are too big... Or they are held by the central pin and disconnect with any slight vibration. Oh, my.

[ftg]

While it is only a small anecdotal data point, all of the MCX patch cables I have pulled from old base station hardware or scrap test equipment have mated fine with even the cheapest RTL-SDR sticks I have managed the buy.

The same cannot be said about the plugs that the included antennas had.
Some of them did not mate properly with any MCX and many were overmolded and did not plug in deep enough for a reliable connection.
But the included "DVB-T" antennas usually were complete bottom of the barrel garbage.
Some did not even have coax and instead just had normal unshielded cable with two wires.

[Co6aka]

normal unshielded cable with two wires

"Twinlead"    At least it's "balanced" rather than lopsided.

[tooki]

So the conclusion seems to be MCX plugs commonly available that do mate with pretty much any modern made "consumer grade" MCX I have (mostly DVB sticks, LTE modems etc) are a "Chinesium" standard that doesn't actually mate with any MCX made by a reputable company like Hubner-Suhner, etc? Hilarious! Every device I've ever owned with a MCX jack up to this point was made in China so I never realised :-)

Don’t confuse simple noncompliance with a “standard”. (Though the existence of a Chinese standard that’s undocumented in the West is of course possible.)

Also it’s Huber+Suhner, not Hubner-Suhner. I love their products, but damn they’re expensive. There are other coax connector brands that are just as good, but I’ve encountered none that are superior. (Though I will complain that their MMCX connectors require an annoying assembly process that requires stripping the jacket, then partial assembly before stripping the dielectric…) I also adore their RADOX 125 polyolefin-insulated wire.

The thing with these drawings I just noticed is that they show the plug outer diameter dimension on the petals and not on the cylindrical part of the diameter. The plugs I have are all 3.7mm min on the cylindrical part and even bigger on the petals (I haven't measured there, I will later).

The Randall page shown on the post above shows something like a thinning of the cylindrical section and this is marked as diameter H at 3.6mm max. The plugs I have do not have this thinned out section and are all 3.7 to 3.75 there. The slots are absolutely tiny. I haven't measured but by the eye they look like 0.05mm at most. Sadly no drawing shows what the width of slots should be or how many there are.

If we allow the petals to flex in and say 3.6mm cylindrical section has to compress to 3.48mm that means in the original plug the slots have to be 0.4mm collectively (roughly). So 4 slots of 0.1mm would make sense probably.

Now if we take a Chinesium plug with cylindrical diameter of 3.7mm (sometimes even more) that would gave to have slots that are collectively 0.7mm wide to compress to 3.48mm. It has 3 slots, so 0.23mm each? They are not even close to being that wide.

I'll be ordering more MCX plugs to see if any made in China are different, but I'm not holding my breath. I'm not spending $20 on an MCX plug (my assumption) by hubner-suhner etc, but I'm happy to modify my Chinesium brand plugs if I can't get in-spec ones.

Measuring such a tiny internal diameter precisely is very difficult, but I'd love to measure the Chinese MCX jacks to see what they are. I have some measuring rods. I should definitely have 3.5mm. If it fits loosely into a Chinese made MCX jack we have our answer.

Thank you for your help in getting to the bottom of this.

Now I can imagine a guy who spent "good money" on quality MCX plugs to make an antenna extension for his LTE modem and his plugs fall out because the Jack's are too big... Or they are held by the central pin and disconnect with any slight vibration. Oh, my.

Assuming that your measurements are correct (which there’s no reason to doubt), then those Chinese plugs are woefully out of spec. It just dawned on me: your plug shown in the very first picture appears to be nickel-plated. Is it possible that they machined the plug to final dimensions (and more specifically, to the maximum dimensions, maybe even a bit beyond) before plating and then plated on such a thick plating that it is now decidedly out of spec? 

(This actually makes me wonder: do RF connector manufacturers have to machine to different dimensions depending on the plating they intend to use? I assume gold and silver are thinner platings than nickel.)



It’s possible that the actual standards documents for MCX make some statement about mating and retention forces, and leave it up to the manufacturer to decide how to slit the barrel to achieve this. Unlike all the RF connectors that are specified in U.S. military standards, and thus freely available, I don’t think the full MCX standard is available for free.)



I don’t think Radiall dimension “H” means the thickness behind the retention rib: the dimension lines don’t extend that far back. So I think they’re just referring to the taper that must exist on the mating end, to guide it into the jack.


Remember that if you have a 3.5mm gauge, it should actually not fit in a standards-compliant jack.



There are plenty of other brands to try before spending $$$ on Huber+Suhner, Radiall, Hirose, etc. Amphenol, Molex, Samtec, Telegärtner, TE, Cinch, … (The only brand available on Digikey that I would avoid is Adam Tech.)

What I think is safe to assume is that Ericsson used a top-quality $$$ connector. So it will be made to spec.