Don't buy SMA cable assemblies from CentricRF

[matthuszagh]

TLDR: Gage the connectors of all SMA cable assemblies upon receipt, even from reputable manufacturers. Don't mate unverified cable assemblies with anything that is expensive or difficult to replace. "Reputable" manufacturers are selling SMA cable assemblies with pin protrusion. This practice seems to be quite common in my admittedly limited sample. I'm directly calling out CentricRF (Centric RF) in the title since I had some discouraging communication with them (more on this below), but I've also experienced the same problem with Mini-Circuits and Crystek.

Recently, I got around to using my Maury A027 kit to gage pin depth on a number of SMA cable assemblies I have. The results were discouraging. I have a number of C589-086 SMA cable assemblies from CentricRF, as well as FL141-12SM+ assemblies from Mini-Circuits, and 1 crystek ccsma18-mm-086f-24. I never bought cheap SMA cable assemblies from Amazon or alibaba believing that pin protrusion was confined to cheap cables. But, this issue seems to be similarly prevalent in $20-$60 cables from brand names. The Crystek and 3/4 minicircuits cables showed pin protrusion on at least one connector (up to about 4 mil). 6/8 centricrf cables showed pin protrusion on at least one connector. Below are all measurements for the CentricRF cables. I verified that my A027 gage was giving correct results by comparing several measurements against a much more accurate 3.5mm gage from an HP 85052B calibration kit. A negative reading indicates protrusion. All units in mil.

C589-086-48: -0.5, +2.0
C589-086-12: -1.5, +2.0
C589-086-36B: -1.0, +2.0
C589-086-36B: -3.0, 0.0
C589-086-60: -2.0, -2.0
C589-086-72: -1.0, -1.0

All of these cables are pretty lightly used. Some were mated just once or twice (like the crystek) and others just in the 10s of mating cycles (minicircuits and some of the centricrf cables). I've also measured the same pin protrusion on new cables from centricrf (more on that below), so I don't think wear explains these results.

I also gaged some very high quality cables I have like Gore phaseflex and H+S sucoflex, and unsurprisingly they're all spot on what they should be. These were all bought used and probably experienced many more mating cycles than the SMA cable assemblies mentioned above.

I think that manufacturers don't care about pin depth of their cheaper cable assemblies. That doesn't stop them from claiming conformance to MIL-STD-348. In my opinion this is dishonest (probably through negligence). It's difficult to imagine these manufacturers don't know to measure pin depth (minicircuits even sells gages). I do have 2 more expensive minicircuits cables (both ULC-1.5FT-SMSM+) and neither showed pin protrusion, though the pin recession was pretty bad for the price (in the range of 5-7mil).

Fortunately, I haven't seen the same problem with SMA adapters - all of the ones I measured (most of mine come from centricrf and minicircuits) were in spec.

I thought I'd give some further details of my interaction with centricrf, since it seems pretty clear that they're knowingly selling cables out of spec and don't care. I reached out to them with my measurements. They didn't want to refund me. In some sense that's understanble, since these cables were all purchased long enough ago to be out of warranty. But, it doesn't change the fact that they're claiming accordance to a spec they're not meeting. Additionally, the CEO (Bob Hawkins) provided a lot of bogus excuses for my measurements. Even though I mentioned that some of these cables were only mated 10 or so times, he said the problem was wear and SMA cable assemblies should only be used once in an application and then left as is. Such an explanation might be applicable to formable cables (though not to the connectors), but these are a very flexible 086 diameter cable with a low min bend radius (ie not formable cables). If these aren't meant to be flexed and mated/demated, what is? Also, if this were true, why would cable mount connectors typically be rated to 500 mating cycles? In any event, after I pointed out these problems with their arguments, they agreed to replace the cables. But, the replacements (new cables) also showed pin protrusion. This was confusing to me. Obviously I was going to gage them so why waste time and money sending replacements that I know are bad and can't even use?

Anyway, I hope people can read this post and avoid wasting money like I did with these cables. Some takeaways: (1) don't assume because you buy an SMA cable assembly from digikey/mouser or a reputable manufacturer that it's to spec, (2) if you do buy cables like these, gage them on receipt so you can send them back for a refund or replacement in case of issue (i wasted quite a bit of money on these). My plan for simple SMA/BNC/N/etc patch cables is not to buy these anymore but to make them myself. This takes a bit of investment in equipment, but I think is worth it. When I need higher performance cables, I buy high quality (eg gore phaseflex, H+S sucoflex) 2nd hand. If you're patient, you can find some reasonable deals.

[KE5FX]

That is not good to hear, re: Mini-Circuits.  We use a LOT of their 086 hand-flex cables.

[joeqsmith]

I wish there was a database available covering a wide selection of cables and adapters. 

I looked around to see if there were any videos about how to properly gage connectors.  With as user dependent as these measurements are I was expecting to find some training videos from Maury, but there doesn't seem to be much out there.   R&S has a basic connector care video that mentions gageing, but nothing in-depth.   Signal Path had a video on cables but skimming it, never appears to gage anything.

Not trying to damage your ego but I assume that you are using the correct gage and master for the connectors you are testing.  Maybe verified these bad cables with a different operator and gage set before posting?  Had you made a video of the entire process, I would have watched it.  Maybe consider putting something out. 

[matthuszagh]

That's ok - ego not damaged.

I did use two different gage kits for these measurements and that information was in the original post, though I understand how that could have been missed within the large block of text. I used both the Maury A027 SMA push-on gages and the (much more accurate) 3.5mm thread-on gages from an HP 85052B calibration kit. Both of my gages are designed to measure pin protrusion/recession. I do not have the Maury kit to measure dielectric protrusion/recession and (of course) this would not be applicable for 3.5mm connectors. I would like to be able to measure dielectric recession. However, based on the way SMA connectors are attached to cables, the position of the dielectric is already fixed and I expect dielectric protrusion is therefore less likely. Additionally, this is just a guess but since the dielectric is directly adjacent to the reference plane, I expect you might actually be able to determine recession or protrusion by slowly sweeping something with a fine point over the boundary (of course, be careful of scratching the reference plane). I've never tried that though.

Push-on gages are quick and convenient, but there is some play in the mating and so there is some change in the reading depending on how you tilt the gage and SMA connector relative to one another. In my experience this typically results in a range of about 1.5 mil or so. The manual says to engage the connector straight on. However, when the center pin of a connector is nearly aligned the reference plane but initially in recession, and I'm able to get it into protrusion by tilting the mating with the gage, I often find that checking the connector on the thread-on gage shows that it is indeed slightly in protrusion. If a reading is far from a specification boundary and you don't care what the actual reading is (just that it's in spec) this is good enough. If it's close, I then also use the thread-on gage and decide based on that. The manual for the HP 85052B says to perform 3 readings as well as other things like tapping the gage to get it to settle, but in my experience this gage is very repeatable and those things barely affect the reading.

To more directly address your concerns. I checked that the readings from the two gages were consistent (at least to within the range of play from the push-on gage) on a number of different connectors and I used the correct master setting gage (and this is obvious).

As for instructions of use, the manuals are sufficiently detailed in my opinion. I've attached both, in case you want to peruse. If you still think it would be useful I can investigate making a video, but I'm not too experienced with doing this.

[joeqsmith]

I wondered if these cables you measure have a stepped contact?  You stated you have A027 so I assume you have the FP and MP gages with the A027T1 master.  But not the MC gage and A027T2 mater.   If you use the MP gage on a stepless contact, does the gage measure from the dielectric? 

I had read about the other gage set but it seemed that you were doing the work.  This is why I asked if you had an independent person verify your results with a different gage set,  prior to making anything public. 

Did you return the suspect cables?  If so, did they measure them in-house and report back?   Just curious. 

[mark432]

"I wish there was a database available covering a wide selection of cables and adapters. "

Take a look at MIL-STD-348.
That has your key interface dimensions for 90% of common connectors.
http://everyspec.com/MIL-STD/MIL-STD-0300-0499/MIL-STD-348A_420/

There are also "precision" grades of some of those connectors, but you'll know if you have them because you paid 10X normal.
They might take a spec that is typically 20 mils tolerance and bring it down to 2 mils.

Improper pin depth is an easy problem to run into.  I've even seen it happen from thermal cycling.  (With cheap connectors)

[joeqsmith]

"I wish there was a database available covering a wide selection of cables and adapters. "

Take a look at MIL-STD-348.
That has your key interface dimensions for 90% of common connectors.
http://everyspec.com/MIL-STD/MIL-STD-0300-0499/MIL-STD-348A_420/

There are also "precision" grades of some of those connectors, but you'll know if you have them because you paid 10X normal.
They might take a spec that is typically 20 mils tolerance and bring it down to 2 mils.

Improper pin depth is an easy problem to run into.  I've even seen it happen from thermal cycling.  (With cheap connectors)

I don't think I stated that well.  I would be interested in database of cables where independent measurements were made and published. 

[matthuszagh]

I wondered if these cables you measure have a stepped contact?  You stated you have A027 so I assume you have the FP and MP gages with the A027T1 master.  But not the MC gage and A027T2 mater.   If you use the MP gage on a stepless contact, does the gage measure from the dielectric? 

I had read about the other gage set but it seemed that you were doing the work.  This is why I asked if you had an independent person verify your results with a different gage set,  prior to making anything public. 

Did you return the suspect cables?  If so, did they measure them in-house and report back?   Just curious.

The connectors I measured all have a stepped contact and that's what the A027 kit is designed to measure.

I'm the only one who performed these measurements. But, in addition to the fact that the procedure is fairly straightforward, I received a number of independent results that indicate I performed this correctly. These were all listed in the original post, but to reiterate some of them: (1) agreement between both gages, (2) all high quality Gore phaseflex and H+S sucoflex cables I have met the tighter 0 to 3 mil center pin recession requirement, (3) all adapters with an SMA male connector measured within the 0 to 10 mil pin recession requirement (and I probably have like 20 adapters or so).

Also, it seems pretty clear the company never gaged the center pin on these cables. They reported several dielectric recession measurements they had on file but never gave me any pin recession measurements. They never admitted not having gaged the center pins but never actually denied it either. But to your question, even when they sent me replacement cables (some of which ended up being bad), they never provided me measurements of those either. And no I never got a refund. I was out of warranty and they seemed pretty pissed at me that I was making any sort of a fuss about this (apparently, I should pay $1000 for each cable if I want this, or something like that). This whole thing was made more difficult by the fact that I was well out of warranty for all of these. I reached out to digikey too about the crystek cable and they just told me it was out of warranty and did not seem concerned about the pin protrusion (even though I used that cable probably just once and the protrusion was not due to wear).

I think it's pretty unlikely measurement error is the problem here. I have a hard time coming up with a self-consistent explanation of how that might work.

It seems more likely the mentioned companies (and probably others) don't bother to gage their cheaper cable assemblies and no one is checking them on it. Or, if they do, clients complain and the company sends replacements but doesn't bother to change the process. It's probably cheaper to send the occasional replacement cable than to improve the process and validate cables more thoroughly before selling them. Who cares about being honest, apparently. The aerospace company I work for buys mini-circuits cables, but doesn't gage them. I expect we're not the only ones. That's just a guess of course. But, how many people gage $40 cables? I imagine most of the people gaging cables are in a higher precision metrology-type setting and buying much more expensive cables.

I feel ok with the fact that I didn't seek a second opinion before outing these companies. But, I have the bad cables here and if someone wants to pay for one to be shipped to them to gage it (if they have the capabilities to do so) I'd probably agree to that.

[joeqsmith]

...all have a stepped contact...

Have you ran into stepless contacts with any of your cables? 

Below, I assume part number of the cable followed by two measurements.  MP , MS gage? Just two measurement with same gage?  The Maury vs HP gage?  What ever these are, is it a single measurement or did you take say five measurements, then averaged?  Just curious of the technique you are using.  Maybe these two number are the max min of 10 readings for each cable?

C589-086-48: -0.5, +2.0
C589-086-12: -1.5, +2.0
C589-086-36B: -1.0, +2.0
C589-086-36B: -3.0, 0.0
C589-086-60: -2.0, -2.0
C589-086-72: -1.0, -1.0

[joeqsmith]

I tried a search to get some context and it seems like gageing connectors may be new to you.   

https://www.eevblog.com/forum/testgear/pin-protrusion-in-commercial-sma-cable-assemblies/
https://www.eevblog.com/forum/repair/how-to-clean-an-sma-gage-kit/
https://www.eevblog.com/forum/metrology/difference-between-push-on-and-thread-on-microwave-connector-gages/


I'm sure you have dug into all of this before dragging these companies through the mud, and not knowing exactly what you are doing,  I am left with questions....

Your MP gage measures from the outer conductor mating plane to the flat surface of the step? 

Today there are wide variations in the SMA designs. I categorize them as the "stepped", "stepless" and  "angled". 

If the step is not a flat surface, the reading will depend in the gage bore and where it touches off on the angle of the step? 

Does the MC gage measure from the outer conductor to the tip of the pin?  Avoiding the variations in the step or angled step?   

Looking at some cables I have, their construction is all over the place.  You have never shown any photos of the pin style used in the cables you tested.     

I could not locate a mechanical drawing for the MP gage.  I assume you contacted Maury about which gage would be suitable for the connectors you are trying to measure.   What are the disadvantages of using the MC gage? 

I also found this 25 year old application note from Maury Microwave that talks about the SMA and 3.5mm.   

While I was not able to locate any decent best practices manuals on proper measurement techniques of RF connectors, it is a mechanical measurement.  I did find a good machinist's manual.  It covers the basics of temperature, storage, cleaning, use of the master gages.... 

[matthuszagh]

Have you ran into stepless contacts with any of your cables? 

No, all of my cables have a stepped center contact. In fact, I'm not even sure I've ever actually seen a stepless center contact, though I was aware of their existence. In any event, stepped center contacts seem to be much more common.

Below, I assume part number of the cable followed by two measurements.  MP , MS gage? Just two measurement with same gage?  The Maury vs HP gage?  What ever these are, is it a single measurement or did you take say five measurements, then averaged?  Just curious of the technique you are using.  Maybe these two number are the max min of 10 readings for each cable?

C589-086-48: -0.5, +2.0
C589-086-12: -1.5, +2.0
C589-086-36B: -1.0, +2.0
C589-086-36B: -3.0, 0.0
C589-086-60: -2.0, -2.0
C589-086-72: -1.0, -1.0

Each cable assembly has two connectors. The list above gives one measurement result for each connector of the assembly.

[matthuszagh]

I tried a search to get some context and it seems like gageing connectors may be new to you.   

https://www.eevblog.com/forum/testgear/pin-protrusion-in-commercial-sma-cable-assemblies/
https://www.eevblog.com/forum/repair/how-to-clean-an-sma-gage-kit/
https://www.eevblog.com/forum/metrology/difference-between-push-on-and-thread-on-microwave-connector-gages/

The first link is basically a duplicate of this thread. I was trying to determine if other people had also found pin protrusion to be very common.

The 2nd link is from a gage kit I ended up returning because I didn't feel that I would be able to clean it properly such that I could achieve reliable results. The kit I have has no such degraded foam on the mating surfaces to remove, since both gages and the gage master were protected by plastic coverings.

The 3rd link is from over 2 years ago and the answer posted there is in agreement with what I've seen having both types of gages.

Your MP gage measures from the outer conductor mating plane to the flat surface of the step?

Correct.

Today there are wide variations in the SMA designs. I categorize them as the "stepped", "stepless" and  "angled". 

If the step is not a flat surface, the reading will depend in the gage bore and where it touches off on the angle of the step? 

As mentioned all of the connectors I gaged are of the stepped variety. I was not aware of angled steps. Can you provide some further information about this? In any event, both the centricrf and mini-circuits cables state conformance to MIL-STD-348 in their respective datasheets. Unless I've missed something, that standard only permits for male SMA connectors with a flat stepped contact and with no contact, but not stepless or angled steps. Also, they don't look like angled steps, though if the angle grade were small I suppose it's possible I wouldn't be able to determine this by sight. But, all SMA connector variants need to be mechanically-compatible with all other SMA connectors as well as 3.5 mm and 2.92 mm connectors. Unless I'm mistaken, this does not allow for pin protrusion on any part of an angled step. So, measuring pin protrusion on an angled step would still indicate a non-comforming connector. Still, unless you can provide standards to explain this angled step variant (and even the stepless variant), these sound like non-standard configurations. Again, the companies I mentioned claim adherence to MIL-348.

Does the MC gage measure from the outer conductor to the tip of the pin?  Avoiding the variations in the step or angled step?   

I don't know. I don't have an MC gage and have never used one.

Looking at some cables I have, their construction is all over the place.  You have never shown any photos of the pin style used in the cables you tested.     

That may be true for your cable assemblies, but not for mine. All of my SMA connectors on cable assemblies and adapters are stepped and as far as I can tell are not angled.

I provided part numbers and you can probably find photos from that. Anyway they're stepped, do I really need to provide you with photos?

I could not locate a mechanical drawing for the MP gage.  I assume you contacted Maury about which gage would be suitable for the connectors you are trying to measure.   What are the disadvantages of using the MC gage? 

The MC gage is designed for a stepless center contact, whereas the MP gage is designed for a stepped center contact. This is all clearly stated in the A027 manual. What precisely do you want me to ask of Maury? It seems like all the necessary information is already available in the manual.

I also found this 25 year old application note from Maury Microwave that talks about the SMA and 3.5mm.   

While I was not able to locate any decent best practices manuals on proper measurement techniques of RF connectors, it is a mechanical measurement.  I did find a good machinist's manual.  It covers the basics of temperature, storage, cleaning, use of the master gages....

I appreciate these resources.

I have trouble believing temperature can account for the 5 mil in protrusion I've measured on some connectors. The datasheets for the mini-circuits and centricrf cables state an operating temperature range of -55C to 105/125C. Moreover, MIL-348 does not state a temperature at which the mechanical requirements apply, as far as I can tell. I believe that means that the connectors must adhere to the stated mechanical specifications over the provided operating temperature range. I performed these measurements in a room kept at normal room temperature. If they're not designed to have the correct mechanical specifications at room temperature, at what temperature do you think they are supposed to be correct?

I'm in general happy to continue this discussion, and you've provided some useful resources, which I appreciate. However, I don't appreciate the implications that I've unjustifiably sullied the good names of companies. If you're not convinced that's your prerogative, but I think the evidence seems pretty clear. I followed the manuals for my gages, used the gages on connectors for which they were designed, and checked the measurements on a second gage set before "I dragged names through the mud". But, no, I didn't pay NIST to ensure all my measurements were accurate to within fractions of a micron before I posted this. I don't think that's necessary. Anyway, I think my claims are justified. I created this post to help people who do not have gage capabilities. You're free to believe whatever you'd like.

[joeqsmith]

Today there are wide variations in the SMA designs. I categorize them as the "stepped", "stepless" and  "angled". 

If the step is not a flat surface, the reading will depend in the gage bore and where it touches off on the angle of the step? 

As mentioned all of the connectors I gaged are of the stepped variety. I was not aware of angled steps. Can you provide some further information about this? In any event, both the centricrf and mini-circuits cables state conformance to MIL-STD-348 in their respective datasheets. Unless I've missed something, that standard only permits for male SMA connectors with a flat stepped contact and with no contact, but not stepless or angled steps. Also, they don't look like angled steps, though if the angle grade were small I suppose it's possible I wouldn't be able to determine this by sight. But, all SMA connector variants need to be mechanically-compatible with all other SMA connectors as well as 3.5 mm and 2.92 mm connectors. Unless I'm mistaken, this does not allow for pin protrusion on any part of an angled step. So, measuring pin protrusion on an angled step would still indicate a non-comforming connector. Still, unless you can provide standards to explain this angled step variant (and even the stepless variant), these sound like non-standard configurations. Again, the companies I mentioned claim adherence to MIL-348.

I really don't know.  This is why I am asking.  The MIL-STD that was previously linked is A, dating back to 1988.  I would need to start with current specs.   It is a bit hard to believe they would not allow for any hardline.  This type of coax is used in a lot of equipment.  The coax center conductor is formed into the male pin.  Steps cause reflections.  Maybe back in 1988, it wasn't much of a problem.   

*** RevB
https://landandmaritimeapps.dla.mil/Downloads/MilSpec/Docs/MIL-STD-348/std348.pdf
***

Looking at some cables I have, their construction is all over the place.  You have never shown any photos of the pin style used in the cables you tested.     

That may be true for your cable assemblies, but not for mine. All of my SMA connectors on cable assemblies and adapters are stepped and as far as I can tell are not angled.

I provided part numbers and you can probably find photos from that. Anyway they're stepped, do I really need to provide you with photos?

I assumed with them being bad, you would pull them apart and post a side profile of the pin.   If they are bad, there's not much reason not to dissect them. 

I did go to their site and look up the PNs but they don't supply any details on the connectors used.   Looking at several part prints, I have not found any that show the dimensions of their pins.   Nor could I find a dimensional drawing for your gage.  Worse, it appears the distance from the start of the step to the tip of the pin varies.

So I really have no way to know if the gage would work with all the countless stepped pins out there.  My first step would be to call Maury and ask about compatibility with all these new pins that are out there.

I could not locate a mechanical drawing for the MP gage.  I assume you contacted Maury about which gage would be suitable for the connectors you are trying to measure.   What are the disadvantages of using the MC gage? 

The MC gage is designed for a stepless center contact, whereas the MP gage is designed for a stepped center contact. This is all clearly stated in the A027 manual. What precisely do you want me to ask of Maury? It seems like all the necessary information is already available in the manual.

Why can't the stepless gage be used across the board?  If it is from the outer contact to the tip, this is really all you want to know?

I appreciate these resources.

I have trouble believing temperature can account for the 5 mil in protrusion I've measured on some connectors. The datasheets for the mini-circuits and centricrf cables state an operating temperature range of -55C to 105/125C. Moreover, MIL-348 does not state a temperature at which the mechanical requirements apply, as far as I can tell. I believe that means that the connectors must adhere to the stated mechanical specifications over the provided operating temperature range. I performed these measurements in a room kept at normal room temperature. If they're not designed to have the correct mechanical specifications at room temperature, at what temperature do you think they are supposed to be correct?

That document was just the basics and covered a lot of facets.   I am not suggesting temperature is the cause of what you are seeing.  Actually, I would think the main sources of error would be poor cleaning, handling, wear. 

I'm in general happy to continue this discussion, and you've provided some useful resources, which I appreciate. However, I don't appreciate the implications that I've unjustifiably sullied the good names of companies. If you're not convinced that's your prerogative, but I think the evidence seems pretty clear. I followed the manuals for my gages, used the gages on connectors for which they were designed, and checked the measurements on a second gage set before "I dragged names through the mud". But, no, I didn't pay NIST to ensure all my measurements were accurate to within fractions of a micron before I posted this. I don't think that's necessary. Anyway, I think my claims are justified. I created this post to help people who do not have gage capabilities. You're free to believe whatever you'd like.

I tend to be data driven.  Beliefs and feelings go out the window.    You ran the tests and made the results public.  I am just asking what I considered some basic questions as I assumed you had dug into this rabbit hole.     

[joeqsmith]

Johnson® Stainless Steel SMA Connectors meet or exceed the performance requirements of MIL-PRF-39012.  All designs are based on 50 ohm system impedance per MIL-STD-348, and operate at frequencies up to 26.5 GHz

Straight Solder Type Plug - without Contact, Thread-on Nut
Connector Interface for .141 Cable

Obviously, no step but does it actually conform to MIL-STD-348? 

https://www.mouser.com/datasheet/2/643/pi-CCS-JOHN-141-0701-612-1290055.pdf

[joeqsmith]

Document showing their semi-rigid connector.  No step but appear to conform.   
https://www.mouser.com/datasheet/2/643/pi-CCS-JOHN-142-0593-421-1290159.pdf

Does the standard require that the connectors have a step, or that they mate with the drawings shown?   I have no idea, which is why I am asking.   

[Smoky]

If it hasn't been mentioned before, I saved this article a while back about making SMA cables. It explains the assembly of semi-rigid cables with good pictures using the proper tools.

The author mentions that a quick-check that the shoulder of the center contact should be flush with connector's dielectric when looking into the business end when it's finished as an indication of a proper assembly.

[Smoky]

 Even though I mentioned that some of these cables were only mated 10 or so times, he said the problem was wear and SMA cable assemblies should only be used once in an application and then left as is.
 

I've been reading the many responses and watching the videos that joeqsmith has linked to his SMA/VNA topics.

He mentions one or more times not to allow the center contact to spin when tightening the coupling nut.

I think that this action is the most important to avoid. How many twists of the center contact will eventually remove the plating from the mating connector 

Now imagine that the center contact is protruding at the same time.

It could be possible that the best connection that the SMA plug ever makes is the first one.

[shabaz]

Unfortunately I'm speculating here, but it seems a bit much to expect these cables to remain at the correct pin depth.
Those flexible 0.086" cables re-purpose RG-405 semi-rigid connectors from what I can tell (at least, that's what I intend to do; I have purchased some RG-405 connectors for that purpose, but need to obtain a solder bath/pot (needed to semi-rigid-ize the cable on the end). The pin depth is set at time of assembly with some of those SMA connectors, i.e. seems a friction fit. As someone who knows nothing in this area, I'm curious, could the pin maybe move slightly when using the cable, maybe if the coax bends many times? Or have enough space for movement when the connector is inserted/removed many times?

There are very high-end SMA connectors, some of which have a screw clamping system designed such that I believe won't allow the pin to move at all (unless not screwed down properly). Example is SV Microwave example: SMA male connector for RG178 and the assembly instructions do not mention having to set pin depth (unless I missed it). But I'm sure they will be exceedingly expensive.

Another option of course is right-angle SMA connectors, since they will have the pin depth permanently fixed.

I've never measured SMA connectors, I don't have any critical requirements for that, and use adapters on my most expensive equipment, so I have never investigated this topic deeper unfortunately, but am interested in this thread.

[Smoky]

Full disclosure, I am a total newb when it comes to SMA cables and I've never had a use for one before, but, and thanks to the sale that went on at Arrow over the past year, I own about 150 SMA cables now so, at the least, I describe myself currently as a serious collector

The pin depth is set at time of assembly with some of those SMA connectors, i.e. seems a friction fit. As someone who knows nothing in this area, I'm curious, could the pin maybe move slightly when using the cable, maybe if the coax bends many times? Or have enough space for movement when the connector is inserted/removed many times?

I have a few of the Amphenol .047/.085/.141 semi-rigid cables with "friction-fit" connectors as you describe. Since the body of the connector is soldered to the tin-soaked outer braid I can't imagine it stretching or compressing all that much to where it would affect the center contact. From what I've read, semi-rigid, hand formable, etc., types are not to be bent within a couple inches of the connector. Yet there are super-flexible variants that can be bent at a right angle directly after the fitting.

As for the "friction-fit", my guess is that galling takes effect on the gold or silver plating that lines the mating surfaces when it gets pushed in and helps to maintain a good contact even with slight movement

[matthuszagh]

I'm in general happy to continue this discussion, and you've provided some useful resources, which I appreciate. However, I don't appreciate the implications that I've unjustifiably sullied the good names of companies. If you're not convinced that's your prerogative, but I think the evidence seems pretty clear. I followed the manuals for my gages, used the gages on connectors for which they were designed, and checked the measurements on a second gage set before "I dragged names through the mud". But, no, I didn't pay NIST to ensure all my measurements were accurate to within fractions of a micron before I posted this. I don't think that's necessary. Anyway, I think my claims are justified. I created this post to help people who do not have gage capabilities. You're free to believe whatever you'd like.

I tend to be data driven.  Beliefs and feelings go out the window.    You ran the tests and made the results public.  I am just asking what I considered some basic questions as I assumed you had dug into this rabbit hole.     

Fair enough. I've done some digging but not exhaustive.

I really don't know.  This is why I am asking.  The MIL-STD that was previously linked is A, dating back to 1988.  I would need to start with current specs.   It is a bit hard to believe they would not allow for any hardline.  This type of coax is used in a lot of equipment.  The coax center conductor is formed into the male pin.  Steps cause reflections.  Maybe back in 1988, it wasn't much of a problem.   

Does MIL-STD-348 not allow for hardline? Can you point me to that? I was under the impression it only covered the interfaces, not the cables attached to the connectors. I took another look and couldn't find any sort of restrictions on cables. But I do have the latest -348 standard and have referenced that.

For what it's worth, based on MIL-STD-348 I still think a stepped center contact is the only "standard" configuration, so I don't see how someone could claim adherence to that standard and use an alternative.

I assumed with them being bad, you would pull them apart and post a side profile of the pin.   If they are bad, there's not much reason not to dissect them. 

I did go to their site and look up the PNs but they don't supply any details on the connectors used.   Looking at several part prints, I have not found any that show the dimensions of their pins.   Nor could I find a dimensional drawing for your gage.  Worse, it appears the distance from the start of the step to the tip of the pin varies.

So I really have no way to know if the gage would work with all the countless stepped pins out there.  My first step would be to call Maury and ask about compatibility with all these new pins that are out there.

Ah ok I see what you want. I haven't pulled one apart yet, but I did look at it (intact) under a microscope and it really appears to be a flat step. Maybe I'll pull one apart. If I do, I'll post pictures.

Why can't the stepless gage be used across the board?  If it is from the outer contact to the tip, this is really all you want to know?

Not entirely. There are a number of ways an out-of-spec connector could damage another connector. Confining the discussion to pin protrusion specifically, it's my understanding that there are two possible mechanisms of damage here. One, as I believe you're pointing out is the tip of the pin protruding beyond the tip/valley of the mating socket. This would place pressure on the socket at its base. The other (and what I believe is typically referred to for pin protrusion) is what I'm referring to and is the step extending beyond the mating plane. This would exert pressure on the socket at the tip of the female fingers. This could cause the fingers to splay or otherwise be pushed out of normal alignment. I believe this is a greater risk too. It's probably not too difficult to bend fingers, but if you're just pushing into the base of the socket, I think that would be less likely to do damage. If you look at the interface dimensions too, the socket is specified to be longer than the max length of the pin. The total pin excursion from the reference plane, including the chamfer can be no longer than 0.1in. The socket, excluding the chamfer, must be at least 5mil longer. So there's at least 5 mil of leeway there (more if you account for the female socket chamfer). By contrast, the pin step can be up to the reference plane on both the male and female connector. So, if the male step goes past the reference plane and the female connector is at the reference plane or behind but by less than the male step, they will bump up against each other before the reference planes mate and damage can occur.

In any event, that's a long-winded way of saying ensuring the tip of the pin does not protrude too far does not guarantee that the step does not protrude excessively.

That document was just the basics and covered a lot of facets.   I am not suggesting temperature is the cause of what you are seeing.  Actually, I would think the main sources of error would be poor cleaning, handling, wear. 

Yep I hear you. Cleaning/handling/wear can't account for issues with the cable assemblies because I got the same results on brand new assemblies. These had quite literally never been mated before I gaged them (unless the manufacturer mated them). Now this could explain results with the gage, but I have a microscope to inspect the gage surfaces and ensure they're clean and a dental air compressor to clean them along with fine-tip q-tips. Moreover, as mentioned I got expected results with other connectors so I think the gage is fine.

[matthuszagh]

Obviously, no step but does it actually conform to MIL-STD-348? 

https://www.mouser.com/datasheet/2/643/pi-CCS-JOHN-141-0701-612-1290055.pdf

Also referencing your other, related post.

Yeah it's a valid question. It's hard for me to understand how this adheres to -348 since -348 clearly shows a flat step. Now I suppose they could interpret -348 as saying there cannot be a step that is in protrusion, in which case no step would meet this spec and an angled step would meet this spec too as long as no part of the step passes the reference plane. That's just a guess though. If that's what people are doing, it still seems like a bit of a stretch to claim adherence to the spec. Or maybe the spec should acknowledge these variants and clarify what qualifies as adherence in these cases.

[matthuszagh]

He mentions one or more times not to allow the center contact to spin when tightening the coupling nut.

Yep, and this is critical for basically all coaxial connectors. The body should not rotate, only the coupling nut should rotate. Failure to adhere to this can definitely wear out mating surfaces. That's not my issue though, as I've known this for a long time and am quite careful.

[shabaz]

I own about 150 SMA cables now

Wow, nice! I've not encountered such a good deal, but have slowly acquired a fair few (nowhere near 150 currently!). I used to use SMB a lot more (because that was the main "mid-size" RF connector that was used at my workplace for a while; internal to equipment) but pretty much stick to SMA these days.

I believe/hope you'll find them extremely useful; they are a very friendly size to work with when prototyping, compared to the more bulky BNC, for instance, plus the sockets are very low-cost for quickly making connections onto a prototype. (Some very low-cost SMA sockets are available at LCSC; they are not top quality at that price, but they are great for general prototyping).

I've been meaning to write a sort of "SMA connectors: How to make use of them" blog for ages since they are great problem-solvers, and a friend suggested it would be useful since not everyone comes across that connector in their lines of work) Unfortunately, I have not gotten around to it so far. The main negative is that it can sometimes be awkward to build up a cable (since tools may be needed), but since you've got ready-made cables, that issue is gone! I, too, try to use ready-made cables if possible but end up occasionally DIY'ing some.

[Smoky]

Shabaz, here's one group of 5 cables I bought on sale from Arrow. These are 24" Amphenol ATC Phase Stable SMA cables (095-902-466-001):



They cost ~$28 each but their regular price is ~$200 each now!



This I think is Amphenols first generation of this cable (dtd 2017) because it used the original manufacturer Harbour strip-braid cable (SB142). Amphenol uses Times Microwave cable now. Even the drawing/datasheet is dated 2017. Arrow had a lot of old inventory on sale:



Get a load of this Matthuszagh, you mentioned a durability of 500 cycles for SMA cables. The datasheet reads that these ATC cables are rated for 5000 cycles... minimum



Look at the design requirements of this cable. One rating states VSWR of 1.27:1 is the max up to 20GHz.

You're not just getting a cable made with expensive parts, you're getting a cable that's fully tested too

[joeqsmith]

If it hasn't been mentioned before, I saved this article a while back about making SMA cables. It explains the assembly of semi-rigid cables with good pictures using the proper tools.

The author mentions that a quick-check that the shoulder of the center contact should be flush with connector's dielectric when looking into the business end when it's finished as an indication of a proper assembly.

That same group wrote a paper on proper care in the lab that is also pretty good.  Guessing a paper like this helped curb the damage caused by the students.