RF connectors smack down ... AliExpress vs Amphenol

[Mr. Scram]

Yep, but a reputable brand 1:1 is still VERY MUCH cheaper then sending a VNA back to Keysight/R&S/Copper Mountain to have the connectors replaced, and even a non preciosn one will save the wear on the connector.

Adding a hundred quid or so to the cost of a £50k instrument is hardly a big deal if it means you can avoid weeks of downtime while the thing is being repaired.

Regards, Dan.

Who here has a "£50k instrument" on his desk?

Where are the numbers, charts and other things that make this quantifiable? Right now it's just a bunch of people claiming things. Labs full of test gear and all we have to show for it are opinions?

[bd139]

TBH the performance difference is probably negligible. Initially. Which is the problem.

If you have shitty PL259’s you can watch your PA finals explode after a few tens of connection cycles which is evidence enough for the performance difference. Then your £80 of FETs works out more than an Amphenol or two.

[capt bullshot]

Do you consider a PL259 a proper RF connector? Afaik they're even worse than a BNC.

Otherwise, e.g with BNC connectors attached to RG174 one can see differences between these Radiall comression style gems and non-branded but not el cheapo crimping connectors on a Tek 1502.

[xaxaxa]

Return loss of an aliexpress sma connector, to 2.45GHz (I simply made a 50ohm termination by soldering two 100ohm resistors to the connector):


Looks good right? problem is I can make the graph jump just by nudging the connector a bit, and it will probably become all rusty after a month of use (because of fake gold plating).

[EVS]

AliExpress vs Amphenol

Just for comparison. Radiall (R161703000) vs Telegartner (J01008C0825) 

[taydin]

To keep it going

From top to bottom: Fairview, HP, Pasternack

[G0HZU]

Yep, but a reputable brand 1:1 is still VERY MUCH cheaper then sending a VNA back to Keysight/R&S/Copper Mountain to have the connectors replaced, and even a non preciosn one will save the wear on the connector.

Adding a hundred quid or so to the cost of a £50k instrument is hardly a big deal if it means you can avoid weeks of downtime while the thing is being repaired.

Regards, Dan.

Who here has a "£50k instrument" on his desk?

Where are the numbers, charts and other things that make this quantifiable? Right now it's just a bunch of people claiming things. Labs full of test gear and all we have to show for it are opinions?

If you mean >$50k when new I've got a few items that qualify.

I have to be very careful when introducing 'anything' into my workroom that has an SMA connector on it. I try and only use the very best brands of SMA connector (and cable) to minimise the risk of damage to my VNA connectors/cables and the Ecal module that comes with it. The Ecal uses Agilent/Keysight's PSC series 3.5mm connector and these are extremely fragile. I think the Ecal module cost £13k here in the UK and it can be trashed with a moment's carelessness with a cheapo connector that has a tolerance issue or any kind of surface defect on the centre pin. So I inspect any new SMA connectors/cables under a microscope before allowing them into the workroom. Even a tiny burr could cause problems.

You can see the Agilent PSC 3.5mm connector in the link below and how easy it is to damage it. I hate inspecting my Ecal kit for damage because it is quite stressful counting all the tiny petals within each PSC connector (4 connectors in total) to look for any defects.

For this reason I usually only use some very expensive Gore cables to connect to the Ecal module. A good Gore cable can easily cost $1k USD. But these cables can easily be damaged by items under test and pass on the damage to the Ecal module. So I try and do all my testing via known Suhner or Johnson SMA connectors.

Nearly all of my N type connectors/attenuators/cables are made by Suhner or Narda or HP. I have a few cheapo Amphenol ones but these only get used on cheaper test gear.

https://www.microwaves101.com/encyclopedias/how-to-not-trash-a-calibration-kit

[Mr. Scram]

If you mean >$50k when new I've got a few items that qualify.

I have to be very careful when introducing 'anything' into my workroom that has an SMA connector on it. I try and only use the very best brands of SMA connector (and cable) to minimise the risk of damage to my VNA connectors/cables and the Ecal module that comes with it. The Ecal uses Agilent/Keysight's PSC series 3.5mm connector and these are extremely fragile. I think the Ecal module cost £13k here in the UK and it can be trashed with a moment's carelessness with a cheapo connector that has a tolerance issue or any kind of surface defect on the centre pin. So I inspect any new SMA connectors/cables under a microscope before allowing them into the workroom. Even a tiny burr could cause problems.

You can see the Agilent PSC 3.5mm connector in the link below and how easy it is to damage it. I hate inspecting my Ecal kit for damage because it is quite stressful counting all the tiny petals within each PSC connector (4 connectors in total) to look for any defects.

For this reason I usually only use some very expensive Gore cables to connect to the Ecal module. A good Gore cable can easily cost $1k USD. But these cables can easily be damaged by items under test and pass on the damage to the Ecal module. So I try and do all my testing via known Suhner or Johnson SMA connectors.

Nearly all of my N type connectors/attenuators/cables are made by Suhner or Narda or HP. I have a few cheapo Amphenol ones but these only get used on cheaper test gear.

https://www.microwaves101.com/encyclopedias/how-to-not-trash-a-calibration-kit

We're talking about replacement value, so it's the current cost of the device we're talking about. A device worth a fraction those 50k isn't a 50k machine. That's like saying what's in the toilet bowl is a top notch dinner.

We were also taking about BNCs, but I haven't anyone actually show issues arising from cheap connectors. I'd love to know about any, as it would be information relevant to my setup and interests. I can understand people being careful with their treasured equipment and wanting to buy the best stuff to go with it, but there's a huge risk of audiophoolery type self deceit. The fact that measurable performance is apparently similar gives pause.

[G0HZU]

A device worth a fraction those 50k isn't a 50k machine.

On decent high grade lab gear I think the manufacturer will still charge huge money for even a basic repair. One of my analysers probably cost >$60k new ten years ago but it still has a minimum repair cost of $15k USD today. This is a fixed initial repair price set by the manufacturer. Also, the $$$ repair could take weeks which I think was the point made by dmills? So suddenly it makes sense to protect those exotic RF connectors rated to many GHz 

We were also taking about BNCs

The connectors in the OP look like N connectors to me?

The fact that measurable performance is apparently similar gives pause

It's the risk of damage and accelerated wear and the difficulty of replacing the damaged connector. Also there is the reliability of the connection over time/temperature/movement and this can be critical if doing vector based measurements at a few GHz.

Even some manufacturers advise (in the user manual) to use good quality 'connector saver' adaptors on the test ports of their precision analysers. This is to try and protect against long term wear but also to protect against damage and accelerated wear from cheap junk connectors that might get used by a sloppy operator.

[hendorog]

One thing to consider is the replaceability of the connectors in the instrument.

e.g. On the HP 8753A/B/C VNA's the N connectors on the front panel are reasonably easy to get, are not too expensive and are quite simple to replace. I have several spares for mine.

OTOH, the 7mm connectors on the test sets are not replaceable to my knowledge (well not without replacing the attached bridge), but no-one is likely to be using them directly and so the adapter becomes a de-facto connector saver.

Of course you may want to protect that adapter with a connector saver...

[G0HZU]

In my case I've got lots of high end RF gear here with N or 3.5mm connectors that are rated from about 10GHz to 20GHz because that is the frequency range of the test gear. I don't think it would be easy to find/fit cheap replacements so I have to be careful what I connect to them. I try and use connector saving adaptors (or DC blocks) as much as possible but it would still be unwise for me to then try and use unbranded/junk connectors to connect to the connector savers. It's a false economy for various reasons.

[BravoV]

Rosenberger LMR-400 clamp type connector vs Amphenol.

[Yansi]

Return loss of an aliexpress sma connector, to 2.45GHz (I simply made a 50ohm termination by soldering two 100ohm resistors to the connector):


Looks good right? problem is I can make the graph jump just by nudging the connector a bit, and it will probably become all rusty after a month of use (because of fake gold plating).

I am not sure how a connector's S parameters should be measured, but I'd say you likely measured the characteristics of the two resistors, not the connector in this case.  You would need to calibrate the reference measurement plane to be right on the connector, so the termination (or second port) is out of the way.

[Yansi]

Yep, but a reputable brand 1:1 is still VERY MUCH cheaper then sending a VNA back to Keysight/R&S/Copper Mountain to have the connectors replaced, and even a non preciosn one will save the wear on the connector.

Adding a hundred quid or so to the cost of a £50k instrument is hardly a big deal if it means you can avoid weeks of downtime while the thing is being repaired.

Regards, Dan.

Who here has a "£50k instrument" on his desk?

Where are the numbers, charts and other things that make this quantifiable? Right now it's just a bunch of people claiming things. Labs full of test gear and all we have to show for it are opinions?

If you mean >$50k when new I've got a few items that qualify.

I have to be very careful when introducing 'anything' into my workroom that has an SMA connector on it. I try and only use the very best brands of SMA connector (and cable) to minimise the risk of damage to my VNA connectors/cables and the Ecal module that comes with it. The Ecal uses Agilent/Keysight's PSC series 3.5mm connector and these are extremely fragile. I think the Ecal module cost £13k here in the UK and it can be trashed with a moment's carelessness with a cheapo connector that has a tolerance issue or any kind of surface defect on the centre pin. So I inspect any new SMA connectors/cables under a microscope before allowing them into the workroom. Even a tiny burr could cause problems.

You can see the Agilent PSC 3.5mm connector in the link below and how easy it is to damage it. I hate inspecting my Ecal kit for damage because it is quite stressful counting all the tiny petals within each PSC connector (4 connectors in total) to look for any defects.

For this reason I usually only use some very expensive Gore cables to connect to the Ecal module. A good Gore cable can easily cost $1k USD. But these cables can easily be damaged by items under test and pass on the damage to the Ecal module. So I try and do all my testing via known Suhner or Johnson SMA connectors.

Nearly all of my N type connectors/attenuators/cables are made by Suhner or Narda or HP. I have a few cheapo Amphenol ones but these only get used on cheaper test gear.

https://www.microwaves101.com/encyclopedias/how-to-not-trash-a-calibration-kit

That is what these are for, if you want to minimize wear on test gear connectors. Also seen them in the N type connector variant, used on some VNAs.

[Yansi]

https://www.microwaves101.com/encyclopedias/how-to-not-trash-a-calibration-kit

This is also "all opinion". I get the fact of destroying the expensive kits and get the procedures of correctly attaching SMA connectors, however I would like to see a proof, what impact the slightly bent finger of the connector "cunt"  will have on the cal kit's performance, if you will be able to measure any. 

[TheUnnamedNewbie]

https://www.microwaves101.com/encyclopedias/how-to-not-trash-a-calibration-kit

This is also "all opinion". I get the fact of destroying the expensive kits and get the procedures of correctly attaching SMA connectors, however I would like to see a proof, what impact the slightly bent finger of the connector "cunt"  will have on the cal kit's performance, if you will be able to measure any. 

While not true proof, I remember some Agilent/Keysigth cal engineers in one of Dave's videos talking about people using poor connector systems and thinking their device was in spec, but was in fact out of spec because the poor connector was ruining the measurement.

Regarding the "one slightly bent finger": Considering you need a torque wrench to get them to spec, I can't imagine it's hard to see how a bent finger would cause issues. It is hard to get true 'evidence' since nobody want's to ruin their $10k cal kit, or the ports on their >>$100k VNA.

[BravoV]

Regarding sacrificial N connector to preserve the one at the instruments, I used the right angle instead of straight, as its less stressful for the cables on my bench.

What you people used most ? Straight or right angle ?

[TheUnnamedNewbie]

Regarding sacrificial N connector to preserve the one at the instruments, I used the right angle instead of straight, as its less stressful for the cables on my bench.

What you people used most ? Straight or right angle ?

I tend to use N-to-SMA sacrificial connectors as all my test cables at home use SMA. I usually get straight ones and then use cables with angled connections instead.

[Yansi]

https://www.microwaves101.com/encyclopedias/how-to-not-trash-a-calibration-kit

This is also "all opinion". I get the fact of destroying the expensive kits and get the procedures of correctly attaching SMA connectors, however I would like to see a proof, what impact the slightly bent finger of the connector "cunt"  will have on the cal kit's performance, if you will be able to measure any. 

While not true proof, I remember some Agilent/Keysigth cal engineers in one of Dave's videos talking about people using poor connector systems and thinking their device was in spec, but was in fact out of spec because the poor connector was ruining the measurement.

Regarding the "one slightly bent finger": Considering you need a torque wrench to get them to spec, I can't imagine it's hard to see how a bent finger would cause issues. It is hard to get true 'evidence' since nobody want's to ruin their $10k cal kit, or the ports on their >>$100k VNA.

The question is by what amount it will be out of spec. By 0.1dB at a 5GHz? Or by what?
I have never seen a proof of the torque wrench effect on an actual measurement either. Do you know about any resources?

Please note I never stated that you should use cheap connectors.  I would just like to see some real figures that would describe the problem scientifically.

[TheUnnamedNewbie]

https://www.microwaves101.com/encyclopedias/how-to-not-trash-a-calibration-kit

This is also "all opinion". I get the fact of destroying the expensive kits and get the procedures of correctly attaching SMA connectors, however I would like to see a proof, what impact the slightly bent finger of the connector "cunt"  will have on the cal kit's performance, if you will be able to measure any. 

While not true proof, I remember some Agilent/Keysigth cal engineers in one of Dave's videos talking about people using poor connector systems and thinking their device was in spec, but was in fact out of spec because the poor connector was ruining the measurement.

Regarding the "one slightly bent finger": Considering you need a torque wrench to get them to spec, I can't imagine it's hard to see how a bent finger would cause issues. It is hard to get true 'evidence' since nobody want's to ruin their $10k cal kit, or the ports on their >>$100k VNA.

The question is by what amount it will be out of spec. By 0.1dB at a 5GHz? Or by what?
I have never seen a proof of the torque wrench effect on an actual measurement either. Do you know about any resources?

Please note I never stated that you should use cheap connectors.  I would just like to see some real figures that would describe the problem scientifically.

I can try and do some tests with torque wrenches next time I'm working on a VNA.
The difference is also mainly in the \$S_{11}\$ and not the \$S_{21}\$.

Quick google I found this, but this talks about cables and not connectors: https://www.tek.com/blog/improving-vna-measurement-accuracy-quality-cables-and-adapters

This paper I found does not show the true impact of cheap connectors, but does illustrate the impact of imperfections:
Yeou-Song Lee and T. Roberts, "Accuracy study on the newly introduced Anritsu W1-connector calibration and verification kit," Conference, 2003. Fall 2003. 62nd ARFTG Microwave Measurements, 2003, pp. 109-118.

Ofcourse, it does show that this is mostly an issue at MM-wave frequencies

[Yansi]

mm frequencies, like 30GHz+ ?

Having such a VNA is out of scope of many of us.

S11 makes sense, as the geometric imperfections have impact mainly on the impedance.

[TheUnnamedNewbie]

mm frequencies, like 30GHz+ ?

Having such a VNA is out of scope of many of us.

I never claimed that was well within everyday hobby territory, but I was not aware that this debate here was limited to hobby applications.

[Yansi]

I do not understand you point. If you have a 30+ GHz VNA, you will not likely care about connectors from china, or will you?

I thought the discussion was about using the chinese conenctors at applications well bellow that and to find where the limits are. I.e. to find connector torquing and chinese SMA has that much of an effect in that kind of application.

Using chinese SMA at K-band and such is out of the question.

[BravoV]

I tend to use N-to-SMA sacrificial connectors as all my test cables at home use SMA. I usually get straight ones and then use cables with angled connections instead.

My gripe is, whenever the sacrificial connector is installed, at my SA I could not install the front cover as I don't use it often, as it protrudes out even the right angle one. Also I don't like to increase the attach & detach cycles at my SA's N connector.   

[rfspezi]

I do not understand you point. If you have a 30+ GHz VNA, you will not likely care about connectors from china, or will you?

I thought the discussion was about using the chinese conenctors at applications well bellow that and to find where the limits are. I.e. to find connector torquing and chinese SMA has that much of an effect in that kind of application.

Using chinese SMA at K-band and such is out of the question.

That's exactly what i thought too.
The real art is not to buy expencive precision adapters (that's easy and every smart ass can do that and come with frightening arguments) but to find the best compromise between affordable price and performance.
From the N type connector in the opening post it's somehow clear that we are not targeting at the 30+ Ghz range.