I assume from your post that this is a surface through panel connector with locking screws on the outside. If this is tyhe case be sure that the stackup of the stackup of the panel and spacer meet the manufacturers specification. The case you have stated with inadiquate depth for the mating connector to seat properly may cause more issues that a drop in current carrying capacity and may lead to faulty or lost connection and arcing in the contacts leading to failure. I have experienced this with a connector designed for pannel or flush mount being used for a trough pannel application with a resulting loss of power amd data lines. As for a specification look to see if there is a MIL-STD or HBK for the connectors but it should be re-quoted in the connector manufacturers spec; the case I stated was a misaplication of the connector manufacturers specification by an equipment manufacturer and was only found when a unit came that could not be screw locked at all and multiple system malfuctions had occured.

Hope you find the source of the problem, but do not rely solely on the LRU manufacturer to have done the job properly.

We always use d-sub for our scientific equipment unless otherwise stated.

We have several types with the only difference being the length of the hex pillar that takes the screw on either side of the plug. This is to facilitate different panel thicknesses as you would get when fitted to aluminium, not so strong therefor thicker, or steel sheet.

We ALWAYS fit a flat washer on the pillar and a lock ring on the nut. This is specified by the connector standards as far as we are concerned. If you have improper engagement, try to see if your pillar is the wrong length. (they are easily shortened as they are only brass with an electroplated coat on them).

The way to see if the engagement is up at max possible is to take the socket out of the panel and connect the plug without the screws on it. Look at the gap between plug body and the socket face, there should not be any gap bigger than 0.5 mm. Now do the same after you fit it back and see if this gap is bigger. If it is, the pillar is too high.

Another note is that if the missed engagement causes an issue with load capacity, the socket and pin configuration has been wrongly choses for the job. The load, and the current, should never get that close to the capacity of the contacts. I do not know from the top of my head what it is supposed to be but there is a safety factor that has to be applied to the max load in equipment which will than decide load capacity of connectors. 0.5 mm less engagement should not make that go over the limit ever. You can however, like other post mentioned, loose contact altogether when it is further away and that leads to loss of data.

Sounds bad and is probably not only an electrical problem, but also a mechanical problem too.

Surely the manufacturer can supply pillars say 1 mm shorter to allow full engagement?

I have had similar problems in the past with HD SCSI cables, the problems were absolute sods to find.....so do not go there in the first place, you will surely regret it !!!

Pillars are brass so any nail file will make them shorter without any problems. These accessories are hard to find as most suppliers concentrate on the connectors themselves and not the hoods, housings or the fittings.

We get ours from Farnell Electronics.

I have plenty of them and they are also made of stainless steel and normal steel as well......!

But the product should be fixed at the source, not fudges at the customer. Filing on a finished PCB could cause a disaster if not done cleanly.....

That is true. You must ensure the correct pile up of fillings on the circuit board for maximum effect

I have never had any made from stainless myself but nothing to say that it cannot be done I suppose. Strange that all ours are brass, we use at least 20 different lengths for various spacers between product and final covers, as well as for d-sub connectors.

Thanks to everyone who has replied with well-considered and useful advice, experience and suggestions. I confirm that the connectors concerned are back mounted on the inside of a front panel. We do not use any connector backshells but do use the screwlock retaining retaining "butterflies". I have checked that the butterflies don't foul anywhere. When fully screwed in place there is about a 1mm gap between the plug and socket housings. Usually I would expect there to be no gap here as mentioned earlier in the thread.

I have carried out an experiment to increase the standoff by about a further 1mm and found that signals were starting to fail, so I can confirm that there is very little engagement at best.

I wlll fill in with a little more information to show why it is such an issue for me...

Unfortunately I am unable to adjust the pillar (female screwlock) height as the retaining nut is on the inside of the equipment. Contamination caused by any filings would be unacceptable as the equipment is to be used in a very clean environment. The equipment has about 60 connectors - all with the same standoff height - so removal of the external spacer would necessitate a major stripdown to gain access to the internal nuts. Unfortunately we are talking about a one-off multimillion Euro equipment that would need to undergo major retest following the repair so the pressure is on to establish whether or not we really have a problem before pulling it apart. I'd really like to be able to show by analysis of the specification dimensions that there is sufficient engagement depth for our signals, which are fortunately all of relatively low current and frequency.

The contacts used are top quality military/space grade with the female contact being in two parts - an inner part with a sprung contact and an outer sleeve for mechanical strength. I am concerned that the pin contact has to go a fair distance into the socket before a reliable contact is made. Just inserting into the first mm or two may provide some unreliable contact (thereby passing any tests we may carry out) as the point of the pin just touches the non-sprung part of the socket.

I'm sorry this is getting into the fine details of the construction of these connectors but I see no alternative to going as far as is possible to avoid the expensive repair.

Hoping for further advice and bright ideas.

You have answered your own question really.

If it is only 1 mm engagement and it is too labor intensive for you to do the job, you have to send it back with a clear specification to which it should have done in the first place. You have no defense against any action from your customer in a years time if they can prove you knew about this. It may take that time for problems to start due to corrosion of the pins.

Hope you will find the best way forward for you and glad we could be of help.

Thanks a lot for the help. It supports the conclusion and way forward I am being forced to consider.

Cheers

The screwlock assemblies I have used in the past were intended for use with a panel thickness of .062 inches. If they have installed the D-sub in a .091 thick panel, this would account for the excess height on the screwlocks. I would ask the equipment manufacturer about this. The only solutions I see are to

1. Use a shorter screwlock assembly or

2. Mount the connector on the outside of the panel. Of course here you would have to mount using the brass nuts which come with the screw lock assemblies.

and of course the lock washer which comes with the screw lock assembly goes between the connector and the little nut (which comes in the package). It does NOT go between the screwlock and the mounting panel.

Hope this helps

Bill

I really do not see a choice, strip down and replace with shorter parts, nothing else is acceptable in the price range you mentioned.......

Have you talked with the supplier of the product, and relayed your possible concerns with the connectors?

Were the connectors and the possible problems with the mating in the spec you gave to the supplier?

If the supplier can guarantee that there will not be a problem in the design and application for the said job, then if there is a problem, the expense should go back to them for repair.

If you decided to go for the cheapest solution for your multimillion euro/dollar/Kroneg unit, then you could just be dealing yourself problems.

How much movement is on the connectors? mechanical vibration? wiring harness weight pulling on the connectors?

whats the connector design tolarence? maybe the supplier is still within the design tolarence of the connector?

Of course getting the manufacturer to correct their problem is the best solution, but schedules and other considerations may make that option not practical. You may want to consider purchasing port savers and counter-boring the mating interface to assure full engagement. They are available from many sources, here is one DB9 Male to Female Port Saver

All connectors are rated for a limited the number of engagements. Some are as low as 20 while others are in the 1,000's. Using port savers on expensive equipment especially test equipment that will have cables connected and disconnected routinely is a prudent thing to do.

Hi Snave,

I am intrigued by your suggestion, what does this item really do to deserve the name "port saver"?

I can see the picture clearly and all I think I see is a gender changer. I can be wrong and often have been, please explain as it may be of use in my line of work where we use thousands of d-subs a year.

Regards and thanks,

case491 A port saver is similar to a gender changer, but it doesn't change gender. It has male pins on one side and female pins on the other.

I was first introduced to them when I worked for a company that manufactured equipment which employed D-subs on the customer interface. When the product went through many test stations the connectors would sometimes become damaged, either the pins or the jack posts. The solution was to install a port saver on each of the connectors that stayed with the unit as it went though test. The port saver was removed before the equipment was packaged up. The final product shipped with a connector that was only used once and not mangled accidentally by handling in the factory.

Although not a common, there are companies that manufacture port savers for DVI connectors as well.

"Port Saver" haven't heard them called that before....

In general usage, a female connector will expire before the male counterpart, we use a similar construction on DIN41612 connectors (3x32 way connectors.. 32A, 64AB, 96ABC) where the interchangeable plate has a Male connector installed, and the system has a Female plugged into a 2nd Female connector.

When you start getting problems, you changeover the center Female connector, haven't seen many of the male connectors getting replaced.

Although I have not seen "Port savers" in the SCSI world, all the normal High Density cables are male to male and sadly get damaged too often to be good......