IET
Decrease font size
Increase font size
Topic Title: D connector screening
Topic Summary:
Created On: 15 February 2012 03:32 PM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 15 February 2012 03:32 PM
User is offline View Users Profile Print this message



johnnmann

Posts: 53
Joined: 18 October 2006

It has been suggested to me that the screening is not good enough on our D connectors. The reason being that the connector sits in grooves in the two halves of the backshell but because it is not clamped tight and is able to "float", EMC continuity between the connector and backshell cannot be guaranteed. Even when the jackscrews are tightened, they only hold the connector to its mating socket and don't act on the backshell.

I've searched extensively, but it seems that all D connector backshells work like this, even the super-duper EMC ones, so - is the suggestion valid and if so what can I do about it? If anyone can point me at some suitable products or information it would be appreciated.

Thanks, John
 15 February 2012 07:47 PM
User is offline View Users Profile Print this message



IanDarney

Posts: 32
Joined: 18 January 2003

You could contact Roy Ediss.
http://www.ediss-electric.com
Ian Darney
 16 February 2012 09:19 AM
User is offline View Users Profile Print this message


Avatar for gkenyon.
gkenyon

Posts: 4488
Joined: 06 May 2002

Originally posted by: johnnmann

It has been suggested to me that the screening is not good enough on our D connectors. The reason being that the connector sits in grooves in the two halves of the backshell but because it is not clamped tight and is able to "float", EMC continuity between the connector and backshell cannot be guaranteed. Even when the jackscrews are tightened, they only hold the connector to its mating socket and don't act on the backshell.



I've searched extensively, but it seems that all D connector backshells work like this, even the super-duper EMC ones, so - is the suggestion valid and if so what can I do about it? If anyone can point me at some suitable products or information it would be appreciated.



Thanks, John
John,

Assume you are talking here about "manually-wired" connectors: some of the moulded-types can address this issue rather better (but to guarantee you'd need to audit the manufacturing process, of course )?

Many of the "EMC hoods" have a metal cable-clamp or spring-clip for the drainwire/screen to be wrapped around for connection to the hood, and the connection to the "shell" is fortuitous at best. I've seen an odd one or two types with an arrangement having springs on the jackscrews to ensure the hood can take the screen from the shell of the connector itself.

For applications that require "good" screening practice, I do often see this:
Metal hood with screw-type cable clamp, under which the braid or drainwire of the overscreen is terminated: if drawinwire, then this is wrapped around the conductive side of the foil a few times to get the recommended 360-degree connection. "pigtail" is then taken from the clamp and soldered onto the shell (if extended drainwire, then this is insulated from the clamp to the shell to prevent inadvertent electrical connection, otherwise something like 24AWG tri-rated is used, soldered to braid before clamping).

However, there may be alternative approaches: can you provide more information on what interface standards you are using, and whether you must take the screening from the shell, or whether the interface pinout arrangement you're employing has a "screen" or "ground" connection intended for the screen. As an example, the original intention of RS232, for example, where the DB25 had a "SCN" pin, which with many manufacturers' applications you could select whether this was connected to the outer shell only, or both the shell and ground?

I think the upshot of this, whatever method we use, is that most of the "manually-wired" D-type connectors is not really suited for the absolute very best EMC practices as stated in BS IEC61000-5-2, particularly if the disturbances we are trying to protect against are very short wavelength signals: there are always little gaps between the hood and shell for the signals to leak through.

If this abolsute very best practice is required, and you are forced to use manually-wired connectors because of the application, then the solution would be to enclose the equipment at each end, in earthed, or functionally-bonded, metallic enclosures, and provide conductive containment (having 360-degree bond to the enclosures) - of course, this provides mechanical protection too.

-------------------------
Eur Ing Graham Kenyon CEng MIET TechIOSH

Edited: 16 February 2012 at 09:32 AM by gkenyon
 17 February 2012 11:19 AM
User is offline View Users Profile Print this message



johnnmann

Posts: 53
Joined: 18 October 2006

Thanks for your helpful reply Graham.

You have confirmed my initial thoughts that the D connector is probably not the best for EMC performance. We are manually wiring connectors but not to RS232 or any particular standard. The connections go to various transducers and peripheral devices. When I first joined the company a few years ago, all screens were taken through on connector pins and metalwork was isolated "in case there was an earth loop". I have now done away with this in favour of 360 degree screen connections to the body of the connector and enclosure metalwork.

I'm not sure that a pigtail connection to the shell would be much advantage as it is unlikely the shell and backshell will not be making contact at some point(s), what we cannot guarantee is there will be contact at every point.

I think this is a limitation we will have to live with.

Thanks again for your help,
John.
 17 February 2012 12:03 PM
User is offline View Users Profile Print this message


Avatar for gkenyon.
gkenyon

Posts: 4488
Joined: 06 May 2002

Originally posted by: johnnmann

Thanks for your helpful reply Graham.
No problem, John.



You have confirmed my initial thoughts that the D connector is probably not the best for EMC performance. We are manually wiring connectors but not to RS232 or any particular standard. The connections go to various transducers and peripheral devices. When I first joined the company a few years ago, all screens were taken through on connector pins and metalwork was isolated "in case there was an earth loop". I have now done away with this in favour of 360 degree screen connections to the body of the connector and enclosure metalwork.
Yes, I concur with this.



I'm not sure that a pigtail connection to the shell would be much advantage as it is unlikely the shell and backshell will not be making contact at some point(s), what we cannot guarantee is there will be contact at every point.
Understand where you're coming from here.

My point, however, was whether the screen connection itself comes from the d-shell on the equipment, rather than a "pin"?

If it comes from the shell, then at least with the pigtail connection, you get a guaranteed connection from the functional bonding to your cable overscreen/braid, meaning you've not reduced the effectiveness of that measure. Of course, if you're relying on the hood alone to bond the braid to the shell, you've potentially got in ancreased impedance in the path, which the pigtail overcomes.



I think this is a limitation we will have to live with.
Without changing connector type, or using metallic enclosures and containment, it seems that the only option for improvement could well be the "pigtail" to the d-shell, if, and only if, that is where the functional earthing/bonding for the overscreen is taken from.

-------------------------
Eur Ing Graham Kenyon CEng MIET TechIOSH
 17 February 2012 12:30 PM
User is offline View Users Profile Print this message



johnnmann

Posts: 53
Joined: 18 October 2006

Originally posted by: gkenyon

My point, however, was whether the screen connection itself comes from the d-shell on the equipment, rather than a "pin"?


Yes, the screen comes from the shell.
 17 February 2012 06:25 PM
User is offline View Users Profile Print this message


Avatar for gkenyon.
gkenyon

Posts: 4488
Joined: 06 May 2002

John,

In that case, do consider whether you need the "pigtail" approach to reduce the impedance of the noise sinking path, or the overscreen/braid may not provide as effective a "sink" of higher frequencies. Not always necessary: depends on the actual EM environment you're installing in.

-------------------------
Eur Ing Graham Kenyon CEng MIET TechIOSH
 25 February 2012 12:14 PM
User is offline View Users Profile Print this message



jencam

Posts: 608
Joined: 06 May 2007

D connectors were originally designed for low voltage / current digital and analogue signals at frequencies up to a few MHz. They were not intended for use with very high bit rate digital signals which is why their effectiveness of EMI screening is invariable. Is there a reason why you have to used D connectors when connectors that are more suitable for high bit rate / frequency signals with guaranteed standards of EMI screening exist?
 25 February 2012 10:46 PM
User is offline View Users Profile Print this message


Avatar for gkenyon.
gkenyon

Posts: 4488
Joined: 06 May 2002

Originally posted by: jencam

D connectors were originally designed for low voltage / current digital and analogue signals at frequencies up to a few MHz. They were not intended for use with very high bit rate digital signals which is why their effectiveness of EMI screening is invariable. Is there a reason why you have to used D connectors when connectors that are more suitable for high bit rate / frequency signals with guaranteed standards of EMI screening exist?
Whilst this is true, these types of interface persist today.

Unfortunately, they are used in applications merging "old" and "new" technologies, in environments where MHz and GHz signals persist (mobile, WLAN, etc. etc.), and of course have the ability to couple to newer equipment (maybe operating at GHz) through these older physical interfaces.

It's not always an easy solution. However, I do agree that if this connector were chosen for a new interface without consideration of the overall issues, maybe it wasn't the best choice.

-------------------------
Eur Ing Graham Kenyon CEng MIET TechIOSH
Statistics

See Also:



FuseTalk Standard Edition v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.