5
\$\begingroup\$

The question

Did someone knows what are the current capabilities of the standard D-sub (aka Cannon) connector. I want to use it for power source connection together with the signal.

I need pretty high currents - something like 10A or even higher and also, I can use several pins for this task. Even more pin connector (15 or 25 instead of 9) is acceptable.

Special modifications (special power D-sub connectors) are not acceptable. I want to keep the element base very standard and cheap.

Did someone has personal experience with such D-sub use?

The experiment

After some not very informative search, I decided to make some experiment. I connected all the pins in a pair of D-sub 15 pin connectors in series and then connected it to a 3A current source. This way, through the whole connector, 45A of current is flowing (15pins x 3A each). The total voltage drop on all pins is 0.12V and the power is 0.36W total. Now I will leave it for a while in order to see how it will degrade with the time.

Experimental results 1

After 1 hour of work on 3A per pin (45A total) the temperature of the external metal body of the connectors raised to 39 deg. Celsius (27 deg ambient temperature). The average contact resistance is 3mOhms per pin. I leaved it to work for a night and tomorrow will try to couple/decouple them several hundred times and to check what will happen.

Experimental results 2

After a night of work nothing changed. The body temperature is stable, 12°C above the ambient temperature. After several tents of times coupling/decoupling under voltage, the contact resistance has been increased from 2.4mΩ to 2.7mΩ per pin.

\$\endgroup\$
  • 4
    \$\begingroup\$ Have you looked at the datasheet? They very often specify maximum current per-pin. \$\endgroup\$ – Connor Wolf Oct 10 '13 at 11:22
  • \$\begingroup\$ I tried with this: store.comet.bg/Catalogue/Product/2989 but there is no such parameters. Anyway, I am asking for the personal impressions as well. \$\endgroup\$ – johnfound Oct 10 '13 at 11:27
  • 1
    \$\begingroup\$ Try a proper manufacturer (Tyco, AMP, etc...). Also, shop for other manufacturers. Some have shittier datasheets then others. In general, all the d-sub connectors I've seen have been rated at >= 1A per pin, so if you use a D25, and 12 pins per rail (assuming only power and GND), you'd likely be fine. \$\endgroup\$ – Connor Wolf Oct 10 '13 at 11:30
  • 2
    \$\begingroup\$ The pin current ratings will depend in the connector design and body material. When pins are paralleled the temperature rise due to the current passing through the contact resistance is additive and so contacts that are in between other contacts will cause higher net temperature rise than pins that are near the outside edge of the connector body. Only some connector manufacturers make an effort to parameterize this behavior and provide contact derating for cases where paralleled contacts are used. \$\endgroup\$ – Michael Karas Oct 10 '13 at 13:17
  • 1
    \$\begingroup\$ You should also consider how exactly you are balancing the connection, particularly during connection/disconnections, as the D-sub connectors don't guarantee any sort of 1st/last connection pin (like the ground pad on a USB connector, which is slightly longer than the others), so during your insertion/removal cycle hopefully you aren't still powering them at 3A per pin & thus overloading the poor first/last pin, thus increasing it's contact resistance by minor arcing/etc. \$\endgroup\$ – user2813274 Jul 15 '18 at 0:10
2
\$\begingroup\$

Have a datasheet: http://portal.fciconnect.com/Comergent//fci/drawing/c-dsub-0071.pdf for http://www.digikey.com/product-detail/en/865609SLTLF/609-1467-ND/1001781 which claims a max of 5A per contact!

Personally your 45A total seems like a horrifyingly high number, but if you wanted to split your 10A across 10 power and 10 ground connections in a 25 way connector that seems OK.

Don't forget the wire gague; http://www.powerstream.com/Wire_Size.htm says 20 AWG or larger (smaller AWG number), but it also needs to fit the D connector.

\$\endgroup\$
  • \$\begingroup\$ The first data sheet I see that to specify the current clearly. And the operation temperature up to +125*C. \$\endgroup\$ – johnfound Oct 10 '13 at 16:13
  • \$\begingroup\$ Note that 125C is a tested maximum not a level you should be working at if you can avoid it; you don't want to have incidents of "I'll just unplug this OUCH MY HAND IT BURNS" \$\endgroup\$ – pjc50 Oct 10 '13 at 16:23
  • 1
    \$\begingroup\$ Yes, of course. BTW, note that these 5A are for their "Economy" series. The "High performance" is rated at 7.5A :) \$\endgroup\$ – johnfound Oct 10 '13 at 16:28
2
\$\begingroup\$

The obvious answer is to READ THE DATASHEET. The current rating is one of the important specs for any electrical connector, so even marginally reasonable datasheets will tell you this.

I haven't looked up the datasheet for a "D-sub" connector since that's your job (you also need to explain which D-sub connector), but anything normally understood as a typical "D-sub" connector, like a DB-9, isn't going to do 10 A per pin. Not even close.

You can parallel multiple pins of a connector to get effectively higher current rating at the expense of fewer connections. You should derate the total current capability somewhat since the pins won't share the current equally, but you don't have to go too far with that since individual connections are mostly resistive.

\$\endgroup\$
  • \$\begingroup\$ Well, according to the experiment, DB-9 can handle 27A and at least to not fail instantly. Can you comment? \$\endgroup\$ – johnfound Oct 10 '13 at 13:47
  • 2
    \$\begingroup\$ @john: That means very little from a product design point of view. That any one connection tested at one set of conditions can significantly exceed the long term rating applicable over a wide range of conditions is not surprising. However, it is meaningless since there is no guarantee the next connector at a different temperature of a different age with different amount or type of dirt will do the same thing. The manufacturer has done the analisys of what you can rely on and told you this in the datasheet. That's what you need to use for product design. \$\endgroup\$ – Olin Lathrop Oct 10 '13 at 13:52
  • \$\begingroup\$ Yes, you are right. But the data sheets (I found) are vague about all these factors you described in your answer. \$\endgroup\$ – johnfound Oct 10 '13 at 14:00
  • 1
    \$\begingroup\$ @john: So go find better datasheets, or better connectors that come with proper datasheets. Any reputable connector company will explicitly tell you the maximum current per pin. \$\endgroup\$ – Olin Lathrop Oct 10 '13 at 14:21
2
\$\begingroup\$

On Megasquirt EFI they regularly run currents of 5A through a single or pair of pins on a D-37 connector. I've run the system on several cars for years and it's been very reliable.

As someone said above, proper manufacturers like Tyco, Amp, ITT Canon adhere to the spec & will publish a data sheet you can believe in.

However, they also do mixed-signal D connectors which have different contacts allowing for high-current, co-axial cable, and normal signal pins in one connector body. They cost more but may be a better solution.

picture of mixed-signal D connectors

\$\endgroup\$
  • \$\begingroup\$ I know about these power versions of the connectors, but don't want to use them because they are not very common. Anyway, vote up for the first paragraph. :) \$\endgroup\$ – johnfound Oct 11 '13 at 8:11
  • \$\begingroup\$ It's an odd guide, but I've probably done 20,000 miles motoring running 2 banks of 4 fuel injectors through 2 pairs of DB37 pins. The heritage of the D-Connector is MOD/Aerospace so they're surprisingly good when you buy decent ones. \$\endgroup\$ – John U Oct 11 '13 at 9:50
0
\$\begingroup\$

While Sub D connectors were never intended to carry a lot of power, 10A is OK if you double or even triple (if you're not using gold-plated contacts) the contacts and wires, both plus and minus. For too much more than that, you should be using a different connector designed to carry the power. Yes, you can parallel the contacts and wires even more but then you start getting into the heat issues of parallel pins and wires in a cable. It's a whole different game if you'll be connecting and disconnecting the connector while powered. Regardless, beware of using any connector that does not list the current carrying capacity of the contacts. Lot of cheap (Chinese) junk out there and there is a BIG difference.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.