Connector choice and current-carrying capacity of wire

Go To Last Post
13 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm fiddling around with a LED controller design - fairly typical.  It will use a TI tlc5926 LED driver (nice family) which provides a constant-current sink for each of 16 channels.  It can handle up to 120mA per channel, though I'm currently only targeting around 50mA or so.  So one provides a power rail to the anodes of all the LEDs, and the cathodes go to each channel pin - pretty simple.

 

I was originally planning on using ordinary headers and IDC connectors with ribbon cable to connect the LEDs - they're on a remote board.  Then I started looking at the actual currents ...  16 channels all on at 50mA is 800mA.  If they're pulling the full 120mA per channel that's almost 2A.  That little rinky-dink 22ga wire in the ribbon cable as the power rail just won't cut it.  Or will it ?  I've seen charts that show 22ga good for up to 5A for "short" runs, whatever short means.  But an online calculator says I need 13ga wire for 2A at 3.3V over 2 meters.

 

I suppose I could double up on the power rail pins, use 2 or 3 in parallel.  So 19 pins/wires to handle the 16 channels.  Except I'm using 4 banks - 4 power rails - and switching between them to get 64 LEDs out there.  Now we're just adding pins and adding pins ...

 

Another method might be to use the IDC/ribbon cable for the 16 channels only, and another connector type with heavier wire for the power rails.  That would certainly work, but begins to get a little kludgy.  It would be nice to have it all in a single connector.

 

Can anyone suggest a convenient connector type for this ?  Also, what gauge wire would I need ?  I've been digging and reading, but I'm a little bleary.  Say worst case to supply 2A at 3.3V over 2 meters.

Dean 94TT
"Life is just one damn thing after another" Elbert Hubbard (1856 - 1915)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Chancy,

 

The real question should be "How much voltage can you afford to drop over those 2 meters?" ... remembering that you have a drop in both the positive rail wire and its matching ground return wire.

 

Cheers,

 

Ross

 

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well, the tlc5926 can handle up to 17V on the power rail supplying the LEDs, and only needs a little over the LED forward voltage (1.5V I think, have to check) to work.  So, so long as the voltage drop doesn't go over that it should be good.  The power rail wire will have the largest drop as it's feeding everything.  The individual channels are only taking 1/16 of the total current each.

 

The earlier project I did with the 5916 worked fine, but was only using small LEDs with low currents.  Now I want to try handling much larger/brighter LEDs, with their associated higher currents.

Dean 94TT
"Life is just one damn thing after another" Elbert Hubbard (1856 - 1915)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Any chance to put the LED driver chips on the PCB holding the LEDs, and not have a wide bus of cables to sink current from each one?

 

That would let your run a respectable size power cable, (V+ and Ground), and a dinky tiny RS-232, (or whatever), serial comm's cable.

 

Much easier.

 

JC

Last Edited: Tue. Dec 16, 2014 - 05:14 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hah, I already have the layout to handle that as well :)  The problem is lack of space on the LED board.  For example, for a car center taillight, there is barely enough room for the LEDs and a connector, let alone the tlc and ancillary stuff.

 

I've seen some connectors a while back that look promising - like regular D-sub but with a few heavier power pins.  Expensive as I recall though.  Perhaps just doubling up on the regular headers/IDC pins will work OK.

 

I still can't figure a straight answer to the current-carrying capacity.  Several say 22ga is good to 5A, while several calculators insist that I need 13ga for 2A at 3.3V.

Dean 94TT
"Life is just one damn thing after another" Elbert Hubbard (1856 - 1915)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What on earth is 13ga wire?    

 

13 AWG is 1.828mm diameter and capable of 18A or so.   Fusing current of 198A.

13 SWG is 2.337mm diameter.

 

Both of those sizes look very unwieldy.    And unless you are running the length of an artic trailer,   well bigger than necessary.

 

Look up the standard wire tables to see the resistance per metre and choose accordingly.    I would guess that your connector specification is the most critical.    You aren't going to lose much voltage in the wire resistance.    The cable won't get too warm.    But anything on a vehicle needs good quality connectors.   e.g. vibration, temperature, water, corrosion, ...

 

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
I still can't figure a straight answer to the current-carrying capacity.  Several say 22ga is good to 5A, while several calculators insist that I need 13ga for 2A at 3.3V.
The difference is likely due to application.  A conductor of a given gauge will have different load ratings based on whether it is used for transmission or for 'chassis wiring'.

 

Have a look at:

http://www.powerstream.com/Wire_Size.htm

 

 

Different sources will give you different values.  Where I live, electrical code requires the use of 14AWG wire for 15A electrical circuits, 12AWG for 20A circuits.

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As Ross mentions, the voltage drop can be critical especially over the 0V conductor. A 1V drop here is significant as it will affect the logic levels seen by the devices- there's not much tolerance at 3V3. Factor in inductance and the situation gets worse. Sending logic levels may not be a good idea. Differential is one solution.

Also be wary of using idc connectors (the usual 0.1" type) for high currents. I seem to recall a figure of 200mA per pin.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ah-ha.  That figure of 200mA per pin is useful.  I don't think I can assign 10 pins to handle the 2A needed :(  Will do more research on header pins etc (after Christmas probably - in FL visiting parents etc).  The returns from the LED cathodes are fine - max will be 120mA each per the tlc5926, and I would be using maybe 50mA each or a little more.  I do want to be able to handle the max though.  But need to get the 2A *out* there ...

 

This is one of the calculators I checked - http://www.solar-wind.co.uk/cabl...  Doing 3.3V source, 2A current and 2m length it comes up with 2.5mm^2 and 13awg cable.

 

joey - what constitutes "power transmisison" as being different from the current-carrying capacity for "chassis wiring" ?  I need to "transmit" 2A over a max of about 2m, being pushed by 3-5V DC.  That chart almost sounds like a single 22ga wire will be well within capacity, since it's not long-distance wiring for transmission.  Heh, define long-distance though.

Dean 94TT
"Life is just one damn thing after another" Elbert Hubbard (1856 - 1915)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Chassis wiring is likely to be 200mm - 600mm.    Power Transmission might be 100m on a single site.    Power Transmission over the country might involve 100km but the voltages are 250kV rather than 3.3V.

 

Just do the sums on 22AWG wire for 2A and 2m.    What is the resistance and voltage drop?   Is it acceptable?

 

I would take anything from "solar-wind" with a pinch of salt.    The name is a bit of a clue.

Mind you,   they will be advising for minimal power and voltage loss over longish site runs.

 

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Chancy99 wrote:
joey - what constitutes "power transmisison" as being different from the current-carrying capacity for "chassis wiring" ?  I need to "transmit" 2A over a max of about 2m, being pushed by 3-5V DC.  That chart almost sounds like a single 22ga wire will be well within capacity, since it's not long-distance wiring for transmission.  Heh, define long-distance though.
Let's take the 22AWG as an example.  It has a resistance of 52.9392 ohms per 1000 metres.  Your 2 metre run then would have a resistance of 52.9392 * 0.002 = 0.1058784 ohms.

 

With 2 A going through, the voltage drop would be:

E = IR

  = 2 * 0.1058784

  = 0.2117568 V

 

That's for each conductor, so the 'send' (source) and the 'return' (ground) would exhibit that drop.  That's a total drop of 0.4235136 V.  This is significant compared to the 3-5 V you mention.

 

Let's take 13AWG.  6.56984 per 1000 metres.  2 metres has 6.56984 * 0.002 = 0.01313968 ohms.

 

E = IR

  = 2 * 0.01313968

  = 0.02627936 V

 

Each conductor, so both is 0.05255872 V

 

Much better.

 

All of these calculations ignore inductance, capacitance, and other factors like skin effect.  These can be important if the current drawn fluctuates over time, as it would with PWM drive.

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

Last Edited: Wed. Dec 17, 2014 - 04:11 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The "send" source would just be the power rail supplying the LED anodes.  The returns would be individual per LED, at a max of 120mA each, for 16 LEDs.  So the power rail would need to supply just shy of 2A while the 16 returns would be 120mA each, worst case.

 

This site http://www.mosaic-industries.com... says that 26ga should be good for up to 4.5A.  So it sounds like I would have to set up the LED supply voltage to be just high enough to account for the forward voltage plus any drop through the wire plus xxV overhead for the tlc5926.

Dean 94TT
"Life is just one damn thing after another" Elbert Hubbard (1856 - 1915)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

At 26AWG, 2 A, and 2 m, you'd see a voltage drop of 0.535 V.

 

At 120 mA, you'd see a voltage drop of 0.032 V.

 

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]