Pin header current capacity?

[technix]

Questions:
1. How much current can standard 1.27mm or 2.54mm pin headers withstand when used as a board to board interconnect?
2. What is a better high current board to board interconnect solution that does not cost me an arm and a leg, easy to find, and can be hand soldered without hot air?

Background: I have been asking a few questions regarding designing a digitally controlled power supply module (actually a series of such modules, all using the same pinout) Now I need to get the power in and out of the modules. I was thinking about using standard 1.27mm or 2.54mm pin headers for this, since they are cheap, plentiful and easy to solder. However unlike the usual low voltage signal, I am running things rated for 50A or 100A current here.

Maximum ratings this interconnect would carry:
* Input voltage: 4.5V - 25V.
* Output voltage: 0 - 24V, lower than input voltage.
* Rated output current: 0-100A (M379,) 0-75A (M378,) 0-50A (M377,) 0-30A (M376,) 0-15A (M375.) All modules can withstand a few amps of overload given ample cooling.
* Control interfaces: I2C, CAN, RS485

[Tomorokoshi]

Can you add some detail on what you need for this connector:

Is this the output connector to the user, the connector to some other power source, etc.

For the board to board interconnect, are these directly connected or is there a length of cable in between?

Will the user be field-wiring the load?

Do the input and output connectors for the 100A through the 15A options need to be the same?

[technix]

It's in the datasheet? 2.54mm pins can handle 3A per contact. I'll let you find the equivalent value for 1.27mm pins. Digikey's parametric search has plenty of options, I expect you'll be able to filter connectors by current capacity and geometry.

For pin headers I can buy in China, they almost never came with a datasheet or a brand. If the Digikey datasheet is generic it would be useful, otherwise I am left with rule-of-thumb guesstimate.

Can you add some detail on what you need for this connector:

Is this the output connector to the user, the connector to some other power source, etc.

For the board to board interconnect, are these directly connected or is there a length of cable in between?

Will the user be field-wiring the load?

Do the input and output connectors for the 100A through the 15A options need to be the same?

1. Everything goes through this pin header, input, output, controls, everything.
2. I don't expect any cabling being used.
3. I hope the user won't be field-wiring it.
4. The pins have to be compatible. The lower current ones can have less pins (to save cost,) but they should still fit in a socket intended for a higher current one, or vice versa.

[T3sl4co1l]

It's in the datasheet? 2.54mm pins can handle 3A per contact. I'll let you find the equivalent value for 1.27mm pins. Digikey's parametric search has plenty of options, I expect you'll be able to filter connectors by current capacity and geometry.

Note that's per contact for a single contact energized.  In practice, you may get away with 3A over a whole header -- or maybe not.  It'll definitely get warmer with all that high.

Connectors are a rather poorly defined thing to begin with, to the point that, you're rather lucky they work at all, let alone as well as they do.  Can you get away with it?  Sure!  Is it reliable over a long time?  Maybe, maybe not?

I wonder what success rate manufacturers actually use on their products.  Very rare to see statistics of any kind, in any datasheet, sadly.

Tim

[westfw]

In my opinion, you'll be limited by the construction and capacity of the female connectors, rather than the pins.  And if you don't have datasheets and a somewhat controlled supply chain, that may be pretty random.
I mean, a pin is pretty much a pin, but the female connectors can vary significantly in terms of number of contact points, size and composition of wiping area, springiness and heat-resistance of metal, and who knows what else.

[lem_ix]

1) Don't forget the return current has to go trough somewhere.
2) Wire ?

[CopperCone]

Well for those headers you can typically find something like a molex connector with the same pitch i think.

I think some connectors are basically a plastic thing that slides over the pins to prevent rotation.. Of course its supplied like a single unit but with a pair of pliers you can rip a very similar post out of it. These will have fairly well defined current characteristics.

[rounin]

50A is enough you might want to use a real high current connector --

Anderson Power Poles PP45 have a PCB mount option, though it is meant more for wire to board than board to board since there is no compliance in the interface. It is probably fine for a few pairs though. I regularly use these @ 40A - 60A.

https://www.andersonpower.com/_global-assets/downloads/pdf/ds-pp1545.pdf

Cheapish. Few dollars per side, very ubiquitous in industry. They are tall though.

I've never used it, but Molex Ten60 and similar are nice.

[technix]

50A is enough you might want to use a real high current connector --

Anderson Power Poles PP45 have a PCB mount option, though it is meant more for wire to board than board to board since there is no compliance in the interface. It is probably fine for a few pairs though. I regularly use these @ 40A - 60A.

https://www.andersonpower.com/_global-assets/downloads/pdf/ds-pp1545.pdf

Cheapish. Few dollars per side, very ubiquitous in industry. They are tall though.

I've never used it, but Molex Ten60 and similar are nice.

What I need is a high current board to board connector though, and it have to be a low profile one since for my use cases it can be height constrained. This is a board to wire one, which sadly does not meet my requirements.

[Amper]

Hi,
i have a similar problem and just did some testing.

I need to connect an aluminium pcb running about 200A DC and have very little space. I assumed i need to do my own machined connectors, but tuns out pinheaders can do the job. I modified a female 0.1" header by taking the soldering tails off and filling the space with solderpaste. Now the plastic sits flush on the PCB. I tested a single pin at various currents up to 15A and several mechanical loads, the worst result i could get was about 3.6mOhm on a round female with a square pin in it. As i have space for a 12x2 header that will be roughly 150uOhm heating with 6W at full power. No problem if your pcb is cooled well, which in such applications is usually no problem. Running at a more usual 100A or so it will be a pretty good solution! attached are two pictures of the desoldered test board and the final product. Testing of the entire connection will hopefully be done soon.

Oh and for comparison, an 8mm Bullet connector used in RC modeling turned out to have about 240uOhm, and these things are massive!

[eliocor]

Maybe not as cheap as pin headers (but designed for high power)...
https://www.samtec.com/connectors/rugged-power/power

[Amper]

Oh,
thanks! I was looking for the name of these for a long time now !

Though for my application sadly i cant fit them, they have quite a lot of plastic around them when they are actually rated for these currents. i can barely fit the 2x header.

[JS]

  Can't you keep your high current in a single board?

That might be the best solution, but sure you have a reason... What you could try cheap and sturdy is to just make a big pad in the boards and screw them together with a cooper bushing in between. The bushing should be nicely machined to provide good contact and not bend the board, it doesn't NEED to be cooper but I guess it could be the best choice.

JS

[JohnnyMalaria]

I find current rating quite confusing. Shouldn't it be at a specified voltage? Isn't it all about heating and likely damage/fire? If so, then the heat is the power, P (Joule/sec aka Watt):

P = VI (at fixed resistance)

Often, fuses are labeled at 250V and 125V, e.g., 250V/500mA, 125V/1000mA.

Maybe I'm having a  moment.

[Amper]

Well, you said it yourself, fixed resistance. P=I²*R -> P=200A^2*0.00015Ohms = roughly 6W. Since i have very good cooling through the board thats no problem at all.

@JS: In my case this is the termination going to the bus bars, not to a second pcb. Problems that led me to my choice:

- Aluminium pcb, so no throughhole / drilling without a lot of work to insulate and relieve stress

- Easy disassembly for prototype reasons, just plugging is nice

- Contact resistance due to oxidation/wear, I have an automotive inverter here, one edge of the igbt module is completely molten, because a connection like you described it ha a tiny bit of oxidation on one of the bus bars.

- Size limit, you can see how crammed full the board is, those fets are all necessary to get RDSon as low as possible (0.5mOhm per set of 3) Also there will be the controller board on top of this one, originally i thought i had to run the high current through it to connect from the top, but this pinheader thing is amazingly low profile so it actually doubled the usable space on the top board.

I have spent a lot of time thinking about this problem and so far this seems to be the best solution.