Can 22 AWG wire be used for a design of up to 7A

[atlbraves49]

For our design, we're sending at most 7A (usually less) to a prototype board with a resistor and MOSFET on it... can 22 AWG wire handle the amperage? the reason i ask is that no other size wire fits in the wholes of the prototype board...

Thanks,

 

[mgb_phys]

The wire standard says max for 22awg is 7amp (although that's probably assuming AC)
In practice it depends on cooling, if the wire is going to be packed in a sealed box with lots of heat generating components i would use larger wire and give yourself a safety margin.

 

[atlbraves49]

but no other wire fits into the prototype board holes.

I mean, I could run a fan inside the box. it won't be completely sealed.

 

[mgb_phys]

If you are selling the thing and it has to meet CE/UL then you have to use what the wiring code says. Otherwise just try it.
The wiring code has a safety margin so if it says 7A you know that more than 7 will work in most cases.
But if the system also has a couple of Pentium CPUs at 4Ghz and is in a sealed box and is going to have to work sitting in the sun in Arizona in summer you need to de-rate it.
Generally it's a bad idea to solder power wires into the holes on a prototype PCB - you are better off putting pins in them (snapped off 0.1 header will do) and solder to the pins. That way you have somewhere convenient to clip meter probes to and it's easy to solder/desolder when you have a problem - repeatably soldering heavy wire into a pcb can cause the track to lift off.

 

[bipolar]

Solid core 22ga is 5A continuous, 7A peak with DC at 25% duty.

The breadboards I've worked with melted when high current (5-10A) was drawn in PWM, so now I run a header out to a small PCB with the High Heat components mounted to it, along with a separate run to the PCB for the power lines in an H-Bridge application.

The separate power line helps with the noise on the breadboard power lines. Always use 0.1uF ceramic or film decoupling caps across all ICs, and mid-value (22-100uF) electros at a few spots along the power bus. Solderless Breadboards are a huge electrical network of inductance/capacitance and resistance. The less power pulled from the Solderless Breadboard, the better the chances of clean switching. Remember to combine the high and low value caps (0.1uF/22uF) to prevent any oscillation from springing up.

 

[Redbelly98]

One could run TWO 22 AWG wires to carry 7A (3.5A each). They would both fit into the breadboard holes.

Presumably you'd need two more wires to handle the return current as well, so that's 4 wires total.

 

[bipolar]

The solderless breadboard itself has small conductors internally, they are designed for small signal prototyping, and have rather high resistance compared to the soldered, wide tracks on a pc board.

The wire size isn't as much of an issue as the voltage fluctuations while switching higher powered loads on and off. Logic ICs, op-ams, microcontrollers, and even analog circuits tend to not work right when the voltage drops when the output is switched.

The power devices should be at least on a separate solderless breadboard with an independent large gauge wire path to the power supply, not daisy chained to the first breadboard. This isn't an option for TO-3 package devices, so a "spare" TO-3 holder with 22ga wires to go to the breadboard would work for prototype verification, but not long term usage. With 1 foot of 22 gauge wire, about half a watt would be dissipated in the wire alone at 7A, about a 0.1V drop, not counting the resistance of breadboard connections.

 

[atlbraves49]

If you are selling the thing and it has to meet CE/UL then you have to use what the wiring code says. Otherwise just try it.
The wiring code has a safety margin so if it says 7A you know that more than 7 will work in most cases.
But if the system also has a couple of Pentium CPUs at 4Ghz and is in a sealed box and is going to have to work sitting in the sun in Arizona in summer you need to de-rate it.
Generally it's a bad idea to solder power wires into the holes on a prototype PCB - you are better off putting pins in them (snapped off 0.1 header will do) and solder to the pins. That way you have somewhere convenient to clip meter probes to and it's easy to solder/desolder when you have a problem - repeatably soldering heavy wire into a pcb can cause the track to lift off.

i don't have any available pin headers to snap pins off of but could i use like... MOSFET pins (i have a few extra MOSFETs that arent going to be used) or even capacitor leads?

 

[mgb_phys]

Sorry I didn't see the bit where you said it was a breadboard.
But when/if you do a PCB you don't want to solder wires to the tracks if you can avoid it.

 

[atlbraves49]

Sorry I didn't see the bit where you said it was a breadboard.
But when/if you do a PCB you don't want to solder wires to the tracks if you can avoid it.

im not using a breadboard. Someone else said that.. not sure why.

Im using a prototype (through hole) board, basically the same as PCB. So if i can't locate pin headers, can i just cut off the pins from MOSFETs that i have lying around and use those to connect the wires to the board?

 

[Pumblechook]

We used to re-wire power supply units which ran hot for many hours a day. The PVC insulation became brittle and literally fell off. We re-wired with PTFE insulated wire which was in fact smaller in diameter.

 

[mgb_phys]

Im using a prototype (through hole) board, basically the same as PCB. So if i can't locate pin headers, can i just cut off the pins from MOSFETs that i have lying around and use those to connect the wires to the board?

Yes, anything will do - although the thicker the better.