# How can 22 awg cable over 100cm deliver 3A

I wonder how 22 AWG cable e.g. USB-A to DC power jack with 100 cm length can deliver 3A.

Why am I asking -> Using this calculator (I hope its correct) by entering values 2% drop, 1 meter, 5V, 3A gives me output 16 AWG. And I am looking for right cable to get.

What I need is stable output 5V, 1.7A (top) over 1.5m from power bank which is capable of 5V/3A over USB-A in winter condition. I mean like max -10C (most likely about -2C)

What would you suggest? Power bank has two outputs USB-C/USB-A. I can use both connectors at same time. There is absolutely not possible to use AC/DC power source.

End of cable will be most likely soldered on-board to avoid contact disruption.

Device powered via cable is ESP-32 with some LED strips (outdoor usage with movement)

• For a straight cable laying on a surface with no airflow and room temperature, I start to notice cables running lukewarm at 10 A/mm^2. If you coil it, you will have a problem after a while. At these low voltages, I would expect you run into voltage drop issues before heat becomes a problem. What voltage does your ESP32 require as minimum? – winny Jan 21 '20 at 19:16
• You will clearly have a voltage drop of about 7%. Do you really need 2%? – Scott Seidman Jan 21 '20 at 19:19
• @winny ESP32 runs at 3.3V. There is built-in 5V to 3.3V converter. On-board are also capacitors. 470uF + 10uF before input to ESP32 – user1085907 Jan 21 '20 at 19:25
• Try not to design to "as much as possible". Shoot for what you need. – Scott Seidman Jan 21 '20 at 19:32
• "What I need is stable output 5V..." - this will not happen. To start with USB voltage spec is 4.75V min. Then there will be some voltage drop in your cable. What is the minimum voltage your circuit can accept? – Bruce Abbott Jan 21 '20 at 19:40

My suspicion is that they've calculated the 3A capability based on the shortest cable they offer in that listing, which is 50cm. 0.5m of 22 AWG copper with 3A over it has a voltage drop of about 160mV, which is reasonable enough. At one meter, however, it's more like a 320mV, which is probably pushing it a bit if you're regulating 5.0V down to 3.3V with a linear regulator, because that reduces your headroom down from 1.7V to 1.38V.

However, if you know you're only ever going to run it at 1.7A max, you could probably get away with it. In that situation the drop would be more like 180mV (~3.6%) giving you about 4.8V to work with at the device, which should be fine for regulating down to 3.3V using an ESP32 board's onboard LDO.

Keep in mind that these are slightly conservative estimates that don't leave you much wiggle-room for operational variances, so I'd recommend trying to find a slightly thicker cable. Even 20 AWG would provide a marked reduction in losses, getting you closer to 100mV drop over a meter.

• That would make sense for me. I am thinking of 18 AWG what I found on ebay. Cant find 16 AWG. I guess its too big for connector. What about idea using both available connectors USB-C/A from powerbank => 2 cables to power one PCB board? Would that make big improvement or not? – user1085907 Jan 21 '20 at 19:47
• @user1085907 you'd half the resistance and hence the voltage drop. Seriously, your approach isn't that great. There's a reason why power transfer over distance always happens at low current but higher voltage; the power losses are lower. So, step up your voltage, and use a step-down converter at the receiving end. It's exactly what PoE (Power over Ethernet) does. – Marcus Müller Jan 21 '20 at 20:52

Your approach isn't that great. There's a reason why power transfer over distance always happens at low current but higher voltage; the power losses are lower.

So, step up your voltage, (so you need less current to transport the same power) and use a step-down converter at the receiving end. It's exactly what PoE (Power over Ethernet) does.

That gives you the ability to use cheap thin cabling – without having to worry too much, because the lower current means less voltage drop, and with the step-down converter on the receiving end, you always get a well-regulated target voltage.

• I agree that approach isnt great. In my case I unfortunately cant have step-up regulator at powerbank side :/ Not enough space. That's why I am solving it via stronger cable. Operational time is about few hours per day – user1085907 Jan 21 '20 at 21:35
• you overestimate the size of step-up converters. It's also possible to have, say 50cm of thicker cable towards the step-up, and then the thin cable to the target. – Marcus Müller Jan 21 '20 at 22:05
• I'll try to make onboard step-down regulator via MP2225. Powerbank can deliver 12V if I emulate quickcharge via ESP32 – user1085907 Jan 22 '20 at 8:08
• can't simulate quickcharge reliably over 20m of thin cable. – Marcus Müller Jan 22 '20 at 11:47
• well I am gonna simulate it over 1.5m of 20 AWG cable – user1085907 Jan 22 '20 at 21:44