They used an iCharger so it’s DC. My theory is that they had high contact resistance between the charger and wire on the high strand wire and that generated more heat, but I’d love to repeat the same test to see if I can replicate their results.
I do find the temperature curve interesting as well. Amp ratings for wires are different depending on what source you look at because they are based off a specific working temperature which can vary depending on so many conditions. That chart showed that the temperature levels off at some point, so it may be possible to create a wire gauge amp chart that shows what temperature a wire gets to at a given current. It would allow the end user to be more flexible in wire selection.
That’s not my experience. There’s no skin effect for our use, or most uses. Any difference in surface area doesn’t affect things much when the conductors are tightly bundled together and wrapped in an insulator (temperature and electrical).
Both should have the same amount of copper with the same net resistance that creates the heat.
Some things don’t make sense, once you get past the extraordinarily useless color choices they made in that graph:
They show a >40°C difference for two types of stranded wire but “only” a 30°C difference or so between the high strand count and solid wire?
The solid wire plot starts at about 10°C higher than the other two. There seems to have been some lack of control of the test conditions?
These kind of huge temp conditions are something I would expect from comparing 12AWG to 18AWG, not from comparing different types of the same gauge.
A link to the full version of the chart is in my post above and even the author was confused by the results. It was a highschool/college lab project and not a high quality peer reviewed study. I’d love to try to replicate it if I get some time, unless anyone has a better source to show temperature differences between various types of copper wire when gauge and amps are the same.
Thanks, I missed seeing that link.
I think the results were badly affected by the shape of the sensor and the thermal resistance of the contact patch between the sensor and the different wires.
If a thermistor bead was used then that rounded shape would have a different level of contact with the different types of wire. But I would have expected the solid wire to have the lowest temp reading.
The stranded wire reading the hottest makes sense though. All those strands would allow the sensor to nestle in them, increasing the surface area of copper that the sensor touched. The contact patch between the sensor and solid wire would be much smaller.
IMO, we can’t make any conclusions from that testing. The results are inconsistent and don’t make sense. I feel it was a big mistake to post the results before a retest was done to confirm the first set of tests.
Use a thermally non-conductive surface and bare wire of a fixed length.
Constant current power supply so connection resistances don’t affect anything.
Use a type-k thermocouple, or other tiny type, with a dab of thermal paste to help increase the transfer of heat.
Twist the stranded wire tight so the sensor rests on it the same way it does on the solid wire.
Kapton tape the wire down except for a section in the middle. Kapton tape the sensor onto the bare wire in the middle (with a thermal paste blob on the sensor).
No moving air in the room.
Let the wire “heat soak” until the temp hasn’t changed for at least a minute.
Test all three types of wires and then go back and retest all three. If the results are different by more than a degree of two then you were not controlling the variables well enough and you need to retest at least two more times. This is critical. The tests cannot be done just once for each wire. A two-channel temp meter can speed this up considerably as two points on the wire can be tested at once.
If using a thermistor (which are typically huge compared to type-k thermocouple) then max contact with the wire is critical. But the level of contact must be the same for all three wire types. Let the temp stabilize for at least a couple of minutes after it appears to have stopped changing since thermistors are so slow to respond.
A better idea might be to use a fixed length of the wire, coiled up like a spring without any coils touching, and immersed in oil or even water with the temp probe. All of the heat would be captured by the oil and there aren’t any worries about the size of the contact patch between the sensor and the wire.
This wouldn’t tell you the wire temp but it would tell you if there were any temp differences. A well insulated container with a cover is needed, with as small a volume of oil that will do the job, being used.
An important thing to remember with all of this…different wire types of equal gauge will all be heated up based on their resistance and the current flowing through them.
Just look up their resistances or measure it, no testing needed. The higher resistance wires will get hotter (if insulated). Bare wires might heat up differently since their differing surface areas can affect how they shed heat. But I don’t think surface area makes much difference when wrapped in an insulator.
Wouldn’t the skin effect come into play in phase wires? If you’re cruising at near top speed the phase wires are effectively at 10,000 Hz AC or more, right?
It’s pulsed DC current, timed to simulate AC current’s back&forth flow but still DC current. There is no skin effect. The type of motor gives us a hint…BLDC.
Pulsed DC has characteristics more similar to AC than DC AFAIK. Also maybe I’ve not understood, but I thought FOC brought the shape of the pulses supplied to the motor into more of a sinusoidal shape. The BLDC motors we use could just as easily be called 3 phase AC motors. They are the same thing.
Random thoughts…
