Hot phase wires...how helpful is thick wire outside only?

[John in CR]

Hubmonster's phase wires, which are larger than 12ga but smaller than high strand count 10ga, are getting too hot for my taste. The controller has doubled up 10ga on each phase, but the motor wires are 3 ft of single cable. I want to try to avoid a reharness, at least for now. I've seen some guys run big wire all the way to right before the motor. I take it that this not only avoids generating the extra heat in that length of wire, but also sinks some heat away from the short run inside the motor to the windings. Is that approach beneficial enough to be worth the effort, or do I really need to just bite the bullet and take the motor down?

I'm always resistant to opening a motor unless I really have to due to risks of water ingress after breaking the factory seal, and more worrisome, unknowingly creating a short putting the cover back on an enlarged harness.

 

[knoxie]

Hi John

I do exactly this on my hub motors and it does work and has avoided me having to rip out the original wires, its simple ohms law v squared losses are directly proportional to the current the voltage and the length of the cable, you reduce any of these factors and you reduce the losses, OK albeit if you increase the voltage the effective current will decrease for an equivalent power :lol: but I digress here.

If you hunt around for the various power loss calcs online you can work it out for yourself, if you look at the size of a 40A clear cartridge fuse for instance, it can take 40A before it blows because its very short in length, this is simple example but this is the basis of the theory, I dont get any major cable heating in the thin phase wires on the BMC motors and can run them at 1.6KW because the section of wire is only 6 inches long.

 

[John in CR]

Thanks Knoxie. I forgot that you were the one using that strategy. I understand the reduction in overall losses, but your short section of thin wires still has the same loss per unit of length, so wouldn't it get to the same temp unless the thick wire is acting as a nice copper heat sink too?

 

[wesnewell]

Someone please correct me if I'm wrong, but phase wires are only hot (with current) one third of the time. Thus for a 60A controller, that's similar to 20A per phase full time. what size controller are you using that larger than 12 gauge wire isn't large enough?

 

[John in CR]

Someone please correct me if I'm wrong, but phase wires are only hot (with current) one third of the time. Thus for a 60A controller, that's similar to 20A per phase full time. what size controller are you using that larger than 12 gauge wire isn't large enough?

Actually it's 2/3rds of the time because the return route of the current flow is backward through another phase. That's not of help though, because it's all about the current. Due to the pulsing nature of PWM to accomplish current limiting and change the apparent voltage the motor "sees", the current flow from the batteries is chopped up into short pieces, so the phase wires see much higher currents, sometimes many times higher.

Though I haven't measured the current, the battery side max current is 120-140A, so the phase wires see as much as 300-400A.

 

[spinningmagnets]

I think that this would help a lot. Not only is the thicker wire outside the motor acting as a heat-sink (as you have mentioned), but...the wire outside the hub is also creating current-resistive heat right now while it is still the smaller size. Put on the fattest wire you can afford. You are removing a source of heat generation, and adding something that will pull a measurable amount of heat away form the short remaining section of the thinner wires.

 

[dogman dan]

though a known dummy, I agree. Certainly harmless to thicken up the wire to within a foot or less of the axle. For sure you need to do it if those wires are getting hot at a point 2-3 feet from the hub. At that distance it's not likely to be heatsink heating from the hub itself.

 

[Philistine]

what size controller are you using that larger than 12 gauge wire isn't large enough?

the battery side max current is 120-140A, so the phase wires see as much as 300-400A

I remember I had a mate who once casually dropped in conversation about how he had a 10 inch dong, and he couldn't find condoms that fit (he asked me if I ever found that condoms "hurt"), I explained no more than my socks hurt my feet.

I simply never believed him, until one day curiosity got the better of me and I finally asked to see it.

I totally know you are telling the truth John, I am just saying I felt inadequate all over again....

Sorry to go OT.

 

[www.recumbents.com]

I replaced the phase wires on my BMC 600 up to about 5 inches from the hub. I had noticed that the phase wires were very hot after a ride and was concerned about a meltdown. I replaced the wires with wires I removed from a huge 220VAC power cord from rack equipment. They are probably 12 Ga. Anyway, no more hot wires, even on the short section by the hub, and I got another couple MPH on my cruising speed.

-Warren.

 

[knoxie]

Hi John

Yes it does make a difference as the losses are proportional to the length, the shorter you make the cable the more current it can take because it is dropping less volts and therefore heat through that section of the cable, it has certainly worked for me, the hubs I had from the box had a very long length of cable on them, I chopped this right down on all the bikes and they have been fine all of them been running for 6 years like that, so anyone who owns a BMC you dont need to change the wires if you follow this method and keep the power around the 1.6KW mark MAX and ride sensibly.

 

[neptronix]

^-- yep

Think about it, if we have 4 feet of wimpy phase cable and replace 3 feet with super thick stuff, we have reduced our phase resistance by almost 3/4th.

It's literally that simple.

 

[John in CR]

^-- yep

Think about it, if we have 4 feet of wimpy phase cable and replace 3 feet with super thick stuff, we have reduced our phase resistance by almost 3/4th.

It's literally that simple.

If you're worried about overall losses or voltage drop, which I am not, then I agree. I'm thinking of it more like a hot wire styrofoam cutter, where you have low resistance wire going to the cutter, and high resistance on the cutting wire so they get hot enough to cut. ie Cool wires leading to the cutter and hot high resistance wires. I'm counting on a different result, because everything is copper which conducts heat so well. I want the thick wires not to just generate less heat, but also to suck some of the heat out of the short thinner segments. Those segments will create the same heat as before, with the same heat dissipation other than what is conducted out to the cooler thick wire outside.

Knoxie's results are good enough for me to count on the heat sinking action of the larger wire that I seek. 6 years of experience trumps anything in my book.

I've decided to splice 3 equal lengths of 8ga wire in parallel with the existing phase wires. That not only reduces resistance, but I also pick up more surface area for better cooling than as a single cable. My main concern at this point is the shortest copper-to-copper joint in the connection. I don't want the heat to pass through a thick layer of solder before getting to the copper conductors, since it doesn't conduct heat nearly as well.

Thanks guys,

John