Ampacity of copper busbar?

[jonescg]

Hi guys,

Quick question. What is the rule of thumb for current carrying capacity of copper bus? I have seen lots of household wiring definitions which are very very conservative. Warming up a bit is OK, getting hot is not cool. I figure 10 amps for every 1 mm2 would be fair for an EV application? Sounds fair? Excessive?

 

[Kingfish]

I use this site for a conservative reference: AWG Table

Take the Diameter (d) in mm, and use π(d/2)^2 to get the area.

Awe heck, here - I've charted it out for you along with the "Maximum amps for chassis wiring":

As you can see, it's a typical inverse-square.

Enjoy, KF

 

[Nuts&Volts]

The larger the surface area the higher the amps you can run through it because the heat can get out quicker. So flat bar can handle more amps than a round wire for the same cross section. An uninsulated piece can handle even more amps. Experiment and see if it gets too hot.

 

[BatteryMooch]

Hi guys,

Quick question. What is the rule of thumb for current carrying capacity of copper bus? I have seen lots of household wiring definitions which are very very conservative. Warming up a bit is OK, getting hot is not cool. I figure 10 amps for every 1 mm2 would be fair for an EV application? Sounds fair? Excessive?

Voltage drop is also something to consider. Often, the voltage drop (i.e., power loss) is much more of a problem than the heating.

And so much depends on what the ambient environment is for the bus bars in question. Are two of them next to each other? You need to derate them then. Are they insulated or in a case? You might need to derate them. Are they exposed to moving air? Now the current rating might double (or more) for a given temperature rise. But, the voltage drop will not change much so you need to still take that into account.

I'm not sure you can derive a rule-of-thumb for this, even limiting the application to EV's. There are just too many variable and a huge range of limits (i.e., max temp rise over ambient, max voltage drop, etc.). One person's absolute max is another person's everyday norm.

I'm with Nuts&Volts, you'll need to pick a bus bar and experiment with your specific setup. If you aren't sure even where to start, pick a bus bar you can double or triple up with. That way it's easy to parallel multiple lengths of copper when needed.

An example...I was working on an ultracap module that was required to handle 400A surges. There was no way to calculate what thickness of copper sheet interconnects we needed to hold down temperatures and voltage drops without testing. So, we used 40mil thick Alloy 110 copper sheet (cut to shape) and tested the setup with single, double, and triple layers of copper. The triple-layer was about what we'd get with 1/8" thick copper and we felt that would definitely (we hoped) be enough or overkill. This allowed us to set an upper limit for thickness. We then selected easily available sheet that we could stack up for easy testing to a max approximating our 1/8" limit. Luckily, a single sheet of 40mil was all we needed.

 

[jonescg]

My question really relates to this problem:

http://www.endless-sphere.com/forums/viewtopic.php?f=34&p=625567#p625567

Is 3 mm copper bus overkill? I think it would be fine because there would be a stack of brass underneath the tabs, and a fair bit of airflow between them.

 

[BatteryMooch]

I checked that thread and couldn't see how much current you want the bus to handle....verrry important.
I might be the wrong guy to ask if something is overkill as my designs are typically very conservative, high-reliability, setups. But, if you feel that 3mm might be overkill, why not drop down to 2mm (or whatever is available) and test? If it runs too hot (or the power drop is too high) you can easily double up the bar thickness with a second piece.

 

[jonescg]

The inverter maxes out at 450 A DC (300 A rms motor side). However, I can't see me getting remotely close to this most of the time. The narrowest bit is about 12 mm, so at 36 mm2, it's probably fine. I have pulled 500 A through insulated 50 mm2 cables on Voltron and it was fine - probably warmed up a bit, but not hot and floppy in any way.

 

[Hillhater]

...we used 40mil thick ....

Could you translate "40 mil" ,..into inches or millimeters
( i am guessing you mean 40 thousands of an inch 0.040" ...or 1.0 mm ?)

 

[BatteryMooch]

Yup...40 mil = 40/1000 of an inch.
Sorry for the confusion. We use mils quite often to prevent confusion when many sizes/measurements are being discussed. It really does help prevent misreading of numbers, e.g. '40 mils' instead of '0.040"' or '40/1000"'.