1
\$\begingroup\$

I am designing a PCB board that will combine 6 cells in series. However, I am thinking this might not work because of the amount of current that is required from those cells. My battery has to be able to provide up to 100A for few seconds, and 50A for continuous.

Here a quick layout I did of what it will look like: enter image description here

The yellow plane (5mm in width) is what will be connecting each cell in series, however with most calculators it is asking for at least 55 mm with a 20C temp rise. I can't go higher on the temp rise because this will be a battery board and don't want my battery to heat up. I also can't make the board any bigger because of my mechanical requirements. However, I can go thicker with more layers, but I have no idea how to calculate how many layers I will need to make this connection work.

\$\endgroup\$
  • \$\begingroup\$ For multiple layers, nominally just sum up the width on each layer. 4 layers each at 5mm would be 20mm. \$\endgroup\$ – Tom Carpenter Jan 3 '18 at 17:46
  • \$\begingroup\$ What is about the section measurement of the conductors? I do not see consideration of thickness of the layer. Would it be better from safety perspective making the assembly using wired P2P and not PCB holding conductors? \$\endgroup\$ – Anonymous Jan 3 '18 at 17:54
  • \$\begingroup\$ What kind of batteries are these..... and are you sure you an pack them that close together at those currents? \$\endgroup\$ – Trevor_G Jan 3 '18 at 18:41
  • 1
    \$\begingroup\$ This is no way to design a LiPo pack. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Jan 3 '18 at 18:45
  • \$\begingroup\$ @Trevor The are LiPo packs that are new to the market. The are pretty thin packs, the board does consider tolerance, so they should fit. \$\endgroup\$ – J. Jones Jan 3 '18 at 19:02
1
\$\begingroup\$

Why don't you just flip all the odd cells?

If you flip them, you can connect them with a much shorter and wider path.

You say the yellow traces are 5 mm wide, I estimate they are 15 mm long. If you keep the exact same landing pattern, but flip every other cell, you can connect the cells with traces approximately 20 mm wide, 3 mm long. That is like 20x better (lower) resistance.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ We actually thought about this, I believe we didn't do it because of the way the cell are constructed, the tabs wouldn't allow us to do it. We made a mock up of it actually and it was a manufacturing nightmare. \$\endgroup\$ – J. Jones Jan 3 '18 at 18:56
  • \$\begingroup\$ The trick is to add a solderable extension to the tabs. Some cells can be soldered only on one side of the tab. \$\endgroup\$ – Lior Bilia Jan 3 '18 at 19:20
  • \$\begingroup\$ Well alternating the cells is what the manufacturers do in Ns batteries. I think it is doable and is the best option in general, but for your particular cells/mfg process I can't be so sure... \$\endgroup\$ – Vladimir Cravero Jan 3 '18 at 21:51
0
\$\begingroup\$

I would add more layers and make them solid planes. Use the heaviest copper available at least 2 oz. What are the dimensions of your board? That will be a factor in how many layers you need. Another option would be to add large holes to solder thick wires or copper strips to carry the current.

| improve this answer | |
\$\endgroup\$
0
\$\begingroup\$

Using this calculator (http://www.4pcb.com/trace-width-calculator.html) it looks like if you use 10 amps and 2oz copper, you need 3.6mm for a 20C temp rise in air. Using the top layer and bottom layer should result in approx 10A passing through each layer, so 5mm should be adequate. It might be good to put some vias in, but not too many otherwise it will start decreasing the conductivity.

Another solution is using bus bars (http://www.epectec.com/pcb/powerlink-technology.html or https://uk.rs-online.com/web/c/automation-control-gear/circuit-protection-circuit-breakers/busbars/), or solder coating.

| improve this answer | |
\$\endgroup\$
0
\$\begingroup\$

At these currents, and such a simple circuit, use discrete copper wires, and a separate mechanical arrangement to hold the cells.

There is little to be gained by putting the high current terminals on a board. You will also have trouble soldering those large terminals, while not overheating the battery terminals.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Yeah, this was another issue with having lots of copper causing us to overheat the battery tabs. I have a very large soldering iron that I did some test with and was able to quickly solder the tabs to a board in a matter of few seconds. However, this was done with a board with 2 layers and a very small sample. \$\endgroup\$ – J. Jones Jan 3 '18 at 19:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.