Longer battery wires

Hello,
So I was wondering if people have lengthened the wires from the main breaker and the PDB (or PDP as it called now) to the battery?

Andymark sells long (36") battery cables so I think it’s possible but can’t say for sure. As for connecting the ends of the wires, no idea.

You’re only allowed 12 inches on the battery half of the Anderson connector <R5>. You could, theoretically, use a longer cable on the other side and still comply with <R36>, but I would urge you not to do so. Every bit of current your robot uses goes through that cable, and more length is more resistance is more wasted power.

You can make the wires as long as you want. Just know that an increase in length increases the resistance in the wire run. Too long, and your robot may have brownout issues. Make the wires as short as you possibly can.

We lengthened our wires by soldering and heat-shrinking extensions onto the PDP side, in order to allow our breaker to be more accessible. I wouldn’t recommend more than about 2 feet of wire total between the two, just in case you run into issues with too much resistance.

I would not recommend soldering 6AWG. The wire acts as a heatsink and you might not get the solder to “connect” to all the strands, giving you a high resistance joint. It would be better to use a proper crimp lug and crimper to make joints in wire this size.

The resistance of an extra few feet of 6AWG should not cause brownouts. The table linked below shows Maximum Resistances in the range of 0.38 to 0.50 Ohms per 1000 feet. Passing an 100 A current through 1 foot of 6 AWG should only cause 50 mV of voltage drop.

A more likely cause for brownouts are bad joints in the power wiring. These could be bad crimps or loose screws or nuts on the lugs/terminals. I found several teams with loose battery cable screws over the last week. One had their radio drop out both matches they used the battery with the loose terminal screw.

*To put things into perspective:

6 CIMs pulling 40 amps each (it happens) is 240 amps. One foot of 6 AWG drops about .1V at 260 amps. That .1 voltage drop causes about 1.7% loss of max mechanical power.

Yes it does. Most teams would probably have a hard time noticing 1.7% power loss, especially if it is a transient condition.

One of the other posters had stated that a few feet of extra 6 AWG could lead to brownouts (which most teams would notice).

Absolutely. I had helped a team in San Diego that had about a 6 FT run of 6 AWG going from their main breaker to the PDP. Shortening the run solved their issue.

Battery voltage can be affected by three factors, temperature, state of charge and current. The one we’re really talking about here is current. A lead acid battery has internal resistance. As you move current through the battery, the voltage changes due to its internal resistance.

Lets simplify it a little bit.

When you’re charging a battery with some low current, you’ll see the battery voltage increase. Charging it with a larger current, you’ll really see it increase.

Same thing applies to the discharge. Discharge it with a small current, you’ll see it decrease; discharge it with a large current, you’ll see the battery voltage really decrease.

Now apply this same theory to having long interconnects with your battery. Say you’re drawing current at some C rating that’s near the limits of the battery, your voltage will likely be pretty low. Combine that with the voltage drop due to large current draw over a long run of cable.

This situation is extremely prone to causing brownout conditions.

I suspect the problem with the robot you helped fix was really some bad crimp or the nut on the breaker not being tight, leading to a high resistance connection (this is assuming that the battery did not get changed in the process). The drop caused by the wire resistance, even at 240 A through 6 feet of 6 AWG would be 0.72 V if we use the maximum rated wire resistance. A bad connection can easily give you more voltage drop, up to 12 V.

I am not advocating that we should have excessively long power wiring but we should solve the real problems.

According to Ether, 6 AWG wire loses .1 V at 260 A PER FOOT. Running an extra 6 feet would lose .6 V, or a 10.2% loss of mechanical power. This is something many teams would notice.

Question - why?

OK,
running worse case calaculations, a four motor CIM drive will draw ~480 amps at stall. 480 amps will drop .48 volts for every 2 feet of #6, please include the red and black wire in your calculations. The 6 foot run would be 12 feet of wire added to the two feet on the battery and about another equivalent 2 feet for properly crimped connectors and terminals. That adds up to almost 4 volts dropped in just the wiring feeding the PDP whenever you start driving or reverse direction. Add that to the internal resistance of the battery and you potentially have only 3 volts available even on a fully charged battery.
Jumping back to my easy calculator the WIRE FOOT, every WF will drop 0.1 volt at 100 amps.
Battery Internal Resistance=11WF
1 ft. of #6=0.5WF
1 ft. of #10=1WF
1 ft. of #12=2WF
When you view your voltage logs, those short dips to the 4 volt level you see are real. Make your #6 run as short as possible. Often you can significantly shorten the run by simply turning the PDP 180 degrees.

Fortunately, that wasn’t the issue. Their crimps were done well, screws and nuts were tightened down sufficiently on both the breaker, battery and PDP.

Edit: Al beat me to it.

I’ll take your word that the crimps were good and the screws and nuts are tight. What sort of motors did they have in their drivetrain? How much wire was removed from the loop in total?

My gut feeling is that if cutting out a few feet of wire made that much difference (you probably didn’t cut out 12 feet or even 6 feet if they started with 6 feet), they are probably still just above the brownout threshold of 6.8 V. It is likely that they still have some other issues such as batteries with high internal resistance and that it is likely that their brownout problems will return.

6 CIM drivetrain and a couple of minicims on an auxiliary mechanism.

The loop was 6 ft for both black and red wires (12 ft in total). When I told them to chop it off to a minimum, I came back and saw that they had just under 1 ft for both black and red wires.

The problem went away.

High current draw across 6AWG DOES result in a large voltage across a long span of wire.

Physics doesn’t lie unless you do the math wrong.

If they really are pulling that much current and had an extra 6 feet in both the red and black, it makes sense that it would make a difference. I am very familiar with how to do the math but this information was not in your original post so I could only make assumptions about your example. Most of the time, when I see a robot with funky wiring (including that much extra), it was built by someone who also does bad crimps, leaves connections loose and abuses their batteries.

But if there were only 3 volts at the motors, can they still pull 480 amps? I know it is a problem, but I don’t know if it is that bad.

No.

Remember Al said it was “worst case”.

In this case, the estimate is quite pessimistic.

You could create a more complete model, including:

  • battery internal resistance
  • PDB resistance
  • wire resistance
  • crimp resistance
  • connector resistance
  • voltage drop across motor controller
  • motor resistance

… and then do a more accurate calculation to determine the various voltages.

Guys,
What I am trying to get you to think about is this…

  1. All robot current flows through the #6 wire from battery to PDP.
  2. Any resistance in that loop will drop voltage at a predictable value.
  3. Worst case voltage drop will affect the voltage available for other systems.
  4. The RoboRio is the most important followed by the radio for keeping available voltage as high as possible.
  5. Regardless of your design, all motors draw stall current when starting.
  6. Brownout voltage doesn’t matter to motors if the control system has already died.
  7. The voltage dips you see in your logs are real as measured by the RoboRio. This confuses most people but those dips do occur. The log is not lying to you. WildStang did some extensive research into this more than ten years ago. We were able to compare video and voltage data to show that certain behavior leads to voltage drops. We had developed an app for Blackberrry at that time to put it all together and generate warning messages for our drivers at that time.

Yes, motors will draw less current at a lower battery terminal voltage but they will also likely stay at near stall longer as drivers push the throttles to max to get the robot to move as they want. Turns will also push the current to extremes if you have a lot of tires or tank treads on the ground.