Fuse/Breaker Recommendations

[ianlee74]

What are the recommended fuse/breakers and their placement in relation to a RoboClaw 2x60A motor controller?

Thanks!
Ian

[Basicmicro Support]

Choosing a fuse or breaker can get very complicated if you want to protect the system as much as possible. You have to take into account the actual current the motors will pull at a given duty cycle. Then exstimate the maximum current and for how long it will need to pass that current before fusing/breaking. This is probably not what you want.

The simple method is to look at your power sources maximum current sourcing capacity. For example a 20C 4500mah LIPO battery can source 90amps(20 * 4.5a). In the simplest situation the fuse/breakers job is to prevent fires. So you can assume that something bad has been damaged on the controller if the fuse/break needs to operate(eg something is dead shorted). Then you can assume the maximum current will be drawn from the power source. Pick a fuse/breaker that will open below that maximum current sourcing point. Look at the fuse/breaks specs specifially, not just rated amps for the fuse. You need to look at the timing table that shows how long it will take a fuse/breaker to open at a given current. Pick one that will open at the maximum power source current in no more than a few seconds. This will assure it will always open in the event short(though slower) even if the power source is partly or mostly depleted).

Note that if this maximum current is less than the combined total the motors can pull at 100% duty(and stalled), then you may need a bigger power source and a higher current limit fuse/breaker. It will depend on if your application ever needs to pull that much current for that amount of time or longer.

Note the battery current is a function of the motor current(s) and the motor duty cycle. For example, a motor pulling 30a at 25% duty is only pulling 7.5a from the battery. This is because battery power can always be assume to be coming into the controller continously(eg 100% of the time).

[ianlee74]

@acidtech - thanks for the detailed reply. I should have known there wouldn't just be a reference design to use as a guide I'm going to have to dig around a bit to figure what I need according to your recommendations.

However, the question that prompted this post and one I'm not sure I still understand the answer to is this... Let's just assume that I'm using a 2x60A RoboClaw and each motor pulls say 45A (total of 90A). So, a minimum 90A fuse would be needed. However, since the RoboClaw only has one supply how could I protect each channel from receiving more than 60A and damaging the controller? I understand that putting a fuse on the motor side of the controller can cause damage to the controller. Correct?

[Basicmicro Support]

You only need to protect the main power source from the Roboclaw. The Roboclaw has its own current limiting control for each individual motor.

Yes, worse case two motors that can pull 45a(at stall) can pull a maximum of 90a from the battery(assuming 100% duty on both motors).

Let me try to simplify this:
1. Find the maximum current you will possibly need to run the motors(eg 45a X 2 in this example case).
2. Find the maximum current the battery can supply(with Lipos its mah * C rating). Lets assume 4500mah and 45C. Thats 202.5 amps. In this case I would probably use a 100amp fuse. It leaves you some margin on your motor current draw but should still fuse very quickly if something shorts out. In this example you probably dont even need to look at the fuse timing table.

[Chuck-Yeo]

Base on the information in this discussion, I assume that 2 motors with a stall current of 20amps on each channel will require a 50amps fuse. I guess that means that all my devices (relay switch, low voltage disconnect, etc.) along that path will need to be 50amps and above as well.

Does the following circuit make sense or needs improvements? I just want to be safe.

Wiring v0.2.jpg (69.08 KiB) Viewed 9414 times

[Basicmicro Support]

Unfortunately it is much more complicated than that.

1. Motors pull power from the battery, not just current. So a motor pulling 20amps at a 50% duty cycle is only pulling 10amps from the battery. Why? Because the duty cycle effectively reduces the voltage at the motor based on the percentage of duty. So, for example, when you are using a 12v battery, a motor runing 50% duty pulling 20amps is only seeing 6v which is 120watts. The battery sees 120watts pulled from 12v which means current is 10amps at the battery. At very low duties with very high stall current motors you can get to crazy current levels at the motor while the battery is barely sourcing any current at all.

2. Fuses do not blow at their rated current. A 50a fuse will not blow at 50a. Ever. To see when a fuse will actually blow you need to look at the fuse table which will state at different currents how fast that particular fuse will blow.

So how do you pick a fuse.

First you must realise that the fuse will not protect the motor controller. It may help prevent unrepairable damage but it will prevent damage. The fuse is there to prevent fire. pure and simple.

So the fuse requirements must meet these two rules:
1. The fuse must NOT blow under any load form both motors combined.
2. The fuse MUST blow under a near dead short of the battery(eg if the controller is damaged this is what the battery will see)

The worse case current draw from the battery under normal operation will be the motor stall current at 100% duty. This is just the stall current rating of the motor. Unfortunately alot of motors dont list this rating(see ebay motors for lack of examples) .

So you need to determine the motors max stall current. If you have a variable power supply and a multimeter its fairly easy. Set the power source at 1v and connect it to the motor through the multimeter(with the meter in current reading mode). Multiple the current reading by the batteries fully charged voltage. For example current at 1v = 1.2amps then stall current for that motor on a 12v battery will be 14.4v * 1.2amps ~= 17amps. The 14.4 is for a 12v lead acid battery at full charge. note different types of lead acids have different peak charge voltages.

So two of the above theoretical motors will pull a maximum of approx 34amps. The fuse needs to never blow at this amperage.

Second rule, the fuse MUST blow if the battery is nearly dead shorted(eg has a few tens of milliohms resistance across the battery leads). If you are using a car type lead acid this will be in the hundereds of amps range. if you are using a deepcycle lead acid this could be only 80amps. if you are using a lipo its the C rating times the milliamp hours. Eg 4500mah * 20C = 90000 milliamps(90 amps).

If the difference between the minimum current and maximum current of the motors/battery is large you can just get a fuse rated for somewhere in between. However if the range is narrow(40a must not vs 80a must blow) then you will have to read the fuse table to find a fuse that will definitely blow before 80a but will not blow below 40a. I recommend you get a fuse that will definitely blow near the batteries max current in .1 to 1seconds. Anything more and if something does go wrong with the controller it will slag the controller beyond repair before it blows, though it should still prevent fires. If you have a narrow range you may have no choice but to use a fuse thay will take tens of seconds to blow.

There is one other factor you can take into account. You can get a slow blow fuse that will allow short periods with the motors at stall but would blow after a few second. This may be a good choice but keep in mind if the fuse blows while the motors are running you need to provide a path back to the battery +( via a power diode across the fuse) otherwise the regen fromt he motors may cause a voltage spike many times higher than the maximum voltage the controller can survive.

Same goes with the power switch you use. If you can switch power off while the motors are running you need to provide a path back to battery + when you switch off for regen.

Note you may think putting the switch/fuse on - would be a good idea then but that is not a good idea. If the ground is disconnected while running it is very likely the electrons will just find(or make) a new path to ground. That could be right through your nice Intel Nuc .

[jwatte]

Additionally, motors don't like being stalled. They don't like being run at close to stall power for too long. They will overheat. Good motors have a specification for duty cycle and "normal" amount of current (which is typically 25% of stall current, or lower!) Bad motors don't have this specified, and thus you should assume 20%.

Higher-precision motors that are smaller, higher power, actually can take less of stall, because the tolerances are smaller and the heat expansion damages them sooner. I've blown expensive Maxon 12V servo motors when stalling them for just a few seconds at 16V 4S LiPo. I've also blown Pololu 12V 37D 34:1 gear motors when running them on heavy load for a long time -- they ended up overheating, and once cooled, didn't come back.

To avoid motors burning themselves out, you want to limit the total current going into them, but you also want to limit the square of the current times time that goes into the motors, because that means how much heating energy you are actually putting in.

You may end up wanting electronic fuses -- some current sensor, comparator, and MOSFET can do that, and can serve as an electronically controlled "soft" power switch as well. The draw-backs are that it's more cost, and if you get it wrong, it won't protect you after all

[Basicmicro Support]

i2t is probably a bit overkill for hobbiests and most of the motors they have access to arnt going to have the specs to calculate the proper limits for i2t but jwatte is right in general. And I would say in general motors are designed to run between 1/5th and 1/10th stall current continously so even 25% is being generous.

Maxon are particularrly easy to damage if you are using their coreless motors. Coreless motors dont have a metal core like the cheaper regular DC motors so they can over heat very quickly. Coreless motors usually have resin based scafolding to hold the wire wrapped on the axel so they melt very easily since there is no where for the heat to go.

[Chuck-Yeo]

Thanks acidtech for the very useful and detailed information and jwatte for the additional information. Even though I am not 100% sure still about what to do. I definitely have a much better understanding on this topic compared to days of searching the internet for such information.

acidtech: To determine the motors max stall current using 1v as you have mentioned. I assume you need to measure the current by stalling the motor as the current when it was free to spin at 1v was very low. I get around 17A for one motor as well. The specs for the motor rate it at 20A.

I definitely want to use a flyback diode as you have recommended. I would not want to fry my RoboClow controller or my Intel NUC. I am using a 10A10 (10A, 1000V) diode. A little too big I think but the next diode I have on hand is the 1N4001 and 1N4004 which is only rated for 1A.

Questions:
1) Any harm in using such a big diode?
2) Instead of using a diode at every possible breaking point (fuse, switch, low voltage cut off, e-stop).
Can I just connect a long separate path back to the battery from the RoboClaw positive terminal with a single diode in between?

[Basicmicro Support]

Yes. You need to lock the axle on the motor when running the 1v test. Most motors wont spin at 1v so I didnt think about mentioning that.

The diode should work fine. With that small a motor you probably could use the 1N4004 but I'd stick with the 10a one since you have it already.

1. The diode size shouldnt matter as long as it is a regular diode/rectifier, not a zenor or something exotic.
2. I would have the switch and fuse in series as close together as possible and then have a single diode parallel to both.