It’s always kind of resistor
About the discussion above between 100-1000W rated resistors, those are the continuous rated power under sufficient cooling conditions. A 100W resistor will burn itself out from the inside if it’s subjected to for example 200W for long enough time, because it simply can’t move the heat from the resistive element inside to the heat mass outside efficienctly enough.
But the resistor itself doesn’t need to be that huge, but rather have good heat transferring capability to a heat mass, which is preferably cooled. I believe Mellow is using their whole drive unit as a heatsink.
Could be a MOSFET run in the ohmic region as well. Behaves like a adjustable resistor, so the braking power can be controlled.
Could be a resistor, a diode, a light, a heating element, another battery, a supercapacitor, a motor & propeller
Never assume 
Actualy you just assumed those components have no resistance 
2 Likes
But talking about supercaps, yesterday I bought mini screwdriver that has 4.6V 320F cap inside. It’s roughly 400mAh if my calculations are correct but pretty large charger that’s in a set charges it within 60 seconds. Also I don’t know how much of this energy can be used but I suppose that it might be more than in battery where you have to care about not destroying internal chemistry.
At what nominal voltage?
Capacitor stored energy: E = 0.5 * C * U^2 = 3240 Ws or J, if charged to 4.5V 3240 Ws = 0.9 Wh Li-ion equivalent capacity, using nominal 3.6 V → 0.9 Wh / 3.6V = 250 mAh
I used formula I found on arduino forum: 320F 4.6V capacitor can store Q=CV=1472 Coulombs of charge, 1mAh is 0.001 Columbs per second (0.001A) multiplied by 3600 seconds or 3.6 Coulombs. So the capacitor should be equivalent to 1472/3.6 = 409mAh
I’m talking about the stored energy inside the cap, which we can then compare to a li-ion cells nominal voltage capacity.
http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capeng.html
Anyhow, how does the driver feel? How much can you do before the voltage drops too low?
1 Like
That’s a shit-ton of capacitance. Can’t believe that says F and not μF
1 Like
I have no rig to test it but feels good, no difference between regular battery operated one except capacity and charging time. I can compare it with regular battery powered one of similar power later. But the only test i can think of is putting screws into pine board until battery dies.
Going back on tracks I think that easiest method would be relatively short wire heater (the one you see on top in your oven) on some kind of ceramic insulator in a cage. It can get extremely hot and can eat 2000W easily. This is also method used to stop small wind turbines.
1 Like
Well WOW I wasn’t expecting such a strong feedback on this thing 
@lock Mellow is ahead in the future ! That’s how every damn skate should be setup (I mean safety and quality wise). I got a bunch of their battery connectors, RoPD connectors 40A, these things are nice ! Waterproof and magnetic at the same time, easy to mount and impossible to mismatch.
Supposedly a 500W break chopper might be enough to burn away excessive loads, given you can also aerodynamically slow down with your body. Spread arms, stand higher a bit, it will eat away them watts.
@webst which kind of ceramic cage were you talking about ? Which length of heat wire ? With a good placement on air pathway can do, I guess it should be put away from the ESC too as much as possible.
What about a string of smaller resistors or wires mounted on a heatsink with thermal compound ?
1 Like
I was thinking about this for a moment and then had a thought that when breaking the voltage rise on the bus voltage not in the batteries itself. I think cutoff should be operated on cell level using BMS which can sense each cell voltage than the whole bus. Because then you brake you generate a voltage which is not instantly consumed by battery, it is like charging a battery with higher voltage when it is. So for voltage cut off, I think cell voltage would be more appropriate.
For e.g. you brake for 10s with 30A power u battery pack is 42V 10Ah so you generate 42V * 30A = 1,2kW for 10s which is 3.33Wh.
So if I understand correctly in those 10s you generate less than 1% of battery capacity which is also not all energy going into battery considered all the loses and other consumptions on the circuit. So even when you apply a higher voltage with current I don’t think it would overcharge the battery but just rise the bus voltage which is kinda ok. But it’s just my 6-year-old physics thoughts on this problem.
P.S. I can be really wrong on this. It’s just my assumptions 
tl;dr Don’t make cut off based on bus voltage but on summed cell voltage.
The big problem is that the cells don’t like fast charging, even more at a high state of charge
Mellow use the braking resistor not only when the battery full, but every time the current going back into the battery is greater than a set level for a given state of charge
1 Like
Cells don’t like fast Charging because of IR. Cells will absorb current as quickly as they can, when the current cannot be absorbed it turns into voltage spikes. Lower IR cells absorb more amps, just like lower IR cells give off more current easily vs higher IR…
As I stated in another post…but it was about antispark. More capacitance would solve this. But it would make the spark larger if you don’t have an antispark…
I also found solution little closer to what we need, I’m unable to evaluate if there’s everything fine with this rather uncomplicated circuit:
http://vedder.se/forums/viewtopic.php?f=15&t=586&start=10#p3635
Edit: update on my last link “This energy dump is intended for running a motor on the lab bench, from a lab power supply that cannot absorb regenerated energy. It would also allow you to do regenerative braking even though your batteries were fully charged if used on a vehicle (and if it was carefully dimensioned for the load), of course.”
1 Like
I thought about it for some time and come to following conclusion. All this heating element idea is a little low tech. It would be much more elegant to redirect excess power to electromagnets surrounding copper discs connected directly to front wheels. That should induce eddy currents stopping the board. This would unfortunately need new esc construction with additional circuit dedicated to this task but I like the idea of frictionless brake.
Would actually be some next Gen braking safety, but required circuitry seems a bit complicated and steep price. First let’s get the brake chopper as reliable and as affordable as possible.
If it is well done and quick enough to trigger the protection, it could even lead to safely play closer to the DRV absolute max (if even transient voltages spikes can never cross the protection, you’re good to play with 13-14S worry free).
I like brake chopper idea but DRV problem lies in compromises taken during vesc design and making that kind of prostetics would not be worth it. Imo the only proper way to deal with it is getting 6th gen that has much higher limits. Nevertheless I don’t think that any of solutions provided can be added with simple switch, well actually there is one: when the voltage goes above 4.2 per cell you’d open this brake chopper circuit assuming it can work as dump load. This should be theoretically done with some minimum effort. Other solutions involving going directly with power into some load would involve esc that can handle this as something has to redirect this power to another place while you remain control over the board.