@Kug3lis … ‘we’ are a company … not FlipSky.
They are a potential supplier. We are testing their FSESC 6.6 as a possible solution and also had them develop a custom remote.
Cheers
Seems to us it’s about watts, not just amps. Our tests show amps quoted by suppliers as not much use for our case of constant 48v with only amps varying. We go nowhere near the amps they quote (at least on the battery / inbound side) … see my previous posts.
Bottom line, it’s not just about constant amps when it comes to generating heat in the VESC. The constant voltage matters.
Cheers
no it does not (at least not significantly) - we had this discussion before.
3 Likes
@Maxid … Doesn’t make sense. Why would it be in the lab we can overheat all VESC 4.12s (including Maytech’s 100a one) with 48v / 30a constant from the battery in 5-10 minutes? Have you done any bench testing you can share or do you have any other reasoning as to why this would happen?
Cheers.
Because a 4.12 VESC is not capable of 30A continuously (irrespective of the voltage) - duh. I have a dual VESC 4.12 setup and get it to overheat in a couple minutes with a battery current of 40A in total (so 20 each) by just doing acceleration tests. Nobody cares because nobody actually needs 30A * 40V = 1200W continuously. I can throw the P=I² * R equation at you again but you don’t want to accept that.
Did you do the same 30A test also at 24V? I am sure the results will be the same.
4 Likes
@maxid In his video, Vedder has the 4.12 at 21v and 34a … says it will do this ‘all day’. Maybe argue with him first, then me.
Cheers.
So you didn’t do the test at 24V. Come back with your numbers once you have them. Cheers
Also which video are you talking about?
Just to give an idea:
1 Like
I’m not sure that I mixed it up, but yes, I am talking about battery current not motor / phase current. What we are trying to deal with is the VESC overheating.
In the video Vedder points to the watts and says ‘and dissipating X watts of losses in the system, so things are going to get really warm’ … so he’s saying watts (battery amps x input Volts) are relevant to heat. He goes on to say it can run at 34a all day but this is only around 50 watts from the battery.
As I’ve said before, our constant voltage is 48-52 per motor / VESC. Our minimum amps from the battery is 6 with an average of around 15 and max 30 but this could be sustained for several minutes on and off.
So, I can see what you are saying, but it’s not what we are focused on the as the motor / phase current is not a good proxy for heat.
Cheers
I’m late to this conversation, but I’m fairly sure the dominant contributor to the temperature of the VESC is the I2R (IxIxR) losses in the FETs. So this means the limiting factor will be motor/phase current.
This means if you are running Bat=50V@5A input to the VESC and it is driving the motor at 5V@50A. You will get 50x50xR (~50x50x0.01=25W) of heat in the fets. Note that I have made up the value of 10mOhm for the FETs.
Increasing the input power to Bat=50V@25A, but running the motor at a higher speed and hence Mot=25V@50A will lead to the same 25W of heat in the fets/VESC. So even though the VESC input power is 250W vs 1250W, the heat/temperature of the VESC is the same.
1 Like
Sounds to me like you really did mix them up. No wonder you thought voltage matters. You can’t use motor current for your calculation as that is not at constant voltage but varies due to the duty cycle. For easy comparison only use battery voltage and battery current when talking about how much current the VESC can handle. Vedders 34A is meaningless as the “real” amps going into the system are only 2A.
@AlexBE @Maxid … maybe our setup is atypical. It’s possible.
Our engineering partner took the various VESCs to a test lab. They put 30a / 48v into the VESC through conditioned power and had a motor drawing max amps running at constant voltage on the other side. With efficiency losses, slightly lower than 30 amps. This was to represent our worst case to allow us to work out what a customer who maxxed the system out until the battery was dead would do to the controllers … an unlikely situation, but one that needs to be allowed for.
Thanks. That is what we are using. Maybe I wasn’t clear.
Cheers.
But then you should have seen that voltage does not matter!
This may not be true when dealing with square waves. You can have pulses of 34A going to the inductive motor windings but drawing 2A into the filter capacitor from the battery – in this case, it actually is the 34A that will be expending energy according to the “on” resistance of the FETs and P = I²R, but even then only through the inductance of the windings
The only real test is empical and see how much the ESC heats up a known volume of water under a known load
I am not sure if it is that easy as i²r when you talk about an AC signal instead of the DC one coming from the battery.
It is P = I²R at any one instant in time. For the time-based “work” expended (watt*hours instead of watts, aka “degrees Centigrade times grams of water”), you need to get calculus involved. Or do an empirical test according to water temperature.
It still can’t be that the 34A are responsible for the heat loss as that would give you a situation where P coming from the battery is less than what the P in the fets would be.
In the example with Vedder (If I remember the numbers correctly he draws 70 motor amps and battery amps of 2).
So the power the VESC has to dissipate is 2² * r and not 70² * r. How should that work otherwise if you have a total of 50W going in but are supposed to dissipate more than that 
Because the phase wires aren’t on constantly, they are pulsing on and off, and, even while on, the voltage is fluctuating due to the inductance of the windings, and isn’t necessarily full battery voltage
Using your Vedder example. 70 motor amps means that the power the VESC has to dissipate is ~70x70xR (r=fet on resistance). This is why he does his temperature testing at 70 motor amps with only 2 battery amps. So he can temperature test the VESC without having to supply it with a huge amount of power.
Of course the power he has to supply it with has to be more that what the VESC needs to dissipate. Power(battery) = Power(VESC)+Power(Motors).
So making up some numbers. If we have Battery=50V@2A=100W. FET/MOTOR current of 50A. and FET on resistance of 0.01ohms.
Battery(50V*2A=100W) = VESC(50x50x0.01=25W) + MOTOR (50Ax(2-(50Ax0.01) = 75W)
You can see that with the motor current of 50A, you have a total of 2V across the FETs and motor windings. 0.5V across the FETs and 1.5V across the motor windings.
1 Like
No, all three phases are switching on and off at irregular times based on complex formulas. The battery amperage is the only somewhat steady flow. Think “calculus versus algebra”
Ye have to understand the difference between something happening right now, and something that happens now then turns off, then turns on, then turns off, and there are three of those things turning on very briefly at weird times controlled by complex formulas.
The definitive truth is in the BLDC firmware source code, feel free to have a look anytime
In this case, the battery current/voltage/power are fairly constant. On a VESC they are not directly measured but inferred from the phase current/duty cycle. I understand the firmware, it’s what I do for a living.
My point from my previous post still stand, my simplified power equation stands.
Power(battery) = Power(VESC)+Power(Motor)
So in general for VESC temperature testing, you need to increase the phase current, because that will increase the power heating up the VESC. You can do this in many ways. You could put a large load on the motor and run a very high battery power. Or you can use the tools built into the vesc to run an open loop with a set motor current. The advantage of that is you don’t need a huge battery power, and you don’t need to load the motor down (and dissipate the heat that generates).
2 Likes