[phpBB Debug] PHP Warning: in file [ROOT]/phpbb/session.php on line 594: sizeof(): Parameter must be an array or an object that implements Countable
[phpBB Debug] PHP Warning: in file [ROOT]/phpbb/session.php on line 650: sizeof(): Parameter must be an array or an object that implements Countable
[phpBB Debug] PHP Warning: in file [ROOT]/phpbb/session.php on line 1110: Cannot modify header information - headers already sent by (output started at [ROOT]/includes/functions.php:3903)
[phpBB Debug] PHP Warning: in file [ROOT]/phpbb/session.php on line 1110: Cannot modify header information - headers already sent by (output started at [ROOT]/includes/functions.php:3903)
[phpBB Debug] PHP Warning: in file [ROOT]/phpbb/session.php on line 1110: Cannot modify header information - headers already sent by (output started at [ROOT]/includes/functions.php:3903)
[phpBB Debug] PHP Warning: in file [ROOT]/includes/functions.php on line 5277: Cannot modify header information - headers already sent by (output started at [ROOT]/includes/functions.php:3903)
[phpBB Debug] PHP Warning: in file [ROOT]/includes/functions.php on line 5277: Cannot modify header information - headers already sent by (output started at [ROOT]/includes/functions.php:3903)
[phpBB Debug] PHP Warning: in file [ROOT]/includes/functions.php on line 5277: Cannot modify header information - headers already sent by (output started at [ROOT]/includes/functions.php:3903)
vedder.se forums • A Rule of Caps ?
Page 1 of 1

A Rule of Caps ?

Posted: 12 Jun 2016, 10:03
by notger
Hi,

besides on a future Mountainboard Project, id like to use the VESC on a Low Power Pedelec, with max 350Watt 10S (~3,5 A)
Is there any "Rule of Thumb", how big Caps (in Farad) we should use for different Amps/watts.
Or is there even a min. Amp where no cap at all is needed?

thanks in advance

Notger

Re: A Rule of Caps ?

Posted: 12 Jun 2016, 11:04
by rew
The caps are required to smooth out the current peaks from the battery to the beginning of the VESC.

If we assume for a moment that the motor is consuming 10A (its maximum) at 18V or 50% of the max speed.
The average battery current will be 5A.

To achieve that the motor controller will switch the motor 50/50 between recirculating that 10A internally and using 10A from the battery. In theory the switch between those two is instantaneous. So... All of a sudden, the battery needs to go from delivering 0A (when the current was recirculating) to delivering 10A. This change CANNOT happen instantaneous. At the very least there is the lightspeed delay between the VESC and the battery. So, while the battery and the leads towards the VESC are gearing up to provide that 10A, the current needs to come from SOMEWHERE. That's why you mount caps. They can provide that large current on short notice before the battery can react.

The main thing that influences the required capacitors is the length of the cables. In practice, it seems that everybody is getting away with about 3 to 6 times less capacitance than I can calculate based on the model of what's going on.

So one of the things that influence this is that the FETs do not switch instantaneously. This gives the battery time to start "gearing up" putting less load on the capacitor.

Oh, by the way, it is when going the OTHER way, what damages your VESC if you omit the caps: The battery is delivering 10A, and suddenly inside the VESC the current is switched into "recirculate within the VESC/motor". Then the battery and its wires are delivering current that has nowhere to go. The voltage spikes that this causes will damage the VESC.

There are SOME caps on the VESC. If your leads are short enough, the induced voltage spikes may be small enough (at 10S, 36V) that it doesn't damage the VESC (which can handle 60V in theory).

Re: A Rule of Caps ?

Posted: 26 Jun 2016, 07:15
by Hummie
With the new vesc it looks to have one less capacitor. Maybe it's better with them so close to the board but how long can the battery wires safely be for that new vesc? If the caps aren't getting warm can I assume all is well?

Re: A Rule of Caps ?

Posted: 26 Jun 2016, 08:46
by rew
Caps getting warm is a bad sign. If you have a too small capacitance, but very good capacitors, the capacitors won't get warm, but the VESC might damage itself at high battery voltage due to the spikes carrying the voltage above maximum. You get more margin by not going close to the "60V Max" of the board.

Re: A Rule of Caps ?

Posted: 26 Jun 2016, 08:56
by notger
Hi,

Ok, i get the general use od the caps.

But is there any Rule of thumb about the Dimension of the Caps ?
For example
That many Farad for that Battery Voltage and that length of battery cable?
or something like this.

Cause i will use the VESC in a very small 10s setup with maximum 350 Watts (~3,5A).
And if some of the Caps are useless with that low Power id rather save the space then.

greets

Gernot

Re: A Rule of Caps ?

Posted: 26 Jun 2016, 11:31
by rew
You need caps that can handle the voltage of your battery. In your case the battery voltage is max 42. So 50V caps would suffice.

Design everything such that the leads from the battery to the VESC are as short as possible. The result might not even need caps at all. The big margin on the voltage (everything on the vesc should be able to withstand 60V) will probably allow you to run with small to no caps... If you change caps, keep in mind that they need to be "low ESR".

Re: A Rule of Caps ?

Posted: 26 Jun 2016, 11:37
by notger
Ok,

So i guess there is no rule to it, Cause sure its clear that the voltage of the caps has to fit to the battery, but the amount of Farad rule the size of the Caps.

Might be trial an error/damage like with some other Vesc-Topics

Notger

Re: A Rule of Caps ?

Posted: 28 Jun 2016, 06:18
by rew
If you have an oldfashioned powersupply the math is easy.

You want 5V, 1A from an 7805. You have a 9V transformer. Transformer peak voltage is sqrt(2) * 9 = 12.7V. The rectifier drops 2*0.7V -> peak after rectifier is 11.3V. the 7805 needs 3V in worst-case conditions before dropout: 5V + 3V = 8V. The capacitors must be able to ride out the time between the peaks. I = C dU/dt -> C = I . dt/dU = 1A * 10ms / 3.3V = 3000uF.

For the VESC: Would you mind if the voltage at the capacitor went up-and-down by 1V? I'd say: no. So we get: C = I . dt/dU = 50A . 40us / 1V = 2000uF.

But this is assuming that the battery provides ZERO of the ripple current. And this way you're putting 50A of ripple current into the capacitor. You'll be at around 70000uF of capacitors before you've gathered enough capacitors to handle the 50A ripple current. So the battery is definitively contributing to the ripple, but how much is hard to quantify.

So... for a worst-case calculation, the ripple current capacity of the capacitors is the limiting factor. And you get enormous quantities of capacitors. Less obviously works. But how much less is difficult to calculate.