Cap banks - the pre-emptive answer

[ray schrauwen]

Here's the answer before you ask.


external cap banks ADD capacitance to your ESC.

When adding a cap bank, DO NOT remove existing esc caps ( replace dead ones with new ones, sure.. rated same )

Just delivered my desk... a new HV ESC where the owner had removed all onboard caps and replaced with an external cap bank.

I know this fellow had the best intentions, so there will be no finger pointing and gesticulating (look it up) for all to see.

[ATTACH=CONFIG]83341[/ATTACH]

Can I ask why is it bad to remove the old ones as pictured and run it as pictured? What happens? What will go wrong?

EDIT: That is if they have all gone bad, not if they were still good.

[turbonutt]

drwayne, where did you get the raw circuit board material?

[tlandauer]

Here in the States, Radio Shack sells them.

[drwayne]

Thats what I was asking or trying to find out.

I'm sorry my posts were poorly written. I do that sometimes when I haven't eaten enough.

I didn't want you to "say" anything. My apologies sir.

Of course if in-line I'd use heavy guage too but, I wanted mine off the side in flotation so it was out of the way. Since it's not carrying a high amp load as far as I know ??? I was hoping after seeing your posts that I was O.k. and didn't have to change it.

Thanks for your posts!

If I miss food.sleep/medication I spend next morning apologizing for my last night's posts.

Im sure you know this.. capacitors are 'miniature super batteries' that charge and discharge super quick. YES, there is a lot of current/energy
flowing through the wire conduits.. so larger wire on long satellite leads is a must.. ( resistance, EMF, delta Tmp )
Ive evidenced caps under load where the flow is sufficient to burn their own leads away...
Id recommend to 'overcompensate' when adding cap banks.. a few $ more can be the saving grace for your setup.

Can I ask why is it bad to remove the old ones as pictured and run it as pictured? What happens? What will go wrong?

EDIT: That is if they have all gone bad, not if they were still good.

as pictured ?
In the first #1 post, that ESC was NEW ( or a few runs old only )... !
If the onboard caps die, Id most certainly add replacements inside or out.. the circuit was clearly stressed with those existing, so ADD MORE !

Remember this, capacitors store energy.
They are waiting to BITE you 'bad'.

W

[properchopper]

Doc, Is it possible or "not good" to have too many or too big a cap addition ? (my understanding of basic electronics stops at OUCH )

DSC03486.JPG

[properchopper]

for giggles I searched "capacitor explosion" on youtube. Some crazy stuff there

http://www.youtube.com/watch?v=kjhoePUJ8Vg

[stractor]

Please excuse my interruption here, but aren't cap banks supposed to be made up of "X" of amount of caps arranged in parallel so the capacitance of the bank is equal to the sum of all the cap valves in microfarads? That is added c1+c2+c3 = total. I ask this question because in looking at several of the posted pictures some of the banks seem to be made up of sets of two caps in series and these sets are then wired in parallel. The caps set of two caps in series doesn't add but equals a value less than one of the caps value. In the example I am refering to putting two 1000uf caps in series will equal 500uf. See the following link http://www.calctool.org/CALC/eng/electronics/cap_series
If there is another reason for wiring them in a series and parallel configuration or I am missing something please tell me.

The following picture seem to illustrate what I am talking about. Of the small board pictures one is from underneath and seems to show three pairs of series caps in parallel . If you give a value of 1000uf to the six caps the total capacitance of the board is 3000uf or so and not 6000 as if they were all in parallel.
The larger board seems to show the same thing, but there is no picture of the underside showing the actual solder connections.
Of course I could be completely wrong in my thinking.

1.JPG2.jpgIMG_0202.jpg

[keithbradley]

Here are some details of the shared bank

[ATTACH=CONFIG]83358[/ATTACH]

[ATTACH=CONFIG]83359[/ATTACH]

for 12S2P feeding 2x 300A esc - 5698/910 - X665 ... child's play ! ( big kids )

Have you happened to log that combo yet Wayne? I'm curious what kind of current you're pulling with that setup...

[drwayne]

On road atm. 1030am here in South Australia.
More when at desk

I deleted this because it confused even me !

[ray schrauwen]

Very interesting DrWayne.

[drwayne]

Words Vs pictures
remember these numbers ... ...
C= 470uf cap, low ESR, rated voltage ~ 25% above your pack voltage
2 in parallel c1+c2=940uf
2 in series offers C1xC2/(C1+C2) = 235uF !!!!!!
series (-)---[C1]---[C2]---(+)

That series pair in 3x parallel array offers 705uf
(-)---[C1]---[C2]---(+)
(-)---[C1]---[C2]---(+)
(-)---[C1]---[C2]---(+)

Initial charge of the 6 caps is same as if all were individually charged

OMG Ive written and rewritten this a few times and each time I start expanding on referenced detail that loses most people.

simply..
In this circuit, the charge/discharge speed is 8x the std , thus smoothing the DC line ripple more effectively ... even though the energy return availability is 705/2820 ( std6 caps ) .. it presents 8x speed of charge/return into of those original individual 6 caps.. gives us twice the total energy available, at a far higher return rate than std. ( 705 x 8 / 2820 )
In the 15S300A5698/910 example I used a total of 1410uF as detailed here, and logged ripple <0.19Vdc.
Unfiltered, ripple measured >1.6Vdc .. but I didnt enjoy the lacklustre performance so I went home...
I use cap banks as detailed to cover both yield(battery recharges the caps) and reply(esc demand on power system) needs of the system.
------
I read above .. What is DC ripple ? . When the esc fires a phase, it draws energy from the 'energy sump' .( DC batteries ).. this causes a sudden voltage drop evidenced by a blip on the voltage display of your oscilloscope.
Each 3 firings ( a rotation of the motor ) makes 3 tiny blips ..... eg at 30,000rpm there are 90,000(at least) tiny blips.. a ripple .. an induced frequency.....EMF inciting ESR .....
This ripple.. creates an AC that is not usable by the DC-AC inverter (ESC) thus the esc now sees lower voltage...V x Kv - zoomzoom
Capacitor banks reduce the size of that ripple, the esc sees higher nett voltage = more fun.

If the cap bank is too large, it will induce it's own ripple as it recharges.


Would I recommend this type of cap array if asked .. depends on your application.
Im sure to edit this once Ive seen it in a fullscreen.. and read it as a mini-novel.
W

[nata2run]

I guess what Doc is trying to say.... as he has explained it to me a 1000 times is... Because the internal resistance of a battery is much higher than a capacitor. It takes a much longer time to get the same amount of energy from a battery, and the timing is very important. In fact, using a different combination of capacitors/layout can give us the exact control we need. I guess the idea here is time/responce over high numbers of uf. It makes perfect sense..... I wouldnt recommend everyone to just play with caps in series (doc knows what he's doing) as it can have a reverse effect.

(my selective hearing and 10c worth)

[drwayne]

I guess what Doc is trying to say.... as he has explained it to me a 1000 times is... Because the internal resistance of a battery is much higher than a capacitor. It takes a much longer time to get the same amount of energy from a battery, and the timing is very important. In fact, using a different combination of capacitors/layout can give us the exact control we need. I guess the idea here is time/responce over high numbers of uf. It makes perfect sense..... I wouldnt recommend everyone to just play with caps in series (doc knows what he's doing) as it can have a reverse effect.

(my selective hearing and 10c worth)

Wow. here I was thinking you had learmomg difficulties

[nata2run]

HA! yes, only 1000 times later LOL

[ray schrauwen]

Holy Cow! I didn't think it was that complicated!
YOu oviously make good sense from what I read.

I would guess the Etti cap bank is all parellel ? Since only 5 caps...?

Copy & paste in archives...

Thank you!

Words Vs pictures
remember these numbers ... ...
C= 470uf cap, low ESR, rated voltage ~ 25% above your pack voltage
2 in parallel c1+c2=940uf
2 in series offers C1xC2/(C1+C2) = 235uF !!!!!!
series (-)---[C1]---[C2]---(+)

That series pair in 3x parallel array offers 705uf
(-)---[C1]---[C2]---(+)
(-)---[C1]---[C2]---(+)
(-)---[C1]---[C2]---(+)

Initial charge of the 6 caps is same as if all were individually charged

OMG Ive written and rewritten this a few times and each time I start expanding on referenced detail that loses most people.

simply..
In this circuit, the charge/discharge speed is 8x the std , thus smoothing the DC line ripple more effectively ... even though the energy return availability is 705/2820 ( std6 caps ) .. it presents 8x speed of charge/return into of those original individual 6 caps.. gives us twice the total energy available, at a far higher return rate than std. ( 705 x 8 / 2820 )
In the 15S300A5698/910 example I used a total of 1410uF as detailed here, and logged ripple <0.19Vdc.
Unfiltered, ripple measured >1.6Vdc .. but I didnt enjoy the lacklustre performance so I went home...
I use cap banks as detailed to cover both yield(battery recharges the caps) and reply(esc demand on power system) needs of the system.
------
I read above .. What is DC ripple ? . When the esc fires a phase, it draws energy from the 'energy sump' .( DC batteries ).. this causes a sudden voltage drop evidenced by a blip on the voltage display of your oscilloscope.
Each 3 firings ( a rotation of the motor ) makes 3 tiny blips ..... eg at 30,000rpm there are 90,000(at least) tiny blips.. a ripple .. an induced frequency.....EMF inciting ESR .....
This ripple.. creates an AC that is not usable by the DC-AC inverter (ESC) thus the esc now sees lower voltage...V x Kv - zoomzoom
Capacitor banks reduce the size of that ripple, the esc sees higher nett voltage = more fun.

If the cap bank is too large, it will induce it's own ripple as it recharges.


Would I recommend this type of cap array if asked .. depends on your application.
Im sure to edit this once Ive seen it in a fullscreen.. and read it as a mini-novel.
W