churchrvb

My Feedback: (9)

My friend says that a capacitor on a esc, preferbly a Novak Brushless system doesnt do anything, i do remember reading about what it does but i forget. Can some one remind me what it does, so i can settle this straight, thanks

Access

If you have a weak battery, a large capacitor will provide some additional burst current.
1F capacitor provides 1 amp * volt * second. So for instance if you had a 50A burst discharge while flooring it coming out of a turn, a 12.5F cap would help hold the voltage up at a rate of 1 volt drop every 1/4th of a second. The problem is most of the 'racing capacitors' sold by RC shops are gimmicks they range in 10,000uF (microfarads) max, which is an insignificant value and won't give you any help. Real capacitors in the 50F range are expensive and lower voltage, you will need to put them in series to get up to the voltage you want and by then you end up adding some weight and cost. The only caps that are high value and lightweight, small enough I know of for something like this are aerogels
http://www.digikey.com/scripts/DkSea...S&Cat=31982111
and as you can see an Aerogel 50F, 2.5V (you'll need 4 in series which would equal a 10V, 12.5F capacitance value) would net over $100.00, you are better off just getting a better battery like a LiPo 4800mAh.

R/C fan-addict

um well a capacitor stores energy sorta like a battery. When more electricity is demanded from the motor it will add to the energy delievered from the esc to give some extra punch.

churchrvb

My Feedback: (9)

wow, thanks i appreciate it

Access

The amount of total energy stored by a capacitor is 1/2(capacitance in farads)(voltage)(voltage).
(voltage) is the voltage of the system, not the maximum voltage rating of the capacitor itself.

So for instance a 'racing' capacitor 4000uF = .004F. (1/2)(.004)(7.2)(7.2) = 0.1J An insignificant amount of energy.

For my prior example in the post above, if a quad-series setup of 50F aerogel supercaps were used, it would net a 12.5F value.
(1/2)(12.5)(7.2)(7.2) = 324J. 3240 times the amount of the insignificant 'racing' capacitor.

Then also consider that this would be the energy released if and only if the capacitor were discharged from 7.2V to 0V, which in reality does not happen, the receiver will stop working if that voltage drops below 5V or so.

This is why those 'racing' capacitors as sold are a gimmick, they are selling you for $10. or more something that costs $1. or less online and gives you no real performance increase.

gunnut

is it true that the esc/motor can be drawing so much current that the radio system is starved? (causing glitching?)

is a capacitor supposed to prevent that?

SN VipeR

Yes it's true. It happens (for example) to Spektrum if you run 4 cells (as 1/12ths do) or bad battery or nearly empty battery packs. But there the capacitor plugs right into a free plug on the receiver. Check Spektrum RC's webpage for details.

Doahh

Ok a cap on the ESC smooth out the power flow so you don't have cogging...
A cap on the reciever makes it stay at a consistant voltage so the reciever doesn't shut down when using a BEC

It doesn't make your car faster r anything all it does is smooth out the power flow

kostaktinos_mt

it depends. its like doahh said, BUT a high-capacity capacitor wired/mounted at the battery leads WILL give better performance if used with high-ir [low or medium quality] packs. that is what access said and it is correct.

Doahh

yeah but he's talking about ESC mounted caps....

kostaktinos_mt

esc to motor leads: less cogging, smoother control, more efficiency, slightly less brushes/comm spikes [slightly less wear on the comm].

esc to receiver bec: less glitching.

battery to esc: more burst, more acceleration and punch [needs very high capacity caps].

all wired in parralel.

we agree doahh..

Access

One other thing is that certain ESCs (if poorly designed) may underrate the in-circuit front end capacitor to cut size and cost. So adding a low-ESR (equivelent series resistance) capacitor on the input close to the ESC to supplement the one that is already on the ESC circuit board can help. But it is never going to help as much as if it were properly mounted on the circuit board itself. The same can be said for capacitors on the back end of a poorly designed ESC. The real solution is not to add a capacitor, it is just to buy an ESC that doesn't skimp on parts to begin with. The parts that go into building a $100. retail ESC are only about $10. anyways, whether the manufacturer spends $6. or $10. on the parts themselves can make a huge difference. Most ESCs are nothing more than a microcontroller and a bunch of PWM FETs with caps on the input and output, so knowing the PWM frequency of the ESC (not the receiver) and the maximum current draw you can calculate if the caps on the front end are sufficient.

Myself I have often had problems with stock ESCs, but never an aftermarket one from a good brand.

michelob78

What are the effects of the incoming voltage ripple? I realize that the battery pack puts out true dc power, on a osc. you would see a flat line. But with the demands put on the battery it would not be a nice clean power but more of a spikey voltage, would the capacitor not smooth out that power creating a more usefull amount of power?

Access

The ripple is caused by the PWM (Pulse Width Modulation) scheme that most ESCs today use. The speed control itself runs at a certain frequency, typically 25KHz to 300KHz. This is a frequency internal to the ESC, it has nothing to do with the receiver. By varying the duty cycle of an off/on wave at that frequency into the gate of a FET, or in most cases, a group of FETs, it causes a change in the output level. For this circuit to work at all, a capacitor is required on the output and one is highly recommended on the input. Good design practice states that the input and output capacitor be placed as close to the FETs as possible, ie. right on the board with traces kept short (5mm or less, preferably) and wide, and that low ESR capacitors are used. A good designer can calculate the necessary value of these caps using the maximum current, frequency, acceptable ripple, and ESR of a target capacitor.

If you are having to supplement a bad design by placing caps on a wire off the board, you are adding to the capacitors own ESR with any additional wiring and such. Far from an optimal solution, it may work to a point but eventually fail. If you cannot afford a well-designed speed control, you would do better to put a capacitor close to the receiver itself and a shottky diode to prevent discharge of that capacitor by undervoltage. The diode ensures that when the system goes undervoltage, the capacitor alone will power the receiver (rather than being 'dragged down' by the system voltage). You could get by with a much smaller capacitor and there wouldn't be any design issues or failure over time.

thecyberdude

Guys, the capacitor on the ESC serves to attenuate the high frequency component of the swithing stage within the ESC. Due to inductance in wires and stray inductances in the PC board large voltage spikes will be generated when the FET's in the ESC turns off. The cap attenuates those spikes and (when placed close to the ESC) prevents them from travel down the wire (acting as an antenna!) thus causing potential interference with the receiver. If the ESC was propperly designed in the first place those spikes would be eliminated by snubber circuits accros the FET's but unfortunately that is not always the case.

All NiCd or NiMh batteris acts as a huge low ESR capacitor - in other words...placing an additional capacitor in the uF rage won't do you any good in terms of "boosting" current or givin' extra punch! Anyway as Access mentioned earlier, to "fill" in for a "weak" battery for even fractions of a second will demand a huge capacitor (a lots of Farrads)....and rated at min. 10V (perferably higher say 16V or 25V) it will be the size of a beercan! Those supercaps mentioned is not meant for high current draw applications but commonly used for memory backup purposes on eg. a PC motherboard with a very little current draw - they will not do well as "current gap fillers" in RC applications.

kostaktinos_mt

there are some high-end electrolytic capacitors, such as those used for the flashes in high-end photo-cameras. their size is equivalent to an AA battery [they are somewhat thicker though], their voltage rating exceeds 50V and their capacity of some of them is close to 15F. the problem with most of them is that although they can be discharged as fast as desired [even short-circuited], the corresponding fast-charging process is killing them eventually. while the voltage stability enhancing process with the parallel capacitors and a battery pack works fine on the screen of an oscilloscope, it is rather expensive for a non-pro racer or basher to invest money into something that will work as desired for no more than 50-60minutes of continious total duty cycle.

Access

Not true for the aerogel supercaps, they are specially made for low ESR and "pulse power" or "DC hold up" type applications. A 50F cap is about half the size of my thumb and has an ESR of 0.025 ohms. You don't have to believe me, look on page 1031 of the catalog www.digikey.com.

Quoted from: http://en.wikipedia.org/wiki/Supercapacitor
Supercapacitors are also being made of carbon aerogel. Carbon aerogel is a unique material providing extremely high surface area of about 400-1000 m2/g. Capacitances of up to 104 F/g and 77 F/cm3 have been achieved. Some corporations, such as Cooper Electronic Technologies, are already producing aerogel-based supercapacitors. Their maximum voltage is 2.5V, but they can achieve an energy density of 325 kJ/kg (disputed as 10.6 kJ/kg, see Discussion), which is about 70% of that provided by the state-of-the-art lithium polymer batteries. Power densities achieved are even higher, up to 20 kW/kg, orders of magnitude higher than what Li-poly offers. Small aerogel supercapacitors are being used as backup batteries in microelectronics, but applications for electric vehicles are expected. [2]

As the article points out, the big problem is that you cannot actually easily use the stored energy in the capacitor past a certain point without wrapping a lot of other electronics around it.

kostaktinos_mt

for the history, aerogel capacitors had their first application at the 80s, used to power the guidance electronics of the AMRAAM 9[s -i think] missile by the collaboration of Boeing and McDonell with USAF

thecyberdude

Access: A comment on the Aerogel supercaps. In order to comply with the voltage range in RC applications you'll connect them in series - consequently the total capacity will be reduced and the resulting ESR raised - the later limiting the abillity to deliver large pulse currents. Due to internal varriations in the capacitors ESR you can't assume that the voltage splits equally over each cap so you'll have to build in some margin in order not to excced the voltage rating. This suggests using at least five 2.5V caps in a 7.2V system (max. battery voltge + 25% margin -> 9.6V*1.25=12V). Compared with the price tag of those caps i'll totally agree on spending the money on LiPo's instead

Anyway, the large electrolytics (in the uF range) frequently seen on the ESC is for filtering purposes rather than engergy storage.

Access

Yeah IMO as far as a RC design principle any kind of capacitor for energy storage, even an optimal one like an aerogel; it seems like a 'lost cause' since you are compensating for a less-than-effective battery and adding weight, cost. Just replace the battery with one that does that job right. I guess this whole conversation started b'cos people thought racing caps actually 'stored' battery energy and I wanted to illustrate the amount they stored was insignificant .1J at best.

michelob78

I run a cap on a LRP sphere esc, so it should have been designed well..... the manufacturer recommends a low esr cap on the batt side of the esc?.... Novak actually recommends not running several of their esc's without a power cap. I think they probably did design it without enough filtering, just to keep the case size and weight down.