Capacitor For MM 4S

[GriffinRU]

Originally Posted by Bernie Wolfard

Hmm, there have been a couple of things discussed in this thread I feel I need to comment on. I will also try to provide some information about the use of capacitors, the original reason for this thread.

First, the motor’s Kv affect on current.
Kv means RPM per volt. It has nothing to do with kilovolt or other electrical measurement. This is a constant all motors have. Part of Kv is determined by the number of turns of copper on the motor but turns affect different motor types differently so are not a constant and hence not really useful in determining a motors performance. When a motor spins up it becomes either a generator (brushed) or alternator (three phase brushless). As the motor goes faster, the amount of voltage it generates goes up. When the voltage of the motor matches the voltage of the battery pack it cannot go any faster, hence it has reached its Kv.
Given two similar motors with different Kvs and the same voltage, the higher Kv motor will pull more current (AMPS) than one with lower Kv. This is simply because the higher Kv is trying to get a higher RPM and it takes more current to go faster. There are other things that will cause the higher Kv motor pull more current such a lower copper to magnet ratio, less backplane, perhaps wider airgap etc. but you get the point. To get the efficiency advantages of getting watts from volts instead of AMPS you need to lower the Kv of the motor as your volts go up or you wind up pulling a lot more current as your voltage goes up.

Second: Capacitors, why?
Sole purpose of adding Caps to a controller is to control ripple current. Ripple current is a side affect of pulling pulsed current from a DC source where nothing in the system is electrically perfect. In this case this means the components of the system, particularly the battery, have electrical resistance*. A speed controller works by switching full throttle current on and off really fast to the motor (in our case about 11,000 times per second). This is called pulse width modulation or PWM. The motor averages these pulses out. If 50% of each pulse if off, 50% on the motor sees 50% power. This means the ESC is also pulling pulsed current at the same rate from the battery. In a ideal world these pulses would form a square sine wave where the on part of the pulse went straight up to full throttle current, straight over to the cut off point, straight downs to off current, straight over to on then repeat over and over, then rinse. Unfortunately, as some of you may have found out by now, we don’t live in a ideal world, electrical or otherwise. Because of the batteries internal resistance each pulse it puts out starts a little late, slopes up, overshoots, flattens out, shuts off a little late and tapers down and undershoots. All of this lateness, sloping and overshooting is called ripple current. The caps simply help smooth this out. If the batteries are inadequate, there is high resistance wire or plugs between the pack and the ESC or a high resistance plug the caps will quickly become drained trying to mitigate ripple current. If this happens the caps overheat. Worst case they blow up. As they are overheating they allow more ripple current to reach the FETS which in turn heat up. All of this is bad and can lead to catastrophic failure of the ESC. However, it is important to remember that capacitors cannot make up of inadequate batteries or poor wiring, they provide no increase in power and if they are the wrong ones make matters worst by increasing resistance. A properly designed ESC with good batteries does not need more capacitance. With inadequate batteries no amount of capacitance will help. To ensure long life in an electric power system just make sure you use the best (lowest internal resistance) batteries available with large enough capacity. Always ensure are least 20% more battery discharge capability than you think you will need and everything is OK an power will actually go up as more amperage equals more power, more voltage equals more motor RPM.

Bernie

* A batteries internal resistance is the one and only number that determines how fast the cell can discharge. The lower the resistance, the quicker the cell can discharge. Wouldn’t it be nice if battery companies published this statistic so we could easily judge which batteries are the best?

-----------------------------------------------

For the more technically minded here is a responce from our chief engineer on why ESCS have caps.

You might notice that every brushless controller on the market has an input capacitor that goes across the battery leads - the plus of the capacitor goes to positive battery and the negative of the capacitor across the negative of the battery. The reason for this capacitor is to smooth the ripple current from the battery, so that the battery sees a smoother current demand.

At partial throttle, the controller is turning the motor on and off at a high rate (for our controllers it is typically 13khz.) During the "on"
cycle, there is a fairly high current demand on the batteries. During the "off" cycle, the motor current is recirculated through the controller, and the battery has a near zero current demand. The capacitor recharges during the controller "off" cycle, and discharges during the controller "on" cycle.

The apparent ESR of the battery is reduced, and some of the strain on the battery is transferred to the capacitor.

In most cases this works well -- the battery efficiency is increased because it sees a lower peak current demand. Also the strain on other components (MOSFETs, back-emf detection circuits etc.) is reduced because of a reduction in the ripple voltage on the battery rail.
Without the capacitor, some setups would see ripple voltages exceeding 50% of the battery voltage. So you have to think of a battery as you would any other device - - there is parasitic resistance (ESR) and inductance (ESL.) Batteries are pretty good DC sources when the load is constant, but when the load makes large step changes, their output voltage also makes large step changes.

If, for example, I tried to run a motor which would draw 10 amps on a "perfect" DC supply, on a battery that had .1 ohm of resistance and an output voltage of 10V at no load, we would see a ripple voltage of just under 1V, with a loaded voltage of just above 9V, and an open voltage of 10V. Placing a capacitor across the battery would average the current demand on the battery, and smooth the output to a constant 9.5V.

Ok, at this point I can assume you understand the role of the capacitor. The reason electrolytic capacitors are used is because they are inrush current surge resistant (tantalums are not -- so they are dangerous to use in this type of application) they have a high Q so they tend to suppress ripple well at lower frequencies, and they have good bulk capacitance at moderate to high voltages (where ceramics do not.) As you pointed out, lower ESR would be desirable, especially in marginal systems where the batteries have very large voltage ripple (high ESR.) Tantalums cannot be used because the inrush currents would destroy them (they burn up) so electrolytics are used instead. With electrolytics, ESR and bulk go hand in hand -- so we selected a capacitor with a low enough ESR for the majority of applications, that won't be too large or too expensive. However, in some marginal systems, the capacitor is forced to do a lot of work, and the ESR of the capacitor creates heat. With an electrolytic, there is a failure mode where the temperature of the capacitor exceeds the boiling point of the electrolyte, causing a catastrophic failure of the capacitor. Usually, this will only happen on systems where the current demands of the motor are significantly higher than the capability of the battery to supply current.

I hope this helps clear it up!

Very well done!

But, your chief engineer simplified capacitor role way technical inclined hobbyists :) Hopefully he knows that and tantalums caps can do the job but cost allot and size/power /temp/voltage wise are not good for high current/voltage ESC's.

But good cover for bad batteries and poor connectors. And most important upon discharge batteries internal resistance goes up as well, so cap helps here as well even when you have top notch batteries and connectors.

[Arct1k]

When we say poor connections / resistance - is it measurable at no load with a multimeter?

PS lets buy castle pizza

[Bernie Wolfard]

Arct1k,

You cannot measure the amount of resistance with a standard multimeter it is too low. Even though it is low, it is significant because resistance goes up with the square of amperage.

To measure very low resistance values you put a set amount of amperage, say 10 amps, accost the part being measured with a lab quality power supply then measure the voltage accost the part with a voltmeter accurate to at least ten thousands of a volt. Then use Ohm law to calculate the resistance, ohms = volts / current. I don't know how to measure internal resistance or a battery but obviously this method will not work with them.

Artur, I think Patrick, Castle Creations owner and chief engineer explained why tantalums are not appropriate for these applications, one of these being they can explode from the amount of inrush current we get in these applications. Patrick is a systems engineer who has a long resume including being lead engineer at Garmin, Allied Signal and others.

Bernie

[BrianG]

Low resistances can also be measured with a conductance meter (reciprocal of resistance). However, measuring low resistances like that is hard because the meter leads and connections can affect the measurement appreciably. For this reason, measuring voltage drop is easier. Why would you need a lab quality PS though? As long as you have an accurate and calibrated voltmeter (with the proper resolution), you could measure the voltage output of any PS and the V drop on a small resistance (shunt maybe).

Also, IIRC, the internal resistance of batteries is not linear. Instead of supplying the resistance value of cells, manufacturers should supply the resistance curve.

[GriffinRU]

Originally Posted by Bernie Wolfard

Arct1k,

You cannot measure the amount of resistance with a standard multimeter it is too low. Even though it is low, it is significant because resistance goes up with the square of amperage.

To measure very low resistance values you put a set amount of amperage, say 10 amps, accost the part being measured with a lab quality power supply then measure the voltage accost the part with a voltmeter accurate to at least ten thousands of a volt. Then use Ohm law to calculate the resistance, ohms = volts / current. I don't know how to measure internal resistance or a battery but obviously this method will not work with them.

Artur, I think Patrick, Castle Creations owner and chief engineer explained why tantalums are not appropriate for these applications, one of these being they can explode from the amount of inrush current we get in these applications. Patrick is a systems engineer who has a long resume including being lead engineer at Garmin, Allied Signal and others.

Bernie

:)

You can do the same measurements with batteries as with caps.

BrignG - You need to have something as precision reference.

P.S. Resistance also would be temp and current dependant not to mention vibration...

[Arct1k]

I guess thats a no for me and my $20 multimeter then :)

[BrianG]

lol, probably not.

@Artur; As long as you measure the output voltage of the PS, know the value of the precision shunt/resistor, and measure the v drop of the shunt, why would you need a good PS? Wouldn't the various measurements account for any PS deficiencies?

[GriffinRU]

Originally Posted by BrianG

lol, probably not.

@Artur; As long as you measure the output voltage of the PS, know the value of the precision shunt/resistor, and measure the v drop of the shunt, why would you need a good PS? Wouldn't the various measurements account for any PS deficiencies?

You will be better with precision PS then shunt. You can improve your measurements with Kelvin type resistors but it is much easier with PS.

[pipeous]

I just wanted to update. I finally ran my RC8 today with the Neu 1512 2.5D. The caps worked awesome. I checked it a few times and ran for quite a while at a BMX track. Not even close to getting warm on 4s lipo. motor or esc. I'll never waste time with fans again. In fact, I am going to add caps to my Sidewinder now to cool the B44 down a bit.

[Cain]

Saw pipeous other post, he is running 4 of the 440 caps. So with my simple logic, would that mean its basically similar to running 1760 cap? If so, Then running the 1 novak 2700 cap could lead to even better cooling seen on a mamba max ESC?

Thoughts on this? Thinking of picking up 1 novak cap, maybe 2, and ditching the fan.

[BrianG]

Yes, if running caps in parallel, you just add the values up. So, 4 x 440uF caps (what an odd value) is the same as 1 x 1760uF cap.

However, there is a point where you get smaller gains. Sure, you can add 20 caps and it would be cool, but would be a waste. Approximately doubling whatever is on the ESC would make the most difference and still be easy to fit on/near the ESC. Adding more certainly won't hurt (although it does make the initial hook-up spark bigger), but you just don't get the same amount of gain for the effort and cost. If your setup NEEDs the extra caps just to function and not overheat, maybe it's time for a different ESC choice.

Also, since these caps heat up, it's better to use several smaller ones in parallel because there is more surface area to help the air cool them. Plus, transient response is typically faster.

[Cain]

Thanks for the response.

The main thing I am looking for is the ability to eliminate the fan currently used with the Mamba max on 4S. I am curious based on pipeous experience, using 1 Novak HV cap will basically give the same or even more benefit that he is seeing.

I think I may give it a shot.

[pipeous]

I do notice the caps get warm, about the same as everything else. The MM works good on the 8th scale with 4s. it's smooth. I can crawl along with no cogging, burn laps and it is just such a consistent feel. I could come arund a corner onto a table top and keep the throttle on enough to keep speed in the corner and roll the table for the next double. carving such tight lines to the big o too. I can't see spending 4 times the money when $5 worth of caps fixes the heat issue. I am getting about 30 mins runtime too.

I let quite a few guys drive it at the track yesterday. pretty soon all the questions of where, how much started coming up. We just had a new electric only track open up. the property owner has a brushless 8th scale, b44, 10th truck... all brushless. Sat night racing under lights will be awesome. big track

[lutach]

My Mtroniks truck controller has a huge 10,000uF, 85C capacitor and I'm just wondering, what's tha max amount of capacitance do you guys recommend?

Can someone post up a picture of were to place the Novak Cap, the pic postest earlier in the thread is dead.

Thanks!