Integra

iTrader: (8)

I've heard some Heli guys quote Much higher Amperage then 200+......as to it being true or not....that's up to some one to prove with a EagleTree Datalogger.

GSMnow

I am an engineer, I work in sound now, but with modern power amps and the other high power stuff I have done, I know quite a bit about ripple issues.

With a low internal resistance, low inductance motor, there is a huge current surge the instant an FET turns on. Even great batteries have a hard time holding up the voltage when this occurs. This is why every decent ESC has a capacitor of some sort to give the ESC the needed current for a split second as the batery recovers. This is all happening in milliseconds. the whine you hear when you pull the trigger a litle is the switching rate that the ESC is turning the FET's on and off. They are true digital devices, there should never be anything besides full on and full off. This is the only way such a small box and heatsink can carry this much power. How I have seen weak batteries make the power capacitor blew up, litterally, pop open and spew it fluid out as steam. If you notice the top of the capacitors bulging at all, you are in trouble as they are working way too hard to hold voltage as the battery voltage is just dipping too much too fast, and the cap is trying to control the voltage ripple. Once you lose the caps, the real damage can occur. If the rail voltage dips during an on cycle, the drive electronics can fail o hold the FET's in the full on state, causing the internal resistance to climb greatly. This will lead to a huge amount of heat n the FET's and make them burn up. Never try to race with damaged or missing power capacitors on an ESC, they are needed for it to work at all. And extra capacitance is a good thing. Just ask any high end car audio guy. They are adding hug caps because a car battery can't suply the quick surges the amplifiers can demand. and alternators are also very slow to respond and electrically noisy.

As a precaution, I may end up adding an extra cap between my battery and ESC. It certainly can't hurt. Those 3 tiny caps on the old Mamba Max are very special caps to suck up the ripple. But don't be fooled by physically large caps. Many bigger ones have to much internal resistance and inductance to be very helpful at the high switching rates of ESC's. The large audio power amps I work with now use a large number of smaller caps for this same reason, they just work better at high frequencies. Most large caps are designed to control ripple on 60 (or 50) HZ line power supplies, not 2,000 to 20,000 HZ that ESC's run at, let alone the 100,000-500,000 hz the switch mode regulators use in some high power systems.

Current and voltage ripple are very real and can destroy components. NEVER run without good caps, and if you run cheap pack, check the caps for visual signs of stress regularly. I have 2 Turnigy 4S 4500 30C packs on order, so I will be watching mine close.

symmetricon

iTrader: (27)

were and how does one buy and put the "caps" in

GSMnow

I have seen the Nvak power cap sold at hobby shops. You can just add it to your battery pack connector, but the best is as close to the batt input to the ESC as you can get it with the shortest wire possible. Be sure to get one rated for the pack voltage you are running, and use the max voltage, 4.2 volts per cell for LiPo. Higher voltage caps must be much larger and will be more expensive to be as good at high frequencies. Use a 20 volt rated cap for 4S, and 25 volts is bare minimum for 6S as a fresh charge is 25.2 volts, I would look into 30 volt caps if you plan to run 6S if you want it to last.

x60Driver

iTrader: (4)

You know all batterys are made in china for the most part. Smc has a lawsuit form them because they where not selling the battery they said they where.

Dave H

Don’t assume just because a lot of them are made in China they are all the same. China is a very large rapidly growing economy, companies are popping up all over. With widely varying quality of materials, processes, and business practices (ethics?). Why do I say this? I’ve been there and toured many manufacturing companies, have had some manufacturing contact going back over 30 years, have former colleagues that live there, current colleague who goes over every couple months for the last 15 years, customers that regularly go over, etc.

Some are made in Korea too. Best I can tell those are of a more consistent quality.

kufman

iTrader: (6)

Honestly, I have never seen and ESC killed from cheap batteries. Most ESC's either go up the second you plug in the battery or seconds after a hard crash.

rccardude04

Adding to what GSMNow said, I've heard the "ripple" effect explained as electronic momentum. As the FETs switch, or even after a HUGE amount of load and lifting off the throttle, you actually have a difference in voltage across each wire itself. Higher voltage near the battery and lower voltage near the ESC. As the FET switches to *off* or if you lift off the throttle rapidly, all the extra voltage/current will *slam* into the ESC's capacitors, and cause a pretty good size voltage surge.

I've actually been told by airplane and heli guys to add caps that are something like double of the battery's rated voltage. So a 50v cap for a 6s lipo would be in the safe range.

I'm not an electrical engineer (soon to graduate Mechanical, however) but it makes perfect sense to me.

GSM or anyone else who actually knows what they're talking about, please correct me on any of this if I'm wrong, but I think I understood it right.

-Eric

Davidka

iTrader: (86)

GSMnow's explanaion makes sense but the fact remains that modern, mediochre lipos are far more voltage stable than the majority of NimH cells that were in use before. It's up to the speed control maker to recognize the flutter problem and design a safety into their units for it. Nimh batteries are still sold to new hobbiests so that problem is not going away.

More interesting to me is the idea that these power caps wear out. Knowing that, I wonder if it would be a good idea for the speed control makers to recognize that and come up with a service recomendation for sending in the unit to have these caps replaced and preventing this damage from happening.

I stand by my opinion though, if a Castle claims a Zippy Lipo isn't "good enough" for their speed control then their speed control's safeguards are inadequate for far to great a percentage of the batteries on the market today. It's beginning to sound like one of these caps failed and took down your fets with it.

GSMnow

I have not seen much voltage over surge from batteries. I have used a digital scope a few times in the past hat sampled quite fast, but all my current testing has all been with an Eagle Tree which only samples at 10 per second, so I can't say for sure as there could be spikes uch shorter than 0.1 sec.

The big issues with voltage spiking in other devices is caused by indictance between the power source and the load. Car alternators are notorious as they not only have a lot of inductance to store up power as a magnetic field, they also have the regulator adjusting just the field current. So in full scale cars, if you are pulling a large load and cut it off quickly, you can see a spike of 100 volts out of the alternator before it settles down again. Newer cars full of computers are working to reduce that, but any well made electronics for automotive use must be able to absorb a 100 volt surge if you want it to live.

In electric RC cars, the only real inductance should be the motor windings. The ESC is a series of switches that put current into and take current out of these windings. The greater the inductance, the larger the lag in the change in current for a change in voltage. Anyone who has messed with an electro magnet will probably remember getting a healthy shock off of just a few small batteries. This is know as inductive kick back. When current flows into an inductive device, in our case the motor winding, it builds up a magnetic field. As the field is building, the current in the winding is actually reduced, and as the field reaches the steady state strength, the current climbs until it reaches the voltage/DC resistance of the wire using our old friend Ohm's Law. Using my Neu 1512 2.5D. It has a DCR or only 0.006 ohms. If the ESC was to just turn on eon winding and hold it there, the current would ramp up until it hit 14.8 volts / 0.006 ohms = 2466 amps!! Obviously this can't happen, something will fry, let alone there is more resistance in other places like the battery pack, the wire and connectors, and of course the ESC FET's. Add it all up, and my car is still down at about 0.027 ohms total or still 550 amps. When the motor is not turning, or turning slow and the back EMF from the motor is quite low, the ESC must pulse the windings on/off at a high rate. The idea being to build up only the desired amount of current, and then switch the FET off. When the FET does turn off, the magnetic field begins to collapse back around the winding. This acts just like a generator with a magnet moving past a coil. The field is moving backwards, so it generates a voltage in the opposite direction. In a perfect world, this voltage would climb until it created enough current to match the original watts that created the original field. There are some losses so it does not go quite that high, but with no load at all, thousands of volts can be produced. In a gasoline engine, this produces about 400 volts on the ignition coil primary which is then stepped up to 40,000 volts to the spark plugs. Obviosly, we don't want to be firing spark plug in our electric RC cars. The ESC uses another FET or FET's to take this inductive voltage kick back and direct it to maintain the current in the motor winding. In most this is done by literally just shorting the winding out. If the coil current was up to say 100 amps, the current into the short would be very close to the same 100 amps the instant the winding goes from voltage source to shorted. As the magnetic field decays down, the current will fall off at about the same rate it built up. During this coild shorted time, the battery is not supplying any current. This is how we can see more current in the motor than from the batery, but at a lower voltage. For example, at 50% commanded voltage, the battery is connected about half the time and the winding is shorted half the time. The current builds to 100 amps, but the average for the build up time is only 60 amps from the battery and is only on 50% of the time, so an amp meter would only show the ESC pulling 30 amps from the battery. The coil in the motor sees the same 60 amp average from the build up from 0 to 100 amps, but them the winding is shorted, the magnetic field collapses arounf the winding generating power and the current ramps back down from 100 amps back down to zero averaging about 60 amps still. The motor winding now sees an average current of 60 amps at half the battery voltage. During this part of the operation of a motor, a weak battery should only cause the input voltage to dip down during the FET on time slowing the current build up rate. This will stress the power capacitor and maybe kill it eventually, but I doubt even a very weak pack will cause a major failure. BRAKING is a whole different ball game.

Now the motor becomes our power source, and it has inductance like an alternator. When you mash the brakes, it takes the generated wattage (voltage x current) from the motor and has to consume it to slow the car down. If there was no friction or other losses, the same power it took to accelerate has to be sucked up to slow it back down the same amount. And I am sure most of you can see that a good brushless system can slow down faster than it can accelerate, which means we are consuming those watt hours of energy faster than we put them into pushing the car. If we tried to pump that power into just a resistive load, it would be like a hair drier. As the motor is turning, it is generating voltage, now an FET shorts the motor winding. The generated voltage now has to build a magnetic field in the winding and build up current. There is almost no resistance in this circuit, just the motor winding the ESC FET's and the wire between. So if my car is rolling at 1/2 speed, it is generating 7.4 volts and a dead short on the motor would come out to 0.006 ohms for the winding, 2 x 0.00013 ohms for the ESC, and figure .001 ohms for the 15 inches total of #12 wire. This is a total of 0.00726 ohms. 7.4 / .00726 = 1019 AMPS!!!! Ever wonder why braking heats up motors and ESC's? Obviously we don't want this much brake current, and the resulting 7543 watts it produces. We want say 740 watts of braking. SO we only want 100 amps, so when the current ramps up to 100 amps, the FET's turn off, but without a load, where is the inductive current going to go? The FET's channel it to the battery pack. This is regenetive braking. At the instant the switchover happens, the inductive kick voltage will jump up to the battery pack voltage, and reach the current needed to match the original wattage. In this case it goes to 14.8 volts at 50 amps. Yes, we are charging our 5000 mah battery at 10C, but only for a few miliseconds at a time, but if the pack is weak, you can see where this is leading. That high charge current could shove the battery to well over 14.8 volts, or even the 16.8 safe maximum. The ESC power capacitors come into play again to help absorb this power, but can only do so much. The last line of defense is a set of diodes that will clamp this voltage to just less than the FET's can handle. This may force them to take huge current spike and this is where failures usually begin. In the large industrial motor drives, they don't have a storage battery to make use of the braking power, but they do use a large capacitor bank to suck up as much as they can to use when the motor has to start again. The caps are allowed to charge to well over the incoming line voltage making the drive pull no power until the caps are pulled back down again. And if the cap bank voltage goes too high, then it has to be shunted. This is done by turning on a bank of large resistors to turn the extra generated power into heat, and lot's of it. A 3 HP motor trying to stop 400 pounds from spinning at 300 rpm will make the resistor bank glow. As far as I know, noone has bothered to use braking shunt resistors on an RC ESC. They depend on the battery and power cap to be able to suck up the power. And since all of the watt hours generated under braking had to come out of the battery to accelerate the car, even though the current may be quite high, the total mah shoved back into the battery will only be a fairly small fraction of what we took out, so there is no risk of overcharging, even if you accelerate and slam on full brakes at the very start. When you figure in the amount of energy it takes to roll the car, the braking amount we can capture back into the battery is small, but can certainly increase the run time a little bit. Going to mechanical brakes will turn all the rolling energy of the car into friction heat, but will allow the motor to run cooler as it is not asked to handle the watts through it a second time.

So in short, I am betting that most weak battery ESC failures happen under braking, and good battery side power capacitors are essential for good performance under accleration and braking.

unsp0ken

iTrader: (25)

Nice write up Gary! That was like my power class all over again (I wasn't very good at it)! I do want to pick your brain a bit since I switched to mech brakes yesterday.

I switched to mech brakes yesterday to see if I'd get more run time under the assumptions (or guesses) that:

1. The ESC's we use in the cars aren't really optimized to make efficient use of regenerative braking with most of the received energy being dissipated as heat.
2. The power applied under braking is possibly greater than the usable power from regen due to point #1.

Any comments on the above?

But yea, I really want one of those dataloggers to play around with this stuff. It's like engineering lab all over again but a lot more fun.