A123 packs are recommended because of their low resistance (not so much their ability to put out power, but rather their ability to absorb it). They allow power to be "put back" into them very quickly(ie:they can be charged in about 10 minutes or so with the A123 charger). This ability to absorb the energy back is what makes them "Castle Approved". The energy doesn't "get caught" in the ESC because the path to the batteries offers low resistance.

Hmmmm...... I know I saw on another brushless esc manufactures website that a123 cells have a comparatively high internal resistance. They did not want a123 cells being used with their esc's. I'll try and find a link.

Hmmmm, I'm interested in who this manufacturer is.

Found it. This pdf applies to this thread. Page three is where they talk about the a123 cells. http://mgm-compro.com/pdf/en-motor-braking-050409-g.pdf

Ah, now I remember.

"In both cases, it is therefore enough to intensively brake only once and the controller can be easily and quickly damaged,
basically after few meters‘ ride.

The paradox is that the better controller you use (the more quality FETs and PCBs, or the more FETs and thicker copper layers it
has on pcbs, that is smaller resistances of FETs and connections it has), the worse the situation. The motor, working as a generator
during braking, causes more energy (current) to flow to the battery. Thus higher voltages are caught on the resistances of the
conductors (Rconductor) and inner resistance of the battery (Rbattery) and therefore higher voltage will also be on the input of the
controller (UCONTROLLER). That means, the destruction of the controller is more likely to occur when there is not enough reserve in its
voltage dimensioning."

Patrick,

I would like your input on what Mr. Grisa mentioned in the PDF file provided by lydiasdad. You have mentioned that you use a lot of copper on your ESCs which I think is good, but Mr. Grisa thinks otherwise form what I gathered.

Lutach,

Mr. Grisa didn't say it was bad to use alot of copper in an ESC. He just said that it made it worse when using higher resistive batteries when running at the voltage limit of the ESC.

During braking with a high quality ESC(low resistant), more current flows back to the battery. But with a high resistant battery and low quality connectors, the back current can't be absorbed well/fast enough. Thus you get a more voltage accross the input of the ESC( taller voltage spike at brakes), than as you would with the use of a higher quality battery pack. With the lower quality batt pack, the ESC will freak out and say "forget this shiat" and let the magic smoke loose.

Dunno if I explained it well enough.

All this talk and no one has stated the obvious:

1) We need ideal batteries; a pure voltage source with no internal resistance and infinite current capability

2) We need ESCs with 100% efficiency. FETs need to have have infinite slew rates, 0 rdson, infinite gate resistance, infinite frequency response. All the copper has 0 resistance.

2) Motors are 100% efficient; no copper losses and 100% of all magnetic flux is utilized. And, the rotor would be made of superconductors since they repel magentic field, so no need for magnets.

:smile:

Cool, I will pay for all of this technology for everyone on RCM with my money tree in the backyard. Right now the 20 dollar bills are in bloom but in late fall the 100's sprout! It's a nice early winter harvest.

So.... where exactly in Houston do you live? :whistle:

Not too far from me it seems. The last storm we had blew some bills over into my back yard.

Ok, so I got this thing inside my head that's driving me nuts. I have a M.Troniks ESC (Sorry Patrick, I know this is Castle's section) the truck one and it came with 3 diodes to add to each phase and then it went to the negative battery connector. Now what are they really for?

http://i206.photobucket.com/albums/b...p/DSCF2201.jpg (Thanks linc for the pic)

This is what I found in the Mtroniks manual. I believe the newer ESC's have them built in.

Mtroniks offer a simply installed power booster diode pack, which you can install to your controller and this will

solve overheating problems in almost all cases.

From the way they are hooked up (cathode to ground), they look like they would clamp the EMF spikes to whatever they are rated for. I would assume they are similar to the TVS devices in operation, except on the motor wires instead of the power input. Another possibility is to add/supplement the built-in diodes in the FETs (older FETs may not have the diode integrated, I dunno).

Makes me wonder as well. Novak, Quark manuals says never to use schottky diodes with their brushless controllers. But Mtroniks always had them. Maybe old school Brushless ESC users can chime in.