Do I need an external BEC?

[rudi]

I've thought about this (for a short while), and if I was adding a separate battery, I would make it a dedicated receiver battery and not use the BEC at all. The reason is, I've had batteries fail due to over charging or poor charging routines, and using a BEC to charge while flying doesn't seem like a robust, reliable solution to me. I'd be so paranoid about whether the battery was good and working properly and its state of charge, I'd need to check it before each flight. If I'm doing that, I might as well just take it out and charge it but on a decent charger where I can properly cycle charge it and monitor the capacity I'm putting back in.

I don't even charge my transmitter NiMh using the transmitter's built in trickle charger (I've had a cell fail during an overnight charge doing this). I always remove and cycle charge it with a determined amount of capacity.

The two big concerns I would have, other than battery charge state, is what happens if the battery is a higher voltage than the BEC (BEC damage?), and could a BEC failure go unnoticed during preflight, resulting in a crash due to low battery.

[muscleflex]

i think you're all making this tiny 450 too complicated!

[rudi]

i think you're all making this tiny 450 too complicated!

In hind sight, I should have probably got a larger heli

[cjcj1949]

There are many opinions on electronic matters given on forums. Not many of them are given by people who understand what is going on. It is further complicated by the fact that manufacturers supply equipment without supplying proper specifications. Some of these modules will work together, but for a reliable system you need compatible modules and an overall system design. The mks approach is best. On a 450 a 3A BEC with a 5A overload will work reliably, especially if you run it at 5V. Running at higher voltages will not be as reliable. As servos vary so much there will be some that work reliably at 6 or even 7V, but even these will be more reliable at a lower voltage. If our servos had proper specs we would be able to read the MTBF (mean time between failures) at different voltages. The Kontronic simple buffer idea may be ok with their esc, but not with all. It would work very differently at 5V or 6V. At 5V the NiMh would stay flat or destroy the esc, but at 6V the NiMh may get overcharged. NiMhs should not be connected to voltage sources and voltage sources should not be connected to NiMhs. I don't think you will have problems on a 450 with a single 3A 5V BEC from a good supplier.

[Bronek]

In hind sight, I should have probably got a larger heli

yup, me too

[Bronek]

There are many opinions on electronic matters given on forums. Not many of them are given by people who understand what is going on. It is further complicated by the fact that manufacturers supply equipment without supplying proper specifications. Some of these modules will work together, but for a reliable system you need compatible modules and an overall system design. The mks approach is best. On a 450 a 3A BEC with a 5A overload will work reliably, especially if you run it at 5V.

OK thanks. Just like Rudi I am running Trex 450 DFC with mks DS95 servos ; my BEC/ESC is Koby 40 LV and I think it's strong enough to deliver required power at all times, it's 5.6V (no way to change it without progdisc) continuous max 3A , peaks max 10A.

I'm still tempted to attach ~800mAh NiMH 4.8V as extra precaution but since its extra 50g weight and relatively large (compared to amount of free space in this heli), it probably won't be anytime soon. Attached to 5.6V power source continuously it would probably do fine.

[rudi]

Just like Rudi I am running Trex 450 DFC with mks DS95 servos ; my BEC/ESC is Koby 40 LV and I think it's strong enough to deliver required power at all times, it's 5.6V (no way to change it without progdisc) continuous max 3A , peaks max 10A.

Oh cool. How do you find the DS95 servos compared to stock?

I plan to run mine at 6V when they arrive.

[Bronek]

Oh cool. How do you find the DS95 servos compared to stock?

with my lack of skill I can tell only one thing - they work

EDIT: actually one thing to look for, the splines are shallower than on Align servos. You need to force Align horns really hard, or buy MKS horns which do fit nice but then on 450 Pro you need to cut half of each horn (which is what I did). I was unable to find other brand horns which would fit well on MKS DS95.

[cjcj1949]

I think a SuperCap might be a better way to cope with surges. I'm not sure if you want to though as they may happen after a crash when servos can't reach their positions and the extra current may just kill the servos quicker. It's not a proper solution to add another power source in parallel with an esc. The normal advice is to remove the red wire from the esc to the rx when using a battery. Diodes would make things reasonable.

[rudi]

I think a SuperCap might be a better way to cope with surges.

In a normal heli receiver/servo system, under normal operation I don't think there should ever be any surges great enough to benefit from a SuperCap buffer. If there are, then the problem should be identified and fixed rather than patched with a SuperCap.

SuperCaps are a nice idea though. What sized SuperCap would be required to power only the receiver (not servos) for 10 minutes, and physically how big would it be?

[cjcj1949]

I agree. I think they would make a better backup than a Nimh. It depends on your bec though. If the Bec is marginal some extra energy would keep the volts above brownout. The Cap wouldn't have to supply the current for long and so the servos would be less likely to burn out when stalled.

For a capacitor the approximate volts drop is proportional to time, current and inversely to capacitance. If you use the right units V=it/C

V is volts
i is amps
t is seconds
C is Farads

For the receiver if it uses 50ma and we run it from 8V to 4V

C=it/v = 0.05 x 600/4 = about 10 F

A battery would be quite good though. Or you could use a schottky diode and a smaller cap the other side giving less protection. Perhaps one schottky diode per servo and one for the rx with a capacitor. The servos might need capacitors as well, depends on their design, as their volts would drop with current and that might make them unstable. The capacitors are called decoupling when used in high speed digital which is similar, although the volts drops are due to inductance of the power tracks and the fast current switching.

The real capacitor equation is more complicated, but capacitors have a tolerance anyway and it is close enough for a drop of about 50% of supply.