Switching BEC protection
 

Due to my recent experience with a UBEC failure taking out a servo and receiver, I started researching methods to help prevent collateral damage if/when it happens again. I came up with a tiny, simple, and cheap "sacrificial" circuit that is designed to blow before the servo(s) and receiver does.

http://scriptasylum.com/forumpics/bec_protector.gif

As you can probably tell, this would be installed between the BEC and the receiver. It is designed to be used with ESC integrated BECs, or external ones.

The circuit consists of a TVS zener diode rated for 7.5v, and a 3A "picofuse". I opted for zener-based protection because zeners have a tendency to fail "closed" (basically become a piece of wire). And the particular zener I am looking at has a response time measured in pico-seconds. The 7.5v rating is to provide a little voltage headroom, yet clamps dangerous voltages. I chose a "picofuse" (non-glass type) due to the environment. You may think 3A is a bit low, but this type of fuse can conduct more current than its rating in bursts without blowing.

Operation is simple. If the BEC is operating normally, the zener and fuse are basically transparent; very minimal voltage drop and full current available to the load. However, if the BEC decides to fail and send full battery voltage to its output, the zener clamps the voltage to 7.5v and shorts the excess voltage to ground. This will generate quite high current and blow the fuse. And even if the zener blows before the fuse does, it becomes a short (as mentioned earlier) so it helps the fuse blow even faster.

It is important to note that this will not protect the BEC. If it fails, it is already too late. The best we can do is try to protect the other devices downstream. So, yeah, if the BEC fails, you'll have to replace it (or the ESC if the BEC is integrated like the MMM), but certainly cheaper than replacing servos and receiver as well. And like I said, this is a sacrificial circuit, which means once it blows, time to get a new one.

Yeah, there are probably more elegant ways to implement protection, but I feel simple is usually better, and it doesn't get much simpler than this.

Ok, all that said, I want to gauge interest. I plan on making an order for the parts necessary to do some testing and make a few for myself. But I am curious if anyone else would invest something like $7-$10 for such a thing. And I will not be selling these myself, but may end up in a certain storefront. :whistle: The most expensive part of this, believe it or not, is the stupid 6" servo extension lead. No matter what, I don't plan on making much (if any) money off this, but I'd like to break even and that's not even including my time. This project is for the benefit of the hobby, not my wallet.

Maybe someday, manufacturers will build in some type of protection into their BECs, but for now, something likw this will have to do.

depending on the size of it, I'm in. i've never had a BEC fail that way for me(i've had them short on themselves and cause smoke, but always cut power to the Rx) but with the price of a spektrum Rx, and a good servo for 1:8 scales you could be down a good 100+ bucks

Also, what about those with linear BECs, like the MMM?

The circuit would be about the size of a postage stamp, maybe a bit smaller and thicker (obviously), plug about 6" for the servo leads. To keep costs down, it would probably be point-to-point wired (instead of on a perfboard or something).

My recent BEC failure was my first one too, but once was enough. Yeah, the cost of a Spektrum receiver and a powerful servo can actually hit $200.

i would like to invest in something like this. Had one blow years back and took out the servo....not the receiver strangely....

Brian, are you going to put this on a board with pwm leads coming off the ends for easy installation?

mabey this is a dumb question, buy why couldnt you use a bar resistor to bring the voltage from the lipo down to power the RX strait from the batt??:neutral: or am I compleatly stupid??:sleep:

A resistor would only limit current not voltage. Think of it this way if you have a hose and it has a restrictor placed in line with it very little water will come out the end limiting flow (amps), but if you plug the end the pressure (voltage) will rise to the same as before the restriction. So when the current draw is near 0 the voltage will rise to a dangerous level and take out the RX and servo.

Brian this would be easy to add to the existing BEC on the board. Good idea using the fuse, I figured that the trace would be good enough.:wink:

Jeff

Brian, I'm interested in the components if you do a large run. If you could send me enough components for 3 sets on a custom tape-and-reel (piece of tape) inside a letter envelope (if they fit) that'd be cool. I can do the assembly myself and have access to servo plugs.

Dont they make shorter servo extension leads, sure I've seen massive lots for sale that are very short and cheap...

I like the circuit though, it really is so simple even I can understand it for a change :yipi:

My spektrum receiver has a voltage regulator built-in so the circuit was voltage protected, so I imagine other receivers will have something similar. But when the servo blew from the high voltage, it drew lots of current through the receiver ports which fried the traces connecting the power pins. I had to solder a bridge wire across those pins and it was fine. So, it was over-voltage protected, but not over-current protected.

Like I said previously; to keep assembly cost low, I may just do point-to-point wiring. But a small piece of perfboard shouldn't be too bad. And it will come with servo leads already installed.


Lile j57ltr said, a resistor wouldn't work well. If it was in series, all it would do is drop more voltage as the load increased. Maybe good for when the BEC goes bad, unless no current was being drawn and voltages rose, but would also have varying voltage drop for normal loads.

Yeah, using the trace as the "fuse" may work, but then there is a risk of fire of whatever. A purpose-built fuse controls where the failure will occur.

Yup, it's simple all right. If it works, I'm amazed that manufacturers haven't included it as part of their design. And I'm sure I can get bulk servo leads for less cost, but would most likely need to order something like 500-1000 to see the benefits. And even at $1 per lead, it gets expensive. Especially if interest is low. Total build cost would be under $3 per unit, but I'd have 1000 units sitting at home not being used.

I'll probably go ahead and order a few component sets just to do testing. If it works like I hope, will assemble around 20 kits and go from there.

http://cgi.ebay.com/50x-300mm-20-Ser...3286.m63.l1177


ebay is your friend...

Are you planning to put these in a small project box or just shrinkwrap it all together?

Yeah, ebay was an option, but would rather have an actual supplier. But for the first small run, ebay will work.

Yes, the package will look like a servo extension with a heatshrunk bubble in the middle.

just curious on the consequences of the short to the batteries?

edit: i know a bad bec will take things out downstream...but any reports of things upstream going out?

Really, any short that bad will turn the BEC input wires into a fuse far before any damage at all can happen to the batteries.

I only have one external bec now, but plan on getting another one. If your idea works out, I´m in for two.