The case against synthetic voltage

by | | 1 comment(s)

Background: ProModeler servos are rated at 7.4V (nominal voltage for a 2S LiPo pack under load), However, it's widely known that fresh off a charge, the output of a 2S LiPo may approach 8.4V. Thus, it's not illogical to wonder; is it OK to use 8.4V as derived synthetically via a voltage regulator (i.e. the BEC on an ESC), or the voltage out of a power expander/box? In fact, we're often asked variations of this same basic question - like this;

Q. Is it OK to set the output-voltage of my XXXX-device to 8.4V with ProModeler servos?

A. The short answer is yes . . . but we recommend using a battery instead of a BEC.

Long Answer: Many hobby shops sell BEC equipped ESCs (BECs are part of an Electronic Speed Control), power expander devices (often used with multiple servos per control surface on large model airplanes), as well as standalone BECs. In fact, most ESC manufacturers include the BEC feature for free because it's extremely cheap to implement (and this fact alone should tell you all you really need to know). As background, BEC means Battery Eliminator Circuit, and the heart of the idea is to use FETs (Field Effect Transistors) to derive a lower voltage from a higher voltage source.

A well known example involves outputting 4.8-8.4V (user selectable) from a 3S (11.1V) LiPo pack, then using the BEC to power the receiver/servo (instead of a separate receiver pack). This is where the battery eliminator part of the name comes in. Thing is, BECs (and built-in BEC on an ESC) are purely a cost savings measure for manufacturers. And the problem with cheap solutions is that while a $10 voltmeter may show 8.4V, if you use a tool like a $1000 oscilloscope, you'll see a different truth. Basically, that 8.4V isn't clean. Not clean like when it comes from a battery. In fact, it's downright dirty and fuzzy . . . like this.

- When viewed on an oscilloscope, the voltage supplied by a BEC is fuzzy - yuk!

And this also holds true for pricey power expanders, and standalone BECs (even the best ones) and as for why the output voltage is fuzzy, This is because of how FETs work. Basically they're fast switches. Remember when you were a kid and discovered you could make the lights in the hallway dimmer by flipping the on-off switch really fast with your hand? FETs work the same way! Anyway, without getting down into the weeds with technical stuff, just remember this; the BEC came into existence solely for manufacturers to save money (or add features cheaply). They have absolutely 'nothing' to do with high performance.

Bottom line? The output from a hobby-grade BEC is – at best – suitable for a $200 toy truck, perhaps for the driving lights, or to power a winch, or maybe for some cheap foamie model airplane . . . but it sure as Hell isn't something we'd recommend for powering good quality electronics! Of course, this isn't exactly breaking news amongst experienced modelers and electronic engineers (but few of our customers are EEs). We're of the opinion hobby dealers who continue to traffic in BECs these days are either ignorant, cavalier in their responsibility toward trusting users, or maybe just in it for the money.

But there's another issue. Performance. This is because when you nail the throttle, the propulsion battery voltage is drawn down. Basically, the BEC circuit is fighting to provide stable voltage to the servo - but it can't. And this happens 'exactly' when you need the most performance from your servos . . . like when a wheel is wedged in a crevasse, or during a waterfall, during the hover, or while performing a torque roll and you're moving all the flight controls at the same time (while also varying motor thrust with throttle). Bottom line? This is when your servos aren't seeing anything like 8.4V, and this means the servos aren't developing the torque or the speed you've paid for. Cutting to the chase; an inexpensive 2S LiPo battery isolates your pricey electronics from all this nonsense.

Anyway, the trick to selecting a LiPo is first, size it for how long you operate your model between charges (an 800-1000mAh 2S is plenty for an hour, or more), and second, to get one with a sufficiently high C-rating to withstand the power demands of a stalled servo (without browning out the receiver). For our part, we offer packs ranging from 250mAH through to a fantastic 2-lead 60C 2S2200mAH pack (probably good for 3 hours). Yes, a 60C pack may be rightly considered overkill but the delta between it and a cheapo 15C pack is minimal. Moreover, with respect to these high capacity packs, we have them built with two receiver plugs instead of one (the benefit of this is a) you double the available current to the servo bus (or halve it through each connector), and b) with two switches it's an inexpensive way to add some redundancy. Anyway, the 250mAH only weighs about 16g (<1/2oz) and the 2.2AH weighs about 123g (3-1/2 oz).

Last thing; since most guys get tricked into buying a cheapo battery charger when they first get into the sport, use that one to charge the avionics-pack while at the same time using the good charger for the propulsion-pack. Frankly, dealing with two packs isn't that big a deal and with clean voltage your expensive electronics won't be subject to glitches, spikes (as a motor unloads), draw down (depressed voltage because of the onrush of current due to opening the throttle), or the ripples and noise from the FETs.

This entry was posted in .

1 comment(s)
Steve Carruthers December 11, 2016 5:03 PM reply

When attempting to power sensitive electronics, i.e. receivers and servos directly from the propulsion battery we have had several major issues and they all revolve around the motor and controller. Several things are happening all at one time and can cause many different symptoms and makes it very difficult to diagnose the problem. The first thing that we noticed that cause a problem is the voltage spikes when loading and unloading the motor. We have seen voltage change as much as 3 to 8 volts with load and a great big spike when the load suddenly goes away. I do not have to tell you what happens to performance of most electronics with large voltage swings in short periods.

The next problem related to voltage is the continued drop in voltage and increase in current during the period of performance and the loss of power to the servos. The other problems we have encountered is the switching power supplies in BEC’s and the chopped power reaching the servos and receiver. We have had many unexplained gremlins in our systems when getting ripple in the power to our systems. Our solution is far too expensive for modelers and more complicated than most would be willing to proceed with. If there is sufficient space and allowance for an extra ounce, the best solution we have found is to use a separate power supply (battery) for the servos and receiver. That has increased our reliability many fold and reduced the overhead in both time and money.

- Project Manager, United States Naval Research Laboratory

You must be logged in to post comments.