As usual, one of these white papers is the result of responding to someone asking a question. Questions, which I suspect many have but don't ask. This one is no exception as it deals with powering the servos in larger models. This time, a 40% class model aircraft . . . a BIG toy airplane!
FYI, my own 40%-er is a Slick spanning 126", powered by a 210cc engine, and which goes 45lbs. And as you'd expect, my experience colors my thoughts. Anyway, this guy packed two questions into one general purpose query.
Q. Is LiPo or LiIon preferred for receiver+servo battery (both being 3.7 nominal volt)? I'm especially wondering how this affects my choices in high power servos. My model is a CARF Extra 330LX and I am considering both ProModeler DS1155BLHV and Savox 2290SG servos. I'd appreciate your thoughts.
A. So the basic query regard LiPo versus LiIon as the source of power. But for large models, there's more to unpack because what he's really asking is how to manage the current demand for a large model equipped with power hungry servos. So this ultimately devolves to a current management question. Added to which, the response can be applied just as well to smaller models, e.g. 33% ranging in span around 104-108" and weighing in around 27-32lbs as they're often equipped with the very same servos.
So checking the specs of both you'll see the performance and current demands are similar.
- Savox SB-2290SG (972oz-in, 0.11sec/60°, 10.5A - 8.4V)
- ProModeler DS1155BLHV (1155oz-in, 0.10sec/60°,10.1A - 8.4V)
Note; both servos exceed the rating of the servo-connector because it's what receivers are standardized on *and* the RF manufacturers haven't come to an agreement of where to go next. This means 3rd party servo manufacturers like us and Savox can't strike out on our own, or put another way, we're stuck with what we have.
Fortunately, current flow of +10A when stalled is momentary, e.g. during a rifle roll, wall, or crankshaft maneuver, meaning not continuous for very long, so in the real world, the 3-pin DuPont style universal connector soldiers on. Anyway, here we go . . .
LiPo versus LiIons
Similar voltage (3.7V/cell vs 3.6V/cell - 7.4V vs. 7.2V nominal with a 2S pack). Similar enough to say macht nichts when selecting since we can measure the differences in the lab but you'd be hard pressed to feel any difference in flight.
Anyway, for the same capacity, the LiIon is a little bit heavier than LiPo because of packaging (LiPos are built within a plastic bag vs. while LiIon are built within a cylindrical metal shell). Both are great for control electronics, but my personal preference is for the LiIon in this application because of the more rugged construction (metal shell protects better than thin plastic bag against the knocks of everyday life). This decision is 100% down to the protective metal shell.
For example, if a pack shifts slightly during a maneuver (G-forces during a crankshaft are incredible) and then bumps against the hard edge of a plywood former, then a LiPo will likely be creased or dimpled, which damages the electrolyte inside. Means it then poses a fire danger going forward. Not the kinds of risks *I* like to take, but you do you. Since a pack built within a metal shell is less likely to be damaged under similar circumstances, then I view this as a no-brainer. Some guys will say, the servo makes more power on the extra 1/0th of a volt, which is true and come to a different conclusion and opt for a LiPo. Again, you do you but I have wife and family living under my roof, hone has a significant component of wood and other flammables, so I'd rather the LiIon.
So going forward, we're discussing control applications of LiIon only
- Real world, LiPo has better power to weight so in those terms, LiPo wins every time.
- Major benefit of LiIon over LiPo is metal shell construction - more rugged, so LiIon wins.
- LiPo delivers more current - but - either delivers enough current so it's a wash.
- Major issue with receiver packs is . . . how to get the power (current) to the servos?
Boxes
There are a bunch of aftermarket 'boxes' whose chief purpose is routing power. Note, when I say routing power, I am speaking about routing current, e.g. amperage, symbol (A) versus voltage, symbol (V).
Examples are a multitude of PowerBox (various models), Futaba DLPH-1 Dual Link, Smart-Fly BatShare DELUXE Ultra, RCCSKY Powerbox Pro, CORTEX Pro, Booma AR824, Powerbox Core, Emcotec DPSI 2018, etc., etc., etc.
So there are a TON of ways of 'solving' the power distribution problem. And yes, it needs to be addressed - but - I'm just not convinced adding more hardware is the answer. More later.
Point being, these boxes have proliferated for a reason. And the above is just a partial list because there are MANY more available than I have listed. Moreover, in order to one up each other, they have added features over time. E.g. some are regulating the voltage, or adding gyros, the ability to connect a 2nd receiver, and a 2nd battery. And all this compounds a problem I perceive, which is added complexity!
Added to which, the sellers don't mention it *but* while the idea of regulated power is nice . . .
- Regulated power is dirty - it's using FETs to synthetically make it - link to learn more.
- Presumably there's enough current (or more than a BEC circuit off an ESC).
- 2 receivers means a 100% greater chance of receiver failure - yet box must 100% work.
- issue with 2nd battery is the same, 2X greater chance of battery failure - box must work.
- Real issue is a box adds complexity - lot more shit to go wrong.
So THE fundamental problem with a power-type box (anybody's box) is everything has to work 100% of the time. If one lousy component fails, then it means chance the entire box fails. Since the box is a critical link in the chain (meaning it sits ahead of everything within the model (like ahead of the receiver and battery), then this also means you're gonna crash your $5000 model. Don't believe me? Open one of those things up for examination and start counting. Then maybe you'll realize how those boxes have a LOT of extra electronic and mechanical components!
So my being a proponent of KISS (keep it simple, silly), means I use packs built with four leads instead of two. Strike you as odd to add two more leads to make the overall model more simple? Yeah, I know but continue reading. And note, we're not unique, others offer dual lead packs, also.
So now you're wondering, four leads . . . what for? Here's what for, two of them power the receiver instead of one. Works because receivers use a 'bus' structure of management so you can plug the power in anywhere!
- Dual leads on a pack mean twice the current delivery to your receiver - reliably
Enter the PXED
Why the XT30? Simple, because it's an inexpensive yet capable power connector. So the PXED, plugs into the XT30 on the battery pack (rated at 30A) and presto, you get two more DuPont connectors to use as you please. Again, remember, each DuPont is rated at 3.5A continuous, or 5A intermittent (and servos are intermittent loads).
- A reliable way of extracting power from a battery without sacrificing mechanical compatibility
So just how much current will a battery give you? The above 2500mAh pack rated at 10C? Easy math, 10C x 2.5Ah = 25A continuous. That's a lot.
Bigger battery, say 5000mAh, also rated at 10C? Same math but now it's 10C x 5.0Ah = 50A continuous (remember, 5000mAh is the same as 5Ah), so there simply aren't enough servos to worry about them making the pack voltage sag, believe me!
What about those who say they can feel the difference in servo performance at the beginning and end of a flight? I can't and seriously doubt anybody can, but who can contradict them? Certainly not me, but it's my opinion a 1/10th of a volt difference during the flight is indiscernible, but modelers are well known to engage in mental masturbation and if this is their justification for using a regulator, despite the known downsides, then once again, you do you.
- Delivering a whopping 50A if the model needs it, this is enough current without a box system
So using the PXED adapter to give you 2 more power leads to an ordinary receiver (4 total), with each rated at 3.5A continuous, or 5A intermittent means the receiver can draw up to 14A of continuous current without heat build up. However, please note, in the real world, two leads to the receiver is plenty. Thing is, the fellow asking the question is asking about a 3m (118") model.
Switches
This brings us to switches and a recent fad . . . magic switches. My view? These represent yet more components to fail. Sure, these things work fine - but - only for long as 100% of the components are functioning. Me? I favor no switches at all - aka, *no* switch to fail.
This is very easy to implement with the typical 3D-type model which opens up from the canopy to the cowl with a few screws as having access to the guts makes it easy-peasy to make/break the power connection to the receiver.
So I'd much rather eliminate the switch altogether. Then I just use a couple extensions, which means the pilot becomes the switch! Believe me, there are a lot of benefits to zero added electronic components (my stock in trade, I know, believe me). Note, these DuPont connectors are good for maybe 500-1000 interconnects so every few years toss the extensions and buy new ones. Cheap insurance.
Best practice
For my own 42% model, I favor two receivers instead of one - but - I like having them independent. That's right, I expressly DO NOT want to tie them together with a box system of some sort because if it goes TU (tango uniform) results in a crash. Why? As usual with me, keeping things simple means zero added hardware where failure means a crash. How?
Very simple, Velcro two receivers in the turtle deck (behind the pilot's head). Each carries only 50% of the model's total current load. This is because one receiver gets RIGHT ailerons and elevator servos (plus rudder). The other receiver gets the LEFT side (plus throttle). Sweet!
So now you get 100% redundancy in your receivers - but - without any added box hardware. Moreover, because there's no box, then what's not there can't fail, capisci? Similarly, with my philosophy, there's no switch to fail, either. Are you starting to grok what I mean about simplicity? Anyway, this method of adding redundancy gives two major benefits.
First, there's no box to fail - AND - the current loads are reduced by 50% because you're splitting them between two receivers! Believe me, reducing current flow through a receiver is a good thing when dealing with servos capable of drawing large amounts of current!
OK, so what 'if' a receiver fails? Simple, you land the model using the other one because half of my flight controls are better than zero. Anyway, in theory these box things with two receivers are great because one fails it switches over and you keep flying *but* if it's the 'box' itself that takes a crap, then the model is toast because the box sits between both receivers and the crash site!
Meanwhile, if a receiver fails with the split load method, yes, you lose half your control power but 50% control is easily enough to land the model. Basically means receiver failure is not the end of the world! Easy peasy.
Note; what I am expounding is a VERY controversial position. And you will find plenty of folks to say I'm crazy. I can't help this any more than Copernicus was called crazy for saying the Earth wasn't the center of the universe.
Summarizing benefits
If a receiver goes teats up . . .
- Land with 50% control authority (servos in this situation tend to self feather). It's good enough.
- Less complexity! I cannot possibly emphasize this enough. Complexity is why planes crash.
- No switches - best switch in the world is the one on the workbench.
Downside of splitting power up? Honestly? It means 2X the odds of experiencing receiver failure. This is just statistics at work. However because the receivers are 100% independent without relying on yet more components, this means statistics work in your favor because odds of both receivers failing at the same time are astronomical. And we already know we can land with a dead one and saving my model is my primary concern as a consequence of component failure.
Basically, it means one receiver may fail but the pilots retains some function. Me? Been doing this too long to favor adding a box to models that don't really need it (more later).
Wrap up
So the above is 100% true for simple models like an Extra. Switch to a large jet, maybe 20 channels in use? Honestly, it probably means I also switch tactics and begin using a 'box', also. Dunno because I haven't faced this issue, yet. Maybe I can make it work with a pair of 14-channel receivers. I'll cross that bridge when I get to it. For now, for my scale and XA models, the independent receiver route makes more sense in practice than relying on a box does in theory.
Last thing; someone always trots out how regulated power is more consistent because a battery's voltage goes down over time. True, but immaterial. Why? Simple, it's because the discharge curve is so flat you can measure this voltage drop over time in the lab - but - in the field it's just one of those things guaranteed to stir up an argument amongst pilots. Saying it doesn't matter, not in the real world. And note, if you're so good you can tell, then recharge the packs between flight!