pchristy
Senior Heliman
Location: Nr London, England
| First a bit of background: I'm a professional broadcast engineer, and have been building RC equipment since the mid 60's. Since the mid 70's I've designed and flown quite a lot of my own RC equipment, and some of it has been manufactured commercially.
I say this not to boost my ego, or to make you think that everything I say should be taken as gospel, but to assure you that I do have relevant qualifications, and do have experience to back up my opinions! But that is all they are - opinions! You may take them, or leave them!
Modern digital servos can draw an awful lot of current - in the region of AMPS! Some early ones were particularly notorious in this resepct! I once witnessed a very well known pilot write a model off due to flat nicads on the second flight following a full charge! And these were 3000mAH nicads!!!
Most modern servos now have current limiters to stop this happening, but can still draw a peak current - even if only momentarily - of several amps!
A modern helicopter may have five digital servos. Four of these (swashplate and tail) can be under extreme loads when being used for violent 3D. Even if they are not drawing maximum current all the time, the "surge" when the motor switches on can still be considerable.
Just taking the swashplate servos alone, you could be talking about a transient current of 10 amps.
This is exacerbated in the Spektrum system, because the swashplate servos (assuming eCCPM) are all driven simultaneously, rather than in rapid sequence, as when using conventional PPM/PCM.
This places an extreme demand on the connectors and switches - particularly the power lead, as this must carry the sum current of all the individual servos!
A resistance of 0.1 ohms would be enough to reduce the voltage seen by the receiver by 1 volt, assuming the figures quoted above. A dirty connector or switch could easily have an impedance this high!
Similarly, if you use a regulator, it must be able to handle sudden transients of high current without complaint!
There are two types of regulator - linear and switching. Linear regulators are very inefficient, and usually require large heat sinks to dissipate the heat. Switching regulators are very efficient, but unless they are *very* carefully designed, generate copious amounts of RF interference. To be efficient, they have to switch hard and fast, and fast edges result in large amounts of RF, which then has to be screened or suppressed somehow.
Suppression is difficult, because it is hard to distinguish between the edges caused by the switching regulator, and those caused by the very sudden high energy demands of three (or more) digital servos all kicking in at once!
Such a transient may not cause a "brown-out", but it *may* momentarily disrupt the smooth operation of the receiver!
The best way to overcome this is to make sure that there is as little as possible in the way of connectors, switches and electronics between the battery pack and the receiver.
A good quality and well maintained 1900 mAH nicad, coupled with a heavy duty switch harness should be able to supply all the current needed without the voltage sagging.
I believe that Spektrum are starting to fit heavy duty spade connectors to some of their newer receivers for the nicad connections to overcome these issues. About time! I've been wondering who would be the first to recognise that our standard 3-pin conectors are being pushed beyond the limit in modern applications!
If you are doing extreme 3D with a large model, I would also recommend topping up the batteries between flights!
If you are running a regulator, and it works for you, then fine! Carry on!
But if you are experiencing glitching problems, try replacing whatever power supply you are using with a *high-quality* 1900mAH (or so) nicad, and a matching heavy duty switch harness.
The results might well surprise you!
--
Pete |
I've found that they also have more than the rated capacity and hold up well under load . At present I'm using some 3300mA GP nimh packs . They also seem to retain their charge very well when stored .
