Voltage drop along servo leads

[Grasshopper]

I have been testing voltage drop through a typical receiver and a servo extension lead and found some  - perhaps unsurprising results.

 

Set-up was a 4 cell AA NiMH battery (2100 and 2500 mAh tested)  through a switch to a Multiplex receiver and then off to 6 servos. The test servo is connected via a 30cm extension and then via a 50 cm extension to the servo connector.The join between the 30m and 50cm leads would typically be that made at the wing/fuselage 'join' in the a/c.

 

Using an oscilloscope I measured the voltage at the battery and at the servo connector input when the servo was made to work for its living. This allowed me to see the short term pulses of lower voltage as the servo motor gulped current from the battery.

 

Results were interesting: at the battery terminals .2 volt drop and at the servo connector .8 volt drop. This means that on a 4 cell setup the voltage supplied to the servo will drop from 4.8 volts to 4 volts (or less!) as the motor runs.

 

The losses are in the battery to receiver wiring, within the receiver itself and on the extension leads to the servo connector.

 

I then connected directly from the battery switch output to the join of the servo extensions using high current wire such as that from Giant Cod - Super flex silicone (0.13 Ohm/m 18 AWG) This bypasses the receiver and its no doubt lossy circuit board and also the 30cm extension. This was now much better and the voltage drop was reduced from .8 volts to .3 volts.

 

There was however a lot of  'noise' on the supply - several volts in fact and after some experimentation I found that a 47u tantalum capacitor placed across the 'end' of the high current wire feed killed the noise stone dead.

 

The final test was to check if high current capacity NiMH batteries are less good at supplying pulse current then lower capacity ones. I compared a 4 cell NiMH 2100 mAh pack against a 2500 mAh pack and found that they were almost identical in terms of output voltage drop (at the terminals) when hit with a current demand such as a servo will give. In fact the 2500 was marginally better.

 

 In practical terms then to reduce the voltage drops, wire directly from the battery switch/plug to the last 'in fuselage' connection using suitable wire and place a 47u capacitor across the supply at this point. Result, voltage drop reduced from .8 volts to .3 volts and high frequency noise all but eliminated.

 

 Worth the effort - who knows !

 

 

Similar theme over HERE  the current capacity of typical servo leads /connectors

Edited By Tim Mackey - Administrator on 09/10/2010 22:36:21

[Myron Beaumont]

Grasshopper

Doesn't this point to the switch contacts ? (we never see the insides do we?) Could you explain how your capacitor (one of those light brown round things from the olden days I presume) reduced the voltage drop . I don't mind if you get a bit technical !

[Grasshopper]

Myron,

 

well yes, there is some loss in the switch of course but the majority occurs in the extension lead and as you would like more info. here it is.

 

Fully charged 2500 MaH NiMH, slide switch plus charge socket, Multiplex RX-7 Synth receiver, extension as noted and Hitec HS 82MG servo.

 

Readings are a a little 'approximate' since these are of voltage pulses below the idle level of 2 - 8 milli second duration as seen on the 'scope screen and as the disturbances are caused by the servo motor turning so the actual amplitude is always varying slightly but I have trued to amortise the readings as best as I can.

 

At battery terminals -- .15 volts

At exit from switch/charge socket (receiver input) -- .2 volts

At exit from receiver -- .28 volts

At end of .5 metre extension -- .7 volts

At end of 1 metre extension .8 volts

 

As I noted in the original post if you use low resistance wire to by pass the switch/receiverextension then you reduce the voltage drop considerably. In this case connecting from the switch/charge socket output to the end of a .5 metre extension would reduce the voltage drop by ((.7-.2) - .1)) .4 volts and with a 1 metre extension (.8-.2) + .1))  .5 volts. This is only practicable of course if the low resistance wire is run to the end of the extension for example to the fuselage/wing join of a wing mounted servo loom.

 

As for the 'light brown  round things from the olden days'  -  in this case a very modern Tantulum capacitor of 47 uF these do reduce the 2 to 8 ms voltage drop pulses by a very small amount as they provide a local source of current to feed the servo motor but unless they are large in value (say 10,000 uF) they will not do a great deal in this respect. What they do in this application is to almost completely kill the 1.2 + volt spikes of low voltage caused when the servo motor starts to run. These occur as a burst of about 15 to 20 pulses occupying .1 ms in total. These may appear insignificant but nevertheless spikes such as these can cause circuitry to misbehave! If for example a servo contains a microprocessor - essential in the case of digital devices with brush-less motors, spikes of this size could cause the micro to be reset with predictable results.

 

 

 

 

 

 

 

 

 

 

 

 

 

[Myron Beaumont]

Thanks Grasshopper .Lots of little losses Eh?

Especially with aileron extension leads and such like ?

Hmm  Interesting as I no longer have access to an oscilloscope etc.

Thanks again  Myron

[Grasshopper]

No problem,

 

I find winding down from  my 'day job' of hardware and software design of embedded micro systems and products - is very difficult !

 

As a result I am always on the lookout for projects - so if you are ever in need of someone to look into anything or to design a 'widget', drop me a line. Always pleased to assist.

 

Peter