HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Mar 18, 2012, 10:21 PM
KE your cub.
Curare's Avatar
in the gutter, again....
Joined Jun 2005
4,098 Posts
Discussion
How to determine how much amperage you need for servos?

Hey all, just a quickie.

Im tempted to start using a regulated power supply for my servos, but I'm worried the reg won't supply enough.

the servos I'm using have a stalled current of 5 amps.

Does this mean that my reg should be able to feed all the servos (say 5) in series?

i.e 5amps x 5 servos = 25 amps?

that sounds excessive!
Curare is offline Find More Posts by Curare
Reply With Quote
Sign up now
to remove ads between posts
Old Mar 18, 2012, 11:49 PM
Registered User
Joined Jun 2011
218 Posts
Hi,
Those sound like fairly large servos but yes if you are using 5 and they can all actuate simultaneously (parallel) then you will need something substantial but perhaps not 5x5A.
They draw most current as they start to move and once moving smoothly the current falls back but in RC applications they are often used with many small corrections so are constantly starting then stopping. The other thing in your favour - are you likely to stall all 5 units at the same time? Probably not.
Even standard servos draw about 1A when starting but quickly fall back to about 1/8 of this so you need high peak current but average current could be considerably less. This suggests you could get by with perhaps a 2A regulator and a large capacitor to supply the peaks much the same as car audio guys do. This may need to be something like 4700uF 6 or 10V but if room does not permit then you may need to trial the regulator under variable actuator conditions and check if it drops out of regulation. You will need an oscilloscope to monitor this as the drop outs will be too quick for a meter to respond.

Cheers,
David
KiwiDavid is offline Find More Posts by KiwiDavid
Reply With Quote
Old Mar 19, 2012, 08:39 AM
Registered User
pilotpete2's Avatar
The Northeast Kingdom, Vermont
Joined Jun 2004
5,206 Posts
That's very excessive. First off, servos should never be able to stall in use if properly setup. If they did you'd cook the wires from the battery in a heartbeat at 25A.
Have you considered running a 2S A123 battery, very stable anf able to deliver the amps when needed
Pete
pilotpete2 is online now Find More Posts by pilotpete2
Reply With Quote
Old Mar 19, 2012, 08:51 AM
Registered User
Romania, Dolj, Craiova
Joined Sep 2007
14,652 Posts
A servo that stalls at 5A should get under 1A in regular service.
The key is not allow them to stall, more exactly limit the current per servo when more than normal operation.
This is done usualy in powerboxes using polyfuses, i.e. electronic fuses that cut current at, say 1.5A, and resets themselves, reestablish normal low conductivity when current surge stop.
renatoa is online now Find More Posts by renatoa
Reply With Quote
Old Mar 19, 2012, 08:57 AM
Registered User
pilotpete2's Avatar
The Northeast Kingdom, Vermont
Joined Jun 2004
5,206 Posts
Oh, I really like the idea of the Polyfuses. In larger planes with high torque digital servos, it might not be a bad idea to add one in each servo lead. Better to lose the bad servo, than take out all power to the receiver
Pete
pilotpete2 is online now Find More Posts by pilotpete2
Reply With Quote
Old Mar 19, 2012, 09:43 AM
Registered User
Romania, Dolj, Craiova
Joined Sep 2007
14,652 Posts
You can get ideas from this clever design, I can call it "the poor man's powerbox"

http://rc-miskolc.emiter.hu/index.ph...140&Itemid=134

The yellow components are the polifuses. If you google LP30 will find they are rated at 1.35A, and source to buy. There are many amp capacities.
The board also features power redundancy, allowing power from two packs, switchable by Schottky diodes, can be seen in the picture.
But the board is too cheap to copy instead buying it, if fits your needs.
renatoa is online now Find More Posts by renatoa
Reply With Quote
Old Mar 19, 2012, 03:14 PM
German Engineering.......
HugePanic's Avatar
Auxsburg
Joined Dec 2007
453 Posts
Several times i saw the same principle used:

Use the blocking amps, add two times the amps when cycling all servos.

So if you cycle all servos unloaded, you might end up with 1.5amps.
And we assume a blocking current of 5ampere.

You end up with: 2x1.5+5 = 8amps constant
HugePanic is offline Find More Posts by HugePanic
Reply With Quote
Old Mar 19, 2012, 09:45 PM
KE your cub.
Curare's Avatar
in the gutter, again....
Joined Jun 2005
4,098 Posts
Quote:
Originally Posted by pilotpete2 View Post
That's very excessive. First off, servos should never be able to stall in use if properly setup. If they did you'd cook the wires from the battery in a heartbeat at 25A.
Have you considered running a 2S A123 battery, very stable anf able to deliver the amps when needed
Pete
I'm currently using A123 cells for everything, they're brilliant, however this was a flight of fancy, caused by the 5amp rating on the powerbox digi switch, it seemed a little weak to me for 5 digital coreless servos.

I'll stick with what I have for now
Curare is offline Find More Posts by Curare
Reply With Quote
Old Mar 20, 2012, 01:06 AM
Registered User
Toowoomba, QLD, AUSTRALIA
Joined Jan 2008
614 Posts
KiwiDavid hit the nail right on the head, a stalled servo and a stationary servo draw the same current, just that the stationary servo current drops really quickly as it spins up.

I'd expect 25A peaks IF all servos were stationary AND asked to move at the same time. Actually my brief tests with the digital servos in the back end of my Aries revealed something interesting. While the signal from the RX was sent to each servo at the same time, the next control pulse from the servo internals was dependant on when the next of the 300hz control outputs were timed.

To cut a long story short, the amplitude of the two current pulses of my 2 x S9650's unloaded but buzzing away in the back end from my Aries was about 1.5 amps each. I was triggered on one pulse, while the other slowly cruised along until they both met and they formed a single 3A peak, then dropped back to two 1.5A peaks. These pulsewidths are only microseconds wide because the servo was unloaded, but press on the surface and the pulses got wider BUT the peak doesn't. As you would expect from a PWM system.

So I run a 10A reg in my Aries, 3x S9650's 2x BLS154's and a S9255, even though in practice my average current draw is only about 0.4A. A side note is that my reg would cook if I asked it to supply 10A continuous (for a couple of minutes) from a 2S li-po as it just isn't heatsink enough to dissipate the watts.

In your case, any reg that can't supply 25A continious will be relying on capacitors to make up the shortfall, which is perfectly acceptable if the shortfalls are rare and tha caps are sized appropriately.
bjr_93tz is offline Find More Posts by bjr_93tz
Reply With Quote
Old Mar 20, 2012, 02:07 AM
KE your cub.
Curare's Avatar
in the gutter, again....
Joined Jun 2005
4,098 Posts
Well Brett, you're basically in the same boat as me. Thinking about it, we do occasionally get close to synchronized servo moment, say in a snap, (but lead with the elevator!) and as you know this does put a fair amount of load on the servos.

My concern was if i had an inadequate reg I may pop it just as I enter a snap, and finsh by snapping every piece of aircraft that can be snapped.

What reg are you running?
Curare is offline Find More Posts by Curare
Reply With Quote
Old Mar 20, 2012, 04:30 AM
Registered User
Toowoomba, QLD, AUSTRALIA
Joined Jan 2008
614 Posts
Home design reg.

It has two linear regulator IC's, low drop-out type, rated for 5A each. A couple of resistors set the output voltage and your input needs to be about 0.7 volt above that (that's the "low drop out" bit). A normal reg needs about 1.5V overhead.

It's really agricultural, but I incorporated some nifty track design to ensure both regs share the current equally. I'll post more details as the design is on my machine at work.

Sizing a voltage reg falls into 3 categories, #1 something that works most of the time but drops out ocassionally, #2 something that works without problems all the time and, #3 something that works without problems all the time AND lets you extract the full potential from your servos under all conditions.

For the vast majority of us, #2 is perfectly fine. I wanted #3 just because I'm that way inclined.
bjr_93tz is offline Find More Posts by bjr_93tz
Reply With Quote
Old Mar 24, 2012, 11:30 AM
yyz
Registered User
yyz's Avatar
USA, CA, Paso Robles
Joined Dec 2004
2,400 Posts
Quote:
Originally Posted by bjr_93tz View Post
Home design reg.

It has two linear regulator IC's, low drop-out type, rated for 5A each. A couple of resistors set the output voltage and your input needs to be about 0.7 volt above that (that's the "low drop out" bit). A normal reg needs about 1.5V overhead.

It's really agricultural, but I incorporated some nifty track design to ensure both regs share the current equally. I'll post more details as the design is on my machine at work.

Sizing a voltage reg falls into 3 categories, #1 something that works most of the time but drops out ocassionally, #2 something that works without problems all the time and, #3 something that works without problems all the time AND lets you extract the full potential from your servos under all conditions.

For the vast majority of us, #2 is perfectly fine. I wanted #3 just because I'm that way inclined.
Thank you very much for articulating your thoughts and ideas and for offering to share your design. I'd be very interested in this.

I currently fly large 1:3 scale sailplanes that can have upwards of 14 servos. I am currently using redundant A123 (two cell) packs feeding a multi-battery input Multiplex receiver and "high-voltage" (6+ volt rated) digital servos.

I have a boat load of the older, lower voltage servos that I would like to use w/ the A123 batteries.

Mike
yyz is offline Find More Posts by yyz
Reply With Quote
Old Apr 04, 2012, 10:27 PM
Registered User
Toowoomba, QLD, AUSTRALIA
Joined Jan 2008
614 Posts
Sorry for the delay in returning. The Altium licence doesn't want to play the game with my work machine, so basically every PCB I've ever designed can't be accessed by me anymore I'm off next week for a week of R&R

Simple story is I use the Micrel MIC29502BT linear regulator, it's good for 5A, two resistors sets the output voltage and there's and on/off pin. There's a 7.5A version available. The datasheet makes good reading.

I have the on/off pin tied to the supply so it's always "ON" but you can tie the pin to the supply with a suitable resistor and turn the regulator "OFF" by pulling the pin to ground with a mechanical switch (or a logic low from a microcontroller). The key here is that the mechanical switch needs to close to turn it off so essentially the regulator fails to the on condition. I don't use it this way.

In my dual reg, I made a couple of tracks of certain lengths and widths to introduce some resistances into the design so that current would share a bit more equally between regs. The output impedance of linear regs is fairly low, in this case you'd only drop 0.01V at 5A so if one reg was at 5.01V and the other was at 5.00V then one reg would supply all the current (and run hotter) before the other would kick in. A little bit of track resistance in the right spots and very carefull voltage matching (no two regs or resistors are the same) means they both run at close to the same temps at 50% load.

In any case, with selecting a voltage reg (linear type) I'd look at the absolute worst case current demand (say 15A if running at 6 volts), then guess that the worst current at 3V would be about 7.5A and use no less than a 7.5A reg. If you could pull 15A at 5V then you'd need 9A capacity to hold 3V. I'd much prefer the long term current carrying capacity (at worst case) of the reg be able to hold at least 3V-3.5V. The capacitors will sort out the transient current demands but you know you can never get pulled into a brownout reset.

Just a note on filtering capacitors, think that if you tried to pull 15A (lets hit the snaproll button) from a 2200uF capacitor charged to 5volts, nearly all your volts will be gone after just 0.5 miliseconds (hundreds of times quicker than a blink of the eye). I doubt the servo's would even twitch...
bjr_93tz is offline Find More Posts by bjr_93tz
Last edited by bjr_93tz; Apr 04, 2012 at 10:46 PM.
Reply With Quote
Old Apr 04, 2012, 11:57 PM
yyz
Registered User
yyz's Avatar
USA, CA, Paso Robles
Joined Dec 2004
2,400 Posts
Thank you very much! That's what I was looking for.
yyz is offline Find More Posts by yyz
Reply With Quote
Old Apr 09, 2012, 08:17 AM
Registered User
Joined Oct 2010
1 Posts
Hi!
My question is:
What the max powerload the receivers himself without powerbox?
For exemple what max servoload without powerbox?
I like use for exemple 6 pcs 10 Kg servo without powerbox,is possible?
If yes,where is the possible Ampersload in the receivers?
Thanks!
George
madgeorge1831 is offline Find More Posts by madgeorge1831
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion How Do You Determine What Lipo C Value Is Needed? itsme2 Batteries and Chargers 15 Aug 11, 2014 02:51 AM
Question How to determine what "C" rating you need? mag1 Batteries and Chargers 26 Jan 18, 2012 12:48 PM
Question How do you determine,(calculate) the number of climbs on any given motor, prop combin hog2soar Electric Sailplanes 4 Jan 10, 2012 04:17 PM
Wanted Just outn of curiosity, how much would you pay for.... BladeScraper Aircraft - Electric - Helis (FS/W) 0 Nov 25, 2011 07:37 AM
When choosing servos how do you determine torque needs? mexico Radios 3 Oct 24, 2005 11:31 AM