Servo current consumption

[John Muir]

I'm strictly a 4 S3003 man myself so this subject is purely of theoretical interest to me but I was wondering about the difference between analogue and digital servos. Here's my take on it, could somebody tell me if I've got this wrong or right.

Analogue servos are only powered when they receive pulses from the receiver 50 times a second. This is the reason they do not hold position under load as well as digitals - they are constantly being 'nudged' back into position by the pulses and 'relaxing' in between. Each servo receives its control pulse from the receiver in turn, i.e. sequentially, so that the the current requirements for a number of servos isn't cumulative. The peak current registered will be the peak reached by the single most heavily loaded servo. Effectively the current peaks 50 times a second for each servo, but those peaks, by definition, can't overlap, so the overall current requirement is kept relatively low.

Digital servos on the other hand, are powered almost constantly when under load, regardless of the control pulses from the receiver, thereby improving their holding and centering capabilities. However, because of this, the current used by multiple servos is cumulative.

So if you have a four servo setup using analogues, where the peak current on one servo is, say 0.3A, then the total current draw should be roughly 0.3A, because the servos are driven sequentially.

A four servo setup using digitals each peaking at 0.3A may have a total current draw of 1.2A, because the servos are being driven simultaneously.

 

Here's the webpage that got me thinking.

If I'm right, then I can see why digital users would be much more concerned about their batteries. Has anybody run a test on multiple analogue servos, to see if the current does add up or not? I expect I've actually got the wrong end of the stick, though. I very frequently do.

[Steve Fish]

For my 88" Edge 540 I decided to engineer some margin of safety and so bought a Spektrum AR9110 receiver system. This is a power bus connected to at least 3 satellite receivers and has dedicated low resistance wire inputs for 2 packs. Recommended voltage is between 6v and 10v (!) but I have servos with torque ratings given for 4.8v and 6v... refer to my prev message re: max working voltage on a servo.

Helpfully however this product comes with a very detailed 26 page manual written in proper English, not Engrish. Will you indulge me while I refer to what it says on page 5?

 

"Battery Capacity"

It's important to select a battery(s) that has more than adequate capacity to provide the necessary flight time. Our staff has been recording in-flight data to determine typical current consumption of aircraft in flight. Following are two graphs that illustrate the in flight current draw of the radio system.

 

"Note: Current draws may vary depending on your servos, installation and flying style.

 

" The following setup is shown as a worst case scenario indicative of some aerobatic pilots' setups. It is not recommended to use this setup without proper voltage regulation for your servos.

 

40% Yak

9 JR8711's + 1 8317 (throttle)

 

Batteries - two 4000mAh 2 cell 7.4v Lipos

Regulator: NONE (Note: The above servos are rated at a maximum of 6 volt 5-cell use. Using higher voltages will void warranty).

 

Engine: DA150

Weight: 40lbs

Envelope: Hard 3D

Average current: 2.62 amps

Peak Current: 17.8 amps

Milliamps used per 10 minute flight - 435mAh "

 

 

It goes on to say that it's recommended that only 60% of battery capacity be used to ensure plenty of reserve. In the above, 8000mAh x60% =4800mAh, divided by capacity per 10 min flight (435mAh), allows 11 flights per charge.

 

 

Another example:

33% sukhoi

7 JR8611's and 1 8317 (throttle)

1x 4000mAh 2 cell 7.4v lipo

Regulated to 6v

Engine DA 100

Weight 26lbs

FLight envelope: Moderate 3d

Average current .82 A

Peak 6.92 A

Milliamps used per 10 min flght: 137mAh.

 

For each example it gives a graph of current vs time.

 

Clearly it's the peak current draw and the ability of the pack to provide this without drop in voltage that is of concern as opposed to current capacity as such.

Interesting that unregulated 2S lipos were used for non-high voltage servos

Steve

[Tim Mackey]

Those are some serious high currents - but of course, they are seriously large and fast models. I appreciate that its the ability of the supply to maintain its level at peak times which is important in order to avoid a brown out however it would have been interesting to see the data graph, and then at least see for what duration these peak figures applied.

I note also the average current in both cases was relatively low.

Returning to the brown out scenario - although of course I would never willingly "permit" or "condone" a PSU that allowed a brown out, I cant help feeling that providing the receiver has the latest QC firmware, the likely impact of a brown out due to brief voltage drop would probably never even be noticed, until the model was landed, and the observant pilot noticed the slow flashing LEDS.

I think my ideal PSU setup for many models would be one which used a BEC supply ( either UBEC or ESCBEC) as primary power, with a separate battery as back up, which only activated in the event of a failure of the primary power, and of course provided visual indication of such an occurrence.

Most "simple" dual battery systems only allow for two battery systems not a mix of systems, and many of course, simply use two batteries and drain them both equally - thats hardly a true back up IMO.

 

I am also somewhat surprised that they used an unregulated 2s lipo on servos which they specifically state are designed for maximum 6V !

[Steve Fish]

Posted by Tim Mackey - Administrator on 27/03/2011 22:48:13:

... it would have been interesting to see the data graph, and then at least see for what duration these peak figures applied.

 

They did provide a data graph and there were dozens and dozens of spikes well above 12A, but in the interests of © I thought better of scanning it in... the manual is available online here and is well worth a read or flick through just for interest's sake.

In case the link does not work - www.spektrumrc.com/ProdInfo/Files/AR9100_Manual_LR.pdf

 

I note also the average current in both cases was relatively low.

Yes Agree but still - 2.6 A is up there compared to workaday sport models.

I think my ideal PSU setup for many models would be one which used a BEC supply ( either UBEC or ESCBEC) as primary power, with a separate battery as back up,

You cannot do this with the Spektrum AR9100/AR9110: although it features true dual battery redundancy, as each battery is isolated and if one fails, the other takes over and the Rx switch is failsafe-on, both packs provide power, it seems, at the same time. I agree with you but I personally feel a Voltage regulator is another complex piece of electronics that has a potential to fail (with what consequence on a LiPO pack ... maybe misplaced apprehension) when it should not be necessary.

 

I am also somewhat surprised that they used an unregulated 2s lipo on servos which they specifically state are designed for maximum 6V !

 

I agree again ( and they do give a caveat to this!) but this relates to my earlier point precisely:

we should not have to rely on voltage regulation to supply 6v, or rather we should be reassured that 5cell and even 2S LiFe packs are ok for servos, as a charged 4cell pack does not give 6v and a 5 cell pack is on its way to being flat at 6v and potentialy well above 7v when charged. What we really need is for servo manufacturers to clearly state the maximum operating voltage (Voltage Not Exceeded or VNE - harking to real filght envelope ;-0) AND bring that in line with charged 5cell and 2cell LiFe packs - packs that a lot of people use to avoid confusion for pedants like me. Or do what Spektrum have done and give a proper working range:

Specs for the AR9110 and 9100 say-

"Voltage Input - 6.0 to 10.0 volts. Note: Consult your servo manufacturer's specifications for the maximum allowable voltage." I have not yet come across this parameter when looking at servos. I have just checked the Futaba website for an example of this and sorry I failed

It is assumed the max v for a servo rated for torque at 4.8v and 6.0 volts is 6.0 volts when this in fact is clearly not so in practice. Lets please campaign for more appropriate information on the servos we need.

 

Cheers for listening

Steve

 

 

Edited By Steve Fish on 28/03/2011 08:15:21

Edited By Steve Fish on 28/03/2011 08:30:12

[Tim Mackey]

Thanks Steve - I found the manual but the link doesnt work - no worries.

The graphs and interface look very much like the eagle tree logger graphs, and I noticed that the second example the peak current lasted only a fraction of a second, and was also only seen once - even before the flight commenced I suspect

I agree on the servo voltage rating thing, and I have often mentioned that its about time we had HV servos - happy on 7.4V - far more widely available, and at a sensible cost.

At the moment they are marketed as some sort of snake oil, which is both misleading, and un-necessary.

I mean - how much more can it cost to use components ratred at 7.4V instead of 6V

[Steve Fish]

Totally agree. But even 7.4v is slightly misleading because we need to know a useful working range as applied to packs available. I mean, 7.4v implies a 2sLipo which in actual fact will exceed 8 volts for some time. And then there's the example within the Spektrum manual of unregulated 2sLipos on standard servos.

 

I understand that an appplied voltage range will carry with it a range of life expectancy. So one may expect a servo to last 20-30% less time (for argument's sake) at higher voltage than at lower voltage. This may pose a bit of a headache to the manufacturer. On the other hand LiFe 2S packs quickly fall to 6.8-6.6v and stay there for some fair amount of capacity. I would hope that 6v+10% would be accommodated by most standard voltage servos without a significant detriment - but how can I tell?!

 

Sorry - out of time to edit the post and fix the link - I can see the problem.

Edited By Steve Fish on 28/03/2011 10:18:10

[Tim Mackey]

BUMP.

Couldnt help a small smirk when I read the latest edition of BMFA news, where it seems a Mr Trevor Waters concludes pretty much the exact same thing as I did way back over  5 years ago, on measuring real time actual current consumption of your average servo.

Edited By Tim Mackey on 21/03/2013 21:14:42

[Martin Harris - Moderator]

Posted by GONZO on 06/10/2010 18:39:00:

Take a charger to the field and recharge after a flight and note how much you put back in mAh. Do this for a number of flights and you will soon get a feel for how many flights you can do on a fresh set of cells( bet its more than you do on a flying session).

I've long advocated noting the charge put back into packs after a flying session with occasional test discharges to check the reserve capacity/condition of the pack.

Recently, I've been flying a model fitted with "high voltage" Corona DS339HV (558 on rudder)servos and running on parallel 1100 mAH A123 2 cell packs. With 6 servos, I've been quite surprised that after a couple of decent length flights (10 - 15 minutes each), recharges have been consistently in the 60 mA per pack range - theoretically worth around 18 flights on a single charge!

[Steve Hargreaves - Moderator]

I've become a devotee of LiFE Rx packs & have just about converted all my fleet over. What has been interesting is how much goes back into the pack.

Typically after three or four flights I put around 150mAh back into a LiFE pack......the 4 cell NiMH packs I used to use would take over 250mAh...even though they were LSD packs. In use I would expect the LiFE pack to flow a little more current due to the higher voltage & hence discharge a little faster than a 4 cell NiMH....I wonder if the NiMHs are in fact self discharging a bit...

I think one of the problems that manufacturers have with "high voltage"...or more correctly a wider voltage range servos is the speed differential between the lowest & highest voltages....a 4cell NiMH pack will be around 5 & a bit volts...a 2S Lipo around 8 volts or just under 60% higher. The difference in speed & torque between the lower & higher voltages would be quite significant.

Like Martin I'm exploring the Corona 558HV servos & whilst they certainly work on a 4 cell NiMH pack they are noticably slower.

[Steve Fish]

I suspect, like all electric motors, servos will demand a certain power determined by the load. Power is the product of current and voltage: watts. if the voltage is lower for a given load, I would expect the current draw will be higher than a situation where the reverse is the case. Perhaps this is why the lower voltage pack drains quicker than the higher voltage LiFe pack. Neither pack is providing any more power than the other for the same load. Techies on the subject may enlighten us further.

[FlyinBrian]

Posted by Steve Fish on 22/03/2013 10:21:53:

I suspect, like all electric motors, servos will demand a certain power determined by the load. Power is the product of current and voltage: watts. if the voltage is lower for a given load, I would expect the current draw will be higher than a situation where the reverse is the case. Perhaps this is why the lower voltage pack drains quicker than the higher voltage LiFe pack. Neither pack is providing any more power than the other for the same load. Techies on the subject may enlighten us further.

I can not understand how a lower voltage will produce a higher current. Increase the voltage for a given load and the current will go up.