Digital Servo

Digital Servo

Post by Wijoyo Utom » Sat, 27 Jan 2001 01:21:46



Hi,

does anyone know what is really digital with the digital servos? It still
receives a pwm from the Rx doesn't it? Or maybe the position is controlled
digitally through optical encoder (maybe <G>)?

Pardon me for my ignorance over this matter.

TIA,
W. Utomo

 
 
 

Digital Servo

Post by Yarble » Sat, 27 Jan 2001 03:58:12


From what I can see it has to do with how the motor is pulsed for accurate
holding power. The have the ability to supply near max torque right away.
This is done digitally.... I guess that might be why the name. They do need
a name to tell them from normal corless servos since they are stronger and
more accurate



Quote:
> Hi,

> does anyone know what is really digital with the digital servos? It still
> receives a pwm from the Rx doesn't it? Or maybe the position is
controlled
> digitally through optical encoder (maybe <G>)?

> Pardon me for my ignorance over this matter.

> TIA,
> W. Utomo


 
 
 

Digital Servo

Post by Pete » Sat, 27 Jan 2001 04:36:05


it has a 13 bit A/D converter on the pot. All the processing can be done
digitally but its not true digital yet.

It needs the PCM encoded data fed directly to it and an optical encoder on
the output shaft. Then we would have something truly digital.

--

Peter Wales in Sunny Orlando

http://home.cfl.rr.com/pwales


Quote:
> Hi,

> does anyone know what is really digital with the digital servos? It still
> receives a pwm from the Rx doesn't it? Or maybe the position is controlled
> digitally through optical encoder (maybe <G>)?

> Pardon me for my ignorance over this matter.

> TIA,
> W. Utomo

 
 
 

Digital Servo

Post by aerografixs.. » Sat, 27 Jan 2001 05:21:03


Go to this site for some fairly good explanation.

http://www.multiplexrc.com/mc_servotips.htm

Roger

Multiplex dealer
Pure Light Fiber Optic (servo extensions)
http://www.aerografixs.com
"The significant problems we face today,
cannot be solved at the same level of
thinking we were at when we created them"



Quote:
> Hi,

> does anyone know what is really digital with the digital servos? It
still
> receives a pwm from the Rx doesn't it? Or maybe the position is
controlled
> digitally through optical encoder (maybe <G>)?

> Pardon me for my ignorance over this matter.

> TIA,
> W. Utomo

Sent via Deja.com
http://www.deja.com/
 
 
 

Digital Servo

Post by Kjell Arne Ga » Sat, 27 Jan 2001 08:28:36


Quote:

> it has a 13 bit A/D converter on the pot. All the processing can be done
> digitally but its not true digital yet.

Why 13 bit? I guess this could vary from implementation to
implementation and that some manufacturers will start marketing servos
with 14 bit converters. And soon we will have the same hype as we did
with the CD players.

I think that the main advantage of digital servos is the improved
controller. It is easier to make a good controller in the digital domain
then doing the mathematics in analog electronics. But this is not fair
against the analog servos as much of the improvement comes from the fact
that they are running the control loop more often. This could also be
done in analog servos and i guess this is what is done in the so called
"Super Servos". So then it all comes down to production cost. The
manufacturers can make better servos at the same or at a lower cost.
Then the question is how we, the pilots, will benefit from this. How
soon will we get the better digital servos at the same or a lower price
then the analog ones? This will finally happen in the low-end market,
but when? The high-end market is insane and will stay like this or even
more so with the digital servos.

Quote:
> It needs the PCM encoded data fed directly to it and an optical encoder on
> the output shaft. Then we would have something truly digital.

I would also like a optical encoder on the output shaft. Not because it
is digital but because it will not wear out as a pot.

--

 
 
 

Digital Servo

Post by Pete » Sat, 27 Jan 2001 09:02:09


The main advantage of digitals is the utter repeatability of the position.
Because they can have high power just off the set point they can move to the
set point easily and the accuracy of the the converter gives them accuracy
in position.

The analogue ones were a balance between accuracy/power and
overshoot/hunting

--

Peter Wales in Sunny Orlando

http://home.cfl.rr.com/pwales



Quote:

> > it has a 13 bit A/D converter on the pot. All the processing can be done
> > digitally but its not true digital yet.

> Why 13 bit? I guess this could vary from implementation to
> implementation and that some manufacturers will start marketing servos
> with 14 bit converters. And soon we will have the same hype as we did
> with the CD players.

> I think that the main advantage of digital servos is the improved
> controller. It is easier to make a good controller in the digital domain
> then doing the mathematics in analog electronics. But this is not fair
> against the analog servos as much of the improvement comes from the fact
> that they are running the control loop more often. This could also be
> done in analog servos and i guess this is what is done in the so called
> "Super Servos". So then it all comes down to production cost. The
> manufacturers can make better servos at the same or at a lower cost.
> Then the question is how we, the pilots, will benefit from this. How
> soon will we get the better digital servos at the same or a lower price
> then the analog ones? This will finally happen in the low-end market,
> but when? The high-end market is insane and will stay like this or even
> more so with the digital servos.

> > It needs the PCM encoded data fed directly to it and an optical encoder
on
> > the output shaft. Then we would have something truly digital.

> I would also like a optical encoder on the output shaft. Not because it
> is digital but because it will not wear out as a pot.

> --


 
 
 

Digital Servo

Post by Kjell Arne Ga » Sat, 27 Jan 2001 10:15:51


Quote:

> The main advantage of digitals is the utter repeatability of the position.
> Because they can have high power just off the set point they can move to the
> set point easily and the accuracy of the the converter gives them accuracy
> in position.

This all comes from a improved controller.

To really hit the set point you will need integration in the control
loop. To increase the power when pushed/pulled off the set point you
will need more control loop gain.

You can increase the control loop gain without loosing stability if the
sampling frequancy is increased. You can also have more gain in certain
regions by adding complexity to transfer functions in the controller.

Most of this can also be achived in an analog circuit but the limiting
factors are cost and stability when in large scale serial production.
The hunt for more performance with analog circuits halt when they have
to add trim pots to achive control loop stability.

When the controller is running in a computer it can also relatively easy
be made adaptive. The control loop parameters will  then be adjusted, in
run time, to compensate for deviation in the motor parameters due to
production and age. Then we can have even higher loop gain and better
servo performance. If this is done in digital servos of today i do not
know. What i know is that this is REALLY difficult in an analog circuit.
Also do remember that the production cost is still the same as the
program has to be written only once.

Quote:
> The analogue ones were a balance between accuracy/power and
> overshoot/hunting

And the digital ones are not???

--

 
 
 

Digital Servo

Post by Pete » Sat, 27 Jan 2001 11:12:53


To a much lesser extent because you can turn up the gain as the servo
approaches its set point.  With an analogue servo the gain is fixed and if
it is too high the servo overshoots and hunts and if too low it will stop
short of the setpoint.

Of course, the gain changes dramatically depending on the damping effect of
the load and a collective servo will be much more damped than a throttle
servo, but the analogue servo has to be capable of producing good accuracy
in either application. With a digital servo it is much easier,because, as
you say, as the control system is much more versatile.

Peter Wales in Sunny Orlando

http://home.cfl.rr.com/pwales



Quote:

> > The analogue ones were a balance between accuracy/power and
> > overshoot/hunting

> And the digital ones are not???

> --


 
 
 

Digital Servo

Post by Alan Jon » Sun, 28 Jan 2001 11:54:19


On Fri, 26 Jan 2001 02:15:51 +0100, Kjell Arne Gai

Quote:


>> The main advantage of digitals is the utter repeatability of the position.
>> Because they can have high power just off the set point they can move to the
>> set point easily and the accuracy of the the converter gives them accuracy
>> in position.

>This all comes from a improved controller.

>To really hit the set point you will need integration in the control
>loop. To increase the power when pushed/pulled off the set point you
>will need more control loop gain.

>You can increase the control loop gain without loosing stability if the
>sampling frequancy is increased. You can also have more gain in certain
>regions by adding complexity to transfer functions in the controller.

>Most of this can also be achived in an analog circuit but the limiting
>factors are cost and stability when in large scale serial production.
>The hunt for more performance with analog circuits halt when they have
>to add trim pots to achive control loop stability.

>When the controller is running in a computer it can also relatively easy
>be made adaptive. The control loop parameters will  then be adjusted, in
>run time, to compensate for deviation in the motor parameters due to
>production and age. Then we can have even higher loop gain and better
>servo performance. If this is done in digital servos of today i do not
>know.


Which sevoes actualy do that?  Or is that just speculation and
whishful thinking?  Where can I find block diagrams or transfer
functions suitable for simulating and analyzing model R/C servos,
gyros, and controlers?

Alan Jones

 
 
 

Digital Servo

Post by Pete » Sun, 28 Jan 2001 22:36:07


It has been a long time since I looked at the insides of a servo, but there
is a chip available which does all ofthe functions needed from receiving a
1-2mS pulse to driving the output transistors. Start with that.

--

Peter Wales in Sunny Orlando

http://home.cfl.rr.com/pwales


Quote:
> On Fri, 26 Jan 2001 02:15:51 +0100, Kjell Arne Gai


> >> The main advantage of digitals is the utter repeatability of the
position.
> >> Because they can have high power just off the set point they can move
to the
> >> set point easily and the accuracy of the the converter gives them
accuracy
> >> in position.

> >This all comes from a improved controller.

> >To really hit the set point you will need integration in the control
> >loop. To increase the power when pushed/pulled off the set point you
> >will need more control loop gain.

> >You can increase the control loop gain without loosing stability if the
> >sampling frequancy is increased. You can also have more gain in certain
> >regions by adding complexity to transfer functions in the controller.

> >Most of this can also be achived in an analog circuit but the limiting
> >factors are cost and stability when in large scale serial production.
> >The hunt for more performance with analog circuits halt when they have
> >to add trim pots to achive control loop stability.

> >When the controller is running in a computer it can also relatively easy
> >be made adaptive. The control loop parameters will  then be adjusted, in
> >run time, to compensate for deviation in the motor parameters due to
> >production and age. Then we can have even higher loop gain and better
> >servo performance. If this is done in digital servos of today i do not
> >know.

> Which sevoes actualy do that?  Or is that just speculation and
> whishful thinking?  Where can I find block diagrams or transfer
> functions suitable for simulating and analyzing model R/C servos,
> gyros, and controlers?

> Alan Jones

 
 
 

Digital Servo

Post by Kjell Arne Ga » Mon, 29 Jan 2001 11:46:39


Quote:

> On Fri, 26 Jan 2001 02:15:51 +0100, Kjell Arne Gai

> >When the controller is running in a computer it can also relatively easy
> >be made adaptive. The control loop parameters will  then be adjusted, in
> >run time, to compensate for deviation in the motor parameters due to
> >production and age. Then we can have even higher loop gain and better
> >servo performance. If this is done in digital servos of today i do not
> >know.

> Which sevoes actualy do that?  Or is that just speculation and
> whishful thinking?  Where can I find block diagrams or transfer
> functions suitable for simulating and analyzing model R/C servos,
> gyros, and controlers?

As i said, i do not know how advanced the controllers in the current
digital servos are. If the contoller is implemented in software it will
be very difficult to find out anything about it. I do not think that any
manufactuerer will publish their source code. If you want to take a look
at the documentation and source code for some more general motor and
servo controllers go to some DSP manufacturer like TI. With
microcontrollers of today you can run a very advanced control loop for a
motor as simple as a brushed DC.

With analog servos it is easier because they are often designed with an
integrated circuit where the data sheet is available from the chip
manufacturer.

Take a look at MC33030 from Motorola, M51660L from Mitsubishi or the
good old NE544.

--