Getting a servo up and running...

358 views
Skip to first unread message

Niels Jakob Buch

unread,
May 26, 2017, 4:42:10 AM5/26/17
to BeagleBoard
I bought the SM 4303B servo motor (see https://www.banggood.com/Spring-SM-4303B-48V6V-5KG-Analog-SERVO-p-932400.html)

But trying out the rc_test_servo command with many different options, has not revealed any movements at all.

I have tried "rc_test_servos -v -s 1.5" as best guess but not sure though. I have checked that polarity is ok.

Cannot find documentation or tips/tricks anywhere to guide me.

Any hints appreciated.

 

Dennis Lee Bieber

unread,
May 26, 2017, 11:04:05 AM5/26/17
to beagl...@googlegroups.com
On Fri, 26 May 2017 01:42:10 -0700 (PDT), Niels Jakob Buch
<njb...@gmail.com> declaimed the following:

>I bought the SM 4303B servo motor
>(see https://www.banggood.com/Spring-SM-4303B-48V6V-5KG-Analog-SERVO-p-932400.html)
>
>But trying out the rc_test_servo command with many different options, has
>not revealed any movements at all.
>

You haven't stated which Beagle or cape is in use, nor to which pins
you have connected the servo leads. I'm guessing at the use of the
"Robotics Cape" on the basis of the documentation for rc_test_servos (
http://www.strawsondesign.com/content/manual-servos ).


>I have tried "rc_test_servos -v -s 1.5" as best guess but not sure though.
>I have checked that polarity is ok.

Have you tried specifying frequency? Possibly this servo has slow
response to pulses.

Have you even confirmed you have pulses? [An {analog} VOM should show a
non-zero voltage between signal lead and ground -- exact voltage would vary
with pulse width and frequency and sensitivity of the meter; a DVM probably
won't respond in time -- an oscilloscope would be the most precise as you'd
get an indication of both pulse width and voltage height... Even an LED
with suitable dropping resistor should vary in brightness as you sweep the
pulse width]


If you've got pulses, you may just have a bad servo. Forgive me, but
based on the specifications/price, I'm not sure I'd trust the quality of
that servo... The closest match (4.8V 3.9kg-cm 0.22 sec/60deg; 6.0V
5.2kg-cm 0.18 sec/60deg) from a quality company (Futaba) is 10 times the
cost -- $40. S9001/FUTM0075 Aircraft
http://www.futabarc.com/servos/analog.html

--
Wulfraed Dennis Lee Bieber AF6VN
wlf...@ix.netcom.com HTTP://wlfraed.home.netcom.com/

David Lechner

unread,
May 26, 2017, 12:32:38 PM5/26/17
to beagl...@googlegroups.com
Are you running as root?

sudo rc_test_servos -c 1 -v -s 1.5

...was working for me a while back, but I have not tried it with the latest software updates.
--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/beagleboard/9ea778ac-70b4-4357-8f46-5d4ab702cd52%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Niels Jakob Buch

unread,
May 26, 2017, 2:46:56 PM5/26/17
to BeagleBoard
Yes, I am running with sudo....

Niels Jakob Buch

unread,
May 26, 2017, 2:52:22 PM5/26/17
to BeagleBoard


On Friday, 26 May 2017 17:04:05 UTC+2, Dennis Lee Bieber wrote:
On Fri, 26 May 2017 01:42:10 -0700 (PDT), Niels Jakob Buch
<njb...@gmail.com> declaimed the following:

>I bought the SM 4303B servo motor
>(see https://www.banggood.com/Spring-SM-4303B-48V6V-5KG-Analog-SERVO-p-932400.html)
>
>But trying out the rc_test_servo command with many different options, has
>not revealed any movements at all.
>

        You haven't stated which Beagle or cape is in use, nor to which pins
you have connected the servo leads. I'm guessing at the use of the
"Robotics Cape" on the basis of the documentation for rc_test_servos (
http://www.strawsondesign.com/content/manual-servos ).


I am sorry..I am running on the Beaglebone Blue, and I have attached the servo to the plug farthest away from the sd-button, with the white wire closest to the CPU. 

uname -a: Linux beaglebone 4.4.54-ti-r93 #1 SMP Fri Mar 17 13:08:22 UTC 2017 armv7l GNU/Linux
 

>I have tried "rc_test_servos -v -s 1.5" as best guess but not sure though.
>I have checked that polarity is ok.

        Have you tried specifying frequency? Possibly this servo has slow
response to pulses.


I have tried frequencies from 5 to 500 with no change...
 
        Have you even confirmed you have pulses? [An {analog} VOM should show a
non-zero voltage between signal lead and ground -- exact voltage would vary
with pulse width and frequency and sensitivity of the meter; a DVM probably
won't respond in time -- an oscilloscope would be the most precise as you'd
get an indication of both pulse width and voltage height... Even an LED
with suitable dropping resistor should vary in brightness as you sweep the
pulse width]


I will check this in the lab later with a scope...
 

        If you've got pulses, you may just have a bad servo. Forgive me, but
based on the specifications/price, I'm not sure I'd trust the quality of
that servo... The closest match (4.8V 3.9kg-cm 0.22 sec/60deg; 6.0V
5.2kg-cm 0.18 sec/60deg) from a quality company (Futaba) is 10 times the
cost -- $40. S9001/FUTM0075 Aircraft
http://www.futabarc.com/servos/analog.html


You are right, I might have bought this one too cheap... :(

Niels Jakob Buch

unread,
May 26, 2017, 4:04:13 PM5/26/17
to BeagleBoard
So, I got the scope attached to the servo-port on the Beaglebone Blue, and realized that the 6V rail might no give the voltage required.

The pulse is fine shaped to the 50 Hz specified with a voltage between 3 and 4 volts, but the + row in the servo pins does hit around 0 volts.


The output when running the command confirms that the 6V rail is switched on using software.


Obvious next step?

Dennis Lee Bieber

unread,
May 26, 2017, 5:32:29 PM5/26/17
to beagl...@googlegroups.com
On Fri, 26 May 2017 13:04:13 -0700 (PDT), Niels Jakob Buch
<njb...@gmail.com> declaimed the following:

>So, I got the scope attached to the servo-port on the Beaglebone Blue, and
>realized that the 6V rail might no give the voltage required.
>
>The pulse is fine shaped to the 50 Hz specified with a voltage between 3
>and 4 volts, but the + row in the servo pins does hit around 0 volts.
>
><https://lh3.googleusercontent.com/-lg9BcrZcTJc/WSiJUE_HkcI/AAAAAAAAE1w/h3OmM4mNk_oT_ehpwQMfmyR07EftgxLewCLcB/s1600/IMG_9459.JPG>
>
>
>The output when running the command confirms that the 6V rail is switched
>on using software.
>
>
>Obvious next step?
>
Well... Tracing backwards to where the 6V rail is supplied power might
be next -- as long as the scope is available and you can try each 6V point
all the way back to the regulator.

Does the Blue use the same control pin as the Robotics Cape? If it uses
a different control pin, then the software you are using to test with may
not be actually enabling the 6V.

Niels Jakob Buch

unread,
May 26, 2017, 5:44:41 PM5/26/17
to BeagleBoard
Just took a look at the schematics... https://github.com/beagleboard/beaglebone-blue/blob/master/BeagleBone_Blue_sch.pdf

Seems that the servos are powered by the batteries, which might make sense to supply enough amps.

Anyway, is there a way to power the servos with another source?

Niels Jakob Buch

unread,
May 29, 2017, 6:13:34 AM5/29/17
to BeagleBoard
Would it be possible to wire a Powerout with 6V from a power-socket directly to the servo, and not connect the middle pin to the servo-connector?

Aaron

unread,
May 29, 2017, 11:01:44 AM5/29/17
to beagl...@googlegroups.com
You can power the servo from an external 6V source, but you'd have to connect the ground of the power source to your BBBlue or the servo will not work right.  ie.  The ground going to the servo, still needs to connect to the ground pin on the BBBlue servo output header.




--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups "BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beagleboard+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/beagleboard/c98b47a1-dc19-4fe4-93c4-b136747e2635%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages