PDA

View Full Version : Propellor USB Servo Controller



OneCoolCat
03-31-2010, 07:19 AM
Why does this servo controller have such a hard time when I use servo cable extensions?· If I connect the servo directly to the board it works fine but if I use an extension cable the servo twitches erratically.· The old discontinued servo controller would work fine with or without the extension cable.

Any ideas how I can get around this issue?· Thanks :)

Eugene

Post Edited (OneCoolCat) : 3/31/2010 12:24:56 AM GMT

NWCCTV
03-31-2010, 01:10 PM
Sounds like possibly a bad extension cable. Do you have another to test with?

MitchT
03-31-2010, 05:30 PM
I've recently encountered this as well.

Did some testing (All with Propeller USB Servo controller)
(All Extensions were tested with at least 4 different PWM extension cables; some futaba some hitec standard)
(All configurations tested with at least 5 futaba s3305's and 2 hitec HS-5485)
Futaba S3305 No extension - Works
Futaba S3305 With short extension - Works with a little chatter (Some servos better than others; no consistency)
Futaba S3305 With medium extension works with significant chatter; barely controllable
Futaba S3305 with Long extension - wont even try to turn on; as if it werent even getting a pulse.

Hitec HS-5485HB Works in ALL cases with no chatter.


Testing with Astro Flight Inc. Servo Tester

Futaba S3305 works with ALL extension lengths in all cases with no chatter
HS-5485HB works with ALL extension lengths in all cases with no chatter


I hope this provides some insight to someone who knows more about this board than i do :)

Its worth noting that the Futaba S3305's are standard analog servos and the HS-5485HB's are from HiTec's "Standard Digital" line.

Additionally, all testing was done using USB access via the PSCI application provided by parallax; not serial from a separate micro controller.

Some testing was done briefly using my own application with the baud mode at 38400 (also via USB) but no significant differences showed up so I didnt bother testing extensively.

Hopefully later today I will get a chance to test via Microcontroller -> Serial just to get all the info up here in hopes of finding a solution :)

Post Edited (MitchT) : 3/31/2010 10:41:03 AM GMT

Chris Savage
03-31-2010, 09:32 PM
The issue has to do with meeting the proper load on the outputs. When adding the extension cables this is thrown off causing the on-board translators to get confused about direction (since they're auto-switching). We're somewhat limited in options. We're currently looking at a few. One option is to remove the bi-directional capability from the I/O, however a few customers have made use of this feature. Whatever we do, the revisions should happen pretty quickly.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage

Parallax Engineering
·

OneCoolCat
03-31-2010, 11:56 PM
Another observation...· The Futaba S5301, a giant scale servo, appears to work fine with the Propellor servo controller, even with a long extension.· Maybe that will help·with the·design of the revision.

Post Edited (OneCoolCat) : 3/31/2010 6:21:45 PM GMT

hover1
04-01-2010, 01:11 AM
Read this thread and see if it gives you a temporary fix.

·http://forums.parallax.com/showthread.php?p=862005

Jim

MitchT
04-01-2010, 01:30 AM
OneCoolCat said...
Another observation... The Futaba S5301, a giant scale servo, appears to work fine with the Propellor servo controller, even with a long extension. Maybe that will help with the design of the revision.


Are you working on the same project i am? :P

We also tested the S5301's (forgot to post) and found no issues at any length.


As for bidirectional autoswitching I REALLY want to use this with HiTec HMI protocol to read servo position as position capture - and we are slowly cycling out to HiTec stuff anyway; BUUUUT i'm hoping this temporary solution works out too :)


Thanks for fast replies!

Chris Savage
04-01-2010, 02:13 AM
I guess I should attempt to obtain some of these servos we don't have for testing. I need to find out what the differences are.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage

Parallax Engineering
·

MitchT
04-01-2010, 03:01 AM
Chris Savage (Parallax) said...
I guess I should attempt to obtain some of these servos we don't have for testing. I need to find out what the differences are.


Where are you physically? We have a ton. PM me :)

(Bay area california here)

hover1
04-01-2010, 03:16 AM
Chris,

This is to one that gave me the most trouble:
JR DS8711 Digital Servo

And this fixed it:



SkalTura,
The Propeller USB Servo Controller cannot drive loads less with than 5 kΩ of impedance. I have never seen servos that do not have high impedance inputs, but the oscilloscope traces you recorded are typical of a TXB0108 driving a low impedance load. If you do not need bidirectional communication, a series resistor of 5 kΩ or more will stop the ringing.

-- David Carrier
Parallax Inc.

Jim

OneCoolCat
04-01-2010, 11:06 AM
Another observation...· I have two S5301 giant scale servos plugged into the Propellor servo controller and they both work fine with the extension cables.· However, when I add one of the middle sized servos supplied by Parallax (also with an extension cable), all three go into the twitchy seizures.· Hope that helps.

Mitch-T:· I'm working on a robotic arm to handle test samples.··http://forums.parallax.com/images/smilies/cool.gif· ·That S5301 is the strongest servo I've seen.

MitchT
04-01-2010, 05:20 PM
OneCoolCat said...
Another observation... I have two S5301 giant scale servos plugged into the Propellor servo controller and they both work fine with the extension cables. However, when I add one of the middle sized servos supplied by Parallax (also with an extension cable), all three go into the twitchy seizures. Hope that helps.


Mitch-T: I'm working on a robotic arm to handle test samples. http://forums.parallax.com/images/smilies/cool.gif That S5301 is the strongest servo I've seen.



Ah :)

Not quite heh

www.xdtech.net/p/gantry.jpg (http://www.xdtech.net/p/gantry.jpg)
www.morav.net (http://www.morav.net)


Anyway - I've recently been swapping out the S5301's (the big servos we use in the legs on these guys) with the HSR-5990TG's - I find that the fact that they are digital simply gives more stability and less chatter under high load. However the S5301's if supported correctly or tensioned for use in an arm (with a spring or bungee) would probably do better because they are far stiffer and basically dont need power to hold certain amounts of weight at strange angles. Not so good for smooth moving robot though :)