PDA

View Full Version : [Interesting] RX-28M with TTL interface & 12V



RobotAtlas
11-29-2010, 08:28 PM
In this link RX 28M servos are mentioned:
http://downloads.sourceforge.net/project/darwinop/Overall/DARwIn-OP_Specification_Ver1.0.1.pdf?r=http%3A%2F%2Fsourc eforge.net%2Fprojects%2Fdarwinop%2Ffiles%2F&ts=1291083835&use_mirror=surfnet

Here are the specs. Did anybody hear about availability?


RX-28M
Holding Torque 24kgf.cm @ 12V
Speed 45RPM @ No Load
Motor Maxon Coreless DC Motor
Position Sensor Magnetic Potentiometer
Gear Ratio 1 : 193
Resolution 0.29

°

Size 35.5mm x 50.8mm x 41.8mm

Weight 72g
Network Interface TTL / Up To 4.5MBPS
Feedback Position / Speed / Temperature / Alarm

lnxfergy
11-29-2010, 08:36 PM
The current RX-28 is spec'd at 28.3kg-cm at 12V -- so that's nothing new. Weight is also the same. I imagine the "TTL" part is probably in error -- I can't imagine they would make a non rs-485 RX series servo.

-Fergs

iBot
11-30-2010, 04:44 AM
The more interesting features of this servo are that it is the first RX with magnetic potentiometer, hopefully more reliable and giving true 360 degree operation.
Also the 4.5M interface speed suggest maybe a processor change in the servo, maybe to the 72MHz Master clock (24MHz or 36Mhz CPU clock) 32 bit Cortex M3. This would be a good move especially if the new servo firmware were open sourced as part of the Darwin-OC program.

DresnerRobotics
11-30-2010, 10:51 AM
The more interesting features of this servo are that it is the first RX with magnetic potentiometer, hopefully more reliable and giving true 360 degree operation.
Also the 4.5M interface speed suggest maybe a processor change in the servo, maybe to the 72MHz Master clock (24MHz or 36Mhz CPU clock) 32 bit Cortex M3. This would be a good move especially if the new servo firmware were open sourced as part of the Darwin-OC program.

Since this information is public, I can confirm that as far as I know this is all correct (including the use of the Cortex). It is TTL, not RS-485.

'Magnetic Encoder' is what I was told, semantics really.

It also uses the same motor, so I believe these can be run at the 16.5v as well for full torque (not confirmed).

RobotAtlas
11-30-2010, 02:26 PM
32-bit processor in _each_ servo?

DresnerRobotics
11-30-2010, 03:54 PM
32-bit processor in _each_ servo?

I'm fairly certain that is correct. I couldn't get a 100% straight answer because I forgot to clarify with Robotis' engineers when we were at their HQ, but that was the impression I got from the presentation.

tician
12-02-2010, 02:30 AM
ARM is not just for mobile phones and PDAs. The cortex-M is intended for the microcontroller market and really is not that much more expensive for the bump in computing power and pin count. Going with a microcontroller capable of 32-bit math at 20MHz to 120MHz with built in motor controller for a few dollars more than an 8-bit microcontroller maxing at 20MHz does not really seem like an unreasonable leap to me.

Of course, I generally consider a prohibition against full-time professional politicians and lobbyists to be perfectly reasonable, so take my input as you may.

tician
05-16-2011, 11:22 AM
Now called the MX-28 (http://support.robotis.com/en/techsupport_eng.htm#product/dynamixel/rx_series/mx-28.htm).