Posted: Thu Dec 15, 2011 1:43 pm Post subject: New Robobuilder SAM Servo soon to be Available on RoboSavvy
After much development and QA Robobuilder is finally releasing their new line of Servos (SAM series) with torques ranging from 3kgf.cm up to 170kfg.cm with Potentiometer, Magnetic Encoder and Optical Encoder (depending on Model)
They have the advantage of offering a Full Duplex TTL bus (which makes them easier to work with compared to the Half Duplex Dynamixel bus) and the bus is designed to operate at 1.5Mbps.
They also offer RS485 half duplex (and some models seem to support both although I don't really know how...)
The servos are Daisy Chained with a Serial protocol just like Robotis and most models servos expose (externally) an Analog port to connect a sensor.
Quoting Robobuilder, this was the information we got:
"SAM Servo Bus Speed
=> Maximum : 2 Mbps, Normally 1.5 Mbps."
They've also mentioned "SAM manual (English version) will be ready by the end of this month." so in depth information is still limited.
I will try and ask your question (although we deal with the Sales person not Support so I'm not sure if I'll be able to get the answer before the release of the manual).
Additionally, with regards to the joints, what we were told si that currently the only joint that is being made available is the one in the pictures I will post shortly.
This means that many of the joints on the brochure are not yet confirmed in terms of availability.
(as a side note this makes little sense to me: to advertise the design of the joints without being sure if they want to produce them or not but then again there are significant cultural differences so we just have to respect that).
These resemble the wCK modules used in Robobuilder a lot. They have however much less torque (and slightly higher price) offering improved bus speeds.
Both SAM 5 and SAM 3 were designed to build the hand/finger structures on their TITAN robot and are noticeably smaller than current wCK servos and significantly smaller than AX-12.
This is a miniature servo (same size as SAM-5; probably different gear reduction). With only 3kgf.cm was designed to build "finger" articulations although its application is not limited to this.
It has the advantage of small size and reduced weight and retains the same high end features of other servos (same High Speed bus and Protocol) with the possibility of being daisy chained with all other (higher torque) servos.
There are some unique advantages to these servos compared to Dynamixel:
- The bus is Full Duplex and by default Operates at 1.5Mbps
A full duplex bus makes integration with Microprocessors a LOT easier than Dynamixel.
Dynamixel, being Half Duplex, means you need to Short RX and TX on the processor and write code to filter out want you send or otherwise add hardware (Tri State buffers)
- Wider range of Servos, from 3kgf up to 170kgf including minuature servo models: the 3kgf and 5kgf versions are interesting solutions for fingers or compact Pan&Tilt and other applications where space and weight are at a premium.
They can be mixed and matched on the same bus with all other servo models up to 170kfg.cm
- Full Metal Gears on all servos from 5kgf.cm on.
The only exception is SAM-3 that mixes and matches Metal with Plastic.
- Finally, I believe the ratio of price per kfg.cm will also be a factor to consider.
We've received pricing for these servos already (although we won't be disclosing it yet) and we can say they're slightly less expensive than the equivalent Dynamixel model (ie the price per kgf.cm is more competitive on SAM servos).
Robotis has also given us a heads up that pricing for Dynamixel should increase in 2012 which should make this difference even more expressive.
- The only downside I can anticipate on SAM servos is the lack of Magnetic Encoders on the lower end models.
We've discussed this with Robobuilder without great response from them (we're getting used to this from Korean companies)
Only real world experience and applications will tell if this has a significant impact in both adoption and performance.
There is no doubt that Dynamixels are the standard in this market segment and that their quality and reliability are undisputed.
We've seen attempts from other manufacturers to challenge the dominance of Dynamixels with no success so far.
Although we know and expect Dynamixels to continue being a great choice, we have hopes that these new servos from Robobuilder present a viable alternative to Robotis.
Competition is always nice for the end customer: performance improves and prices come down
I see the bus speed is 1.5Mbps. Could you check if this is 1.5M exactly, and are any lower baud rates related to sensible values (not 921.6K and 460.8K like wCK).
I've received some new information from Robobuilder regarding the most common bus speeds being supported.
- Maximum baud: 2 Mbps
- Typical baud: 1.5 Mbps
Other supported Baud Rates:
- 1 Mbps (like Dynamixel)
- 115 200 (ie the same used now by default on current wCKs).
These 4 servo baud rates are directly supported on their interface board for RQ-TITAN (ie the equivalent tot he CM-730 board on Darwin).
I'm not completely sure if these bauds are preciselly 1Mbps = 1 000 000 bps or if they are a rounding of a different number (we've seen this "rounding" before on wCKs).
We will need to wait for the Servo Specification before being able to completely understand the capabilities of the servos.
The issue with bus speed - as you very well mentioned - is more about deficiency than speed itself.
It would be interesting to compare bus efficiency on MX-28 at 3Mbps and SAM series a 1.5Mbps (their recommended speed) as this directly affects the maximum frame rate on a closed loop control system.
We don't have an oscilloscope here but maybe Limor can have someone at the University have a look at this once we receive the first units of SAM servos.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum