Darwin OP CM730
Darwin OP CM730
by i-Bot
Fri Feb 18, 2011 2:08 pm
There is some more information on Darwin OP here:
http://darwin-op.springnote.com/pages/6861909
It especially answers some questions about thecost of CM730 which Gort raised in his blog:
http://mike-ibioloid.blogspot.com/2011/01/how-other-guys-can-get-darwin-op-part.html
The photos of the actual board suggest a price around $100 would be reasonable even allowing for the inclusion of the gyro and accelerometer.
Also interesting to see the slight divergence of the Robotis product from the originally posted code from Univ. of Pennsylvania. More inegrated to existing Robotis products like Roboplus and motions. Also appears to be all C++ and not using Lua.
Looks like Roboplus may be remote interface.
http://darwin-op.springnote.com/pages/7293299
http://darwin-op.springnote.com/pages/6861909
It especially answers some questions about thecost of CM730 which Gort raised in his blog:
http://mike-ibioloid.blogspot.com/2011/01/how-other-guys-can-get-darwin-op-part.html
The photos of the actual board suggest a price around $100 would be reasonable even allowing for the inclusion of the gyro and accelerometer.
Also interesting to see the slight divergence of the Robotis product from the originally posted code from Univ. of Pennsylvania. More inegrated to existing Robotis products like Roboplus and motions. Also appears to be all C++ and not using Lua.
Looks like Roboplus may be remote interface.
http://darwin-op.springnote.com/pages/7293299
by Dewey
Fri Feb 18, 2011 3:55 pm
The upcoming CM730 looks like a great controller!
I am a little confused however as the picture above shows all AX 3pin ttl connectors, no 4pin 485 Rx connectors.
Darwip op uses RX28m's which are 4pin!!!???
It does look like from the circuit that perhaps two versions may be made available as it appears the 3pin plugs could be swapped out to 4pins!
Oh well time will no doubt clarify things!
Dewey
I am a little confused however as the picture above shows all AX 3pin ttl connectors, no 4pin 485 Rx connectors.
Darwip op uses RX28m's which are 4pin!!!???
It does look like from the circuit that perhaps two versions may be made available as it appears the 3pin plugs could be swapped out to 4pins!
Oh well time will no doubt clarify things!
Dewey
by Tyberius
Fri Feb 18, 2011 4:38 pm
Dewey wrote:The upcoming CM730 looks like a great controller!
I am a little confused however as the picture above shows all AX 3pin ttl connectors, no 4pin 485 Rx connectors.
Darwip op uses RX28m's which are 4pin!!!???
It does look like from the circuit that perhaps two versions may be made available as it appears the 3pin plugs could be swapped out to 4pins!
Oh well time will no doubt clarify things!
Dewey
The RX-28M uses TTL, not RS-485. It also incorporates the new Dynamixel 2.0 protocol which is capable of 4.5mbps.
i-Bot: How do you figure $100? The CM-700 is $119 and doesn't have this many features. I'd wager $200+
by i-Bot
Fri Feb 18, 2011 7:28 pm
My price estimate is based on 3 times the 1000 up component cost. This is usually a fair guess for low volume and predominantly hardware products. The CM700 has associated firmware costs too, which may push that price up. This does raise questions on how the firmware in CM730 is sourced and funded, and also if that will be open source too.
There are some oddities in the specs related to the Dynamixel 2. All the Darwin documents show only current support for baud rates up to 1 or 2 Mbps, not the 4.5 of Dynamixel 2. Also it appears the link from main controller to sub controller is 2Mbps serial, and the board contains something looking like a FT232R. This is odd if the sub controller IC has native USB support.
There are some oddities in the specs related to the Dynamixel 2. All the Darwin documents show only current support for baud rates up to 1 or 2 Mbps, not the 4.5 of Dynamixel 2. Also it appears the link from main controller to sub controller is 2Mbps serial, and the board contains something looking like a FT232R. This is odd if the sub controller IC has native USB support.
Great!
by Gort
Sat Feb 19, 2011 2:56 am
<table><tr><td>
</td></tr><tr><td>From DARwIn-OP unveiling at the The 2010 IEEE-RAS International Conference on Humanoid Robots.</td></tr></table>
</td></tr><tr><td>From DARwIn-OP unveiling at the The 2010 IEEE-RAS International Conference on Humanoid Robots.</td></tr></table>
by Dewey
Sat Feb 19, 2011 9:53 am
Thanks Tyberius my mistake wrt the new RX-28M!
Its interesting though, the 4 wire 485 is more immune to noise than 3 wire ttl 232, yet Robotis clearly have shown/believe that they can quadruple the transmission speed and stay with 232 without transmission error problems!
Dewey
Its interesting though, the 4 wire 485 is more immune to noise than 3 wire ttl 232, yet Robotis clearly have shown/believe that they can quadruple the transmission speed and stay with 232 without transmission error problems!
Dewey
by i-Bot
Sun Mar 13, 2011 12:02 pm
As far as I know there is no proposal yet by Robotis to make available the Bioloid GP frames alone. Even the Bioloid GP itself is not available from there web store. I think the frame set would be rather expensive just to bling up a premium. To get the best from GP, you will need to upgrade the legs to AX18F servos, so the full GP kit is probably the best deal.
For the CM730, then again this is part of the Darwin-OP and no detail of seperate availability is given. The CM-730 is the subcontroller in the controller pair of the OP including the Fit PC. The pair togther are fairly large and heavy. In the OP they sit horizontally. This means this main/sub controller combo is more suited to the larger RX servo based OP frames.
The GP uses the CM510 controller.
For the CM730, then again this is part of the Darwin-OP and no detail of seperate availability is given. The CM-730 is the subcontroller in the controller pair of the OP including the Fit PC. The pair togther are fairly large and heavy. In the OP they sit horizontally. This means this main/sub controller combo is more suited to the larger RX servo based OP frames.
The GP uses the CM510 controller.