Playstation controller interface
Playstation controller interface
by StuartL
Fri Nov 07, 2008 1:02 pm
The PSX controller board I've been hyping up is here. We have one of the prototypes built and tested and it's looking good.
We're hoping to have the demo boards built in time for the robotics convention on Sat 29th November in Birmingham UK so if you can make this you should see them in action.
We WILL be offering these test/demo boards to the beta users and resellers at a very low cost, this will include Trossen, Robosavvy, Tribotix and anyone else I can think of who is actively selling Bioloid kit or actively developing robots. In return for participating in the beta programme we expect people to actively assist in resolving any remaining bugs (we're quietly confident here).
The demo boards will be restricted in a minorly inconvenient way to discourage resale and uniquely identifiable so we can track which one people are using when bug finding.
Stealing from the wireless-control thread the feature list not currently implemented but planned for the demo release:
1) Programmable double-click detection on-board.
2) SYNC_READ support (will include the AX-12 firmware) to allow the polling of many devices with one bus transmission.
3) Loss of signal detection.
Please do feel free to request features in this thread and hopefully we can get them in.
We're hoping to have the demo boards built in time for the robotics convention on Sat 29th November in Birmingham UK so if you can make this you should see them in action.
We WILL be offering these test/demo boards to the beta users and resellers at a very low cost, this will include Trossen, Robosavvy, Tribotix and anyone else I can think of who is actively selling Bioloid kit or actively developing robots. In return for participating in the beta programme we expect people to actively assist in resolving any remaining bugs (we're quietly confident here).
The demo boards will be restricted in a minorly inconvenient way to discourage resale and uniquely identifiable so we can track which one people are using when bug finding.
Stealing from the wireless-control thread the feature list not currently implemented but planned for the demo release:
1) Programmable double-click detection on-board.
2) SYNC_READ support (will include the AX-12 firmware) to allow the polling of many devices with one bus transmission.
3) Loss of signal detection.
Please do feel free to request features in this thread and hopefully we can get them in.
by StuartL
Fri Feb 06, 2009 12:49 pm
As you may have spotted there has been a delay in these units. The delay is in software, fortunately, so it's fixable. Matt has been working his bottom off the last few weeks identifying what's going on and progress has been made.
The symptoms are that the controller (and, indirectly, the servo firmware Matt also wrote) occasionally fail to respond to packets. This isn't usually a problem but we're striving for a release that we believe will work first time without flaws.
Watch this space
The symptoms are that the controller (and, indirectly, the servo firmware Matt also wrote) occasionally fail to respond to packets. This isn't usually a problem but we're striving for a release that we believe will work first time without flaws.
Watch this space
by RandomMatt
Tue Mar 10, 2009 9:41 pm
Progress report....
The hardware is built.
I believe the firmware code is all done (for version 1). I am now trying to finish the bootloader so that it is easy to upgrade to version 2(*).
The manual / datasheet isn't even started.
--
More information...
The firmware is very close to that in the servos(**) - so it should be tested. I have, on several occasions, run the robot for 1-2 hours at a time without a single error on the dynamixel bus (which is a good sign).
Currently the code just reflects the state of the controller as memory locations (updated at just under 60Hz). I was thinking that a nice feature to have would be double click detection, but I haven't sorted out (in my head) how to present that information in the memory locations. Any (sensible) suggestions on how to expose this will be helpful.
The manual should be easy to write (i imagine it'll take an evening).
* obviously, I intend not to make any errors at all. But I am not a fool, therefore I recognise that one needs to be able to fix them in the field!
** I have been running with my own servo firmware since oct-vember last year - you'll be amazed just how much better(***) the servos can be with a bit of effort.
*** I am tempted to give the servo firmware away with the psx-b boards, so you won't have to trust me on this one.
The hardware is built.
I believe the firmware code is all done (for version 1). I am now trying to finish the bootloader so that it is easy to upgrade to version 2(*).
The manual / datasheet isn't even started.
--
More information...
The firmware is very close to that in the servos(**) - so it should be tested. I have, on several occasions, run the robot for 1-2 hours at a time without a single error on the dynamixel bus (which is a good sign).
Currently the code just reflects the state of the controller as memory locations (updated at just under 60Hz). I was thinking that a nice feature to have would be double click detection, but I haven't sorted out (in my head) how to present that information in the memory locations. Any (sensible) suggestions on how to expose this will be helpful.
The manual should be easy to write (i imagine it'll take an evening).
* obviously, I intend not to make any errors at all. But I am not a fool, therefore I recognise that one needs to be able to fix them in the field!
** I have been running with my own servo firmware since oct-vember last year - you'll be amazed just how much better(***) the servos can be with a bit of effort.
*** I am tempted to give the servo firmware away with the psx-b boards, so you won't have to trust me on this one.
by BillB
Wed Mar 11, 2009 4:52 pm
I am really looking forward to your PS2 controller solution. I have tried using the Bioloid Zigbee but the reception is absolutely terrible. I am sure that your PS2 controller solution will be far better.
I am still using your Bioloid C library and it works wonderfully. I am sure that if your PS2 controller code is as high quality as your C Library then there should be very few problems.
I am still using your Bioloid C library and it works wonderfully. I am sure that if your PS2 controller code is as high quality as your C Library then there should be very few problems.
by StuartL
Sat May 16, 2009 8:53 am
These boards are now production ready. Version two of the firmware is complete and tested and we have a few samples we're going to sell at cost to the resellers.
New for version two of the firmware we have hot-plug of controllers (handy if you don't want to power down the robot) and 50/60Hz switch for PAL/NTSC controllers which will only operate at their native refresh rate.
I shall contact Limor off-topic regarding reselling these boards.
New for version two of the firmware we have hot-plug of controllers (handy if you don't want to power down the robot) and 50/60Hz switch for PAL/NTSC controllers which will only operate at their native refresh rate.
I shall contact Limor off-topic regarding reselling these boards.
by StuartL
Mon Sep 07, 2009 8:23 pm
Limor has made the decision that these boards are not a viable proposition for Robosavvy. On that basis Matt and I have decided to sell them direct and having checked with Limor that Robosavvy are not interested this post does not contravene the acceptable use policy of the forum.
Pedro's thoughts on the demo board can be found here:
http://robosavvy.com/forum/viewtopic.php?t=3710
Unlike the demo board we sent to Pedro we have decided that there will be two versions of the board offered.
The BO (board only) board will be a populated and programmed PCB with Dynamixel connectors, instructions, firmware and solder pads for the connector or Playstation interface of your choosing. We have had some success with the Matcatz or Cobra wireless controllers.
The BC (board and controller) board will be as above but include the Madcatz wireless Playstation 2 controller and wireless receiver pre-assembled and soldered. Although this controller does not feature vibration feedback it does seem to have the best analogue response of the controllers we've tested.
The PDF documentation and latest firmware can be found here:
http://www.braincell.cx/bioloid/dist/psx-latest.tgz
Prices are:
PSX-BO (preassembled and soldered board with Dynamixel connector) : £40 + P&P
PSX-BC (preassembled and soldered board, Dynamixel connector and preassembled/tested Madcatz controller): £65 + P&P
For purchasing information or bulk discounts please contact my first name at braincell dot cx.
Pedro's thoughts on the demo board can be found here:
http://robosavvy.com/forum/viewtopic.php?t=3710
Unlike the demo board we sent to Pedro we have decided that there will be two versions of the board offered.
The BO (board only) board will be a populated and programmed PCB with Dynamixel connectors, instructions, firmware and solder pads for the connector or Playstation interface of your choosing. We have had some success with the Matcatz or Cobra wireless controllers.
The BC (board and controller) board will be as above but include the Madcatz wireless Playstation 2 controller and wireless receiver pre-assembled and soldered. Although this controller does not feature vibration feedback it does seem to have the best analogue response of the controllers we've tested.
The PDF documentation and latest firmware can be found here:
http://www.braincell.cx/bioloid/dist/psx-latest.tgz
Prices are:
PSX-BO (preassembled and soldered board with Dynamixel connector) : £40 + P&P
PSX-BC (preassembled and soldered board, Dynamixel connector and preassembled/tested Madcatz controller): £65 + P&P
For purchasing information or bulk discounts please contact my first name at braincell dot cx.