PDA

View Full Version : Bioloid Boards Again



JonHylands
10-21-2009, 07:28 PM
Hi Everyone,

I finally have some time again, so I'm going to start building Bioloid boards again.

If anyone has any requests for any new board ideas, please post them here.

One new board I'm thinking about is an AX/RX bridge. That will allow you to drive a mix of AX-12 and RX-series servos from an AX bus. This board will allow you to drive the bus from either direction (RX master or AX master), and will allow you to power the "other" bus separately.

Any other ideas?

- Jon

Suicidal.Banana
10-22-2009, 03:11 AM
Heys,
Well, i recall seeing a board on some japanese page, wich had ports for all robotis servos, but in effect that caused the board to be kind of crammed, maybe you could think of some form of connection cable where the 'end user' can just plugin the servo ports needed (if thats in anyway possible) would be win imo.

Atm im covered though, using an Arbotix, and i cant afford better servos yet either, i will definitely need something like this in the future though (for a v2 or maybe v3 of the bot im making now)

Think about the cable port thing, i see it like theres a large row of pins on one side of the board, and u can put 6(?) pin plugs on them, and u get these cables with it that have sayd plug on one end, and 1 (or maybe more, 2, 3?) servo sockets on the other end.

If you could; how about make it possible to use the same row of pins for sensors/sensor cables?

Just my 2 cents, i love to see what you come up with, i love all projects i've seen from you, so i have high hopes :)

darkback2
10-22-2009, 10:18 AM
could you post a link to your list of existing boards John? I am really really really interested in a couple of things.

DB

JonHylands
10-22-2009, 10:33 AM
I haven't updated it yet, but here they are:

http://www.huvrobotics.com/shop (http://www.huvrobotics.com/shop/)

I'm not going to do the IMU-01 anymore, since I'm building my own custom sensor board. I will probably reduce the price on the IMU a bit, since the sensors themselves are cheaper now.

- Jon

darkback2
10-22-2009, 11:17 AM
Ok...A few things I would really appreciate...

One would be an interface from the serial bus for a standard servo

Another would be a board that takes the place of the seral cable...as in a wireless interface ala zigbee but that works the same as the current cord plug in...

so maybe a xbee - bioloid board.

DB

JonHylands
10-22-2009, 11:34 AM
darkback2,

The first one is relatively easy - the only real trick is providing enough power at a low voltage (6.0 volts) for the servo(s) from the bus power (9-12 volts). I'm a big fan of switching regulators, so I'll have to do a bit of research to figure out what to use. Assuming a single servo is going to pull less than an amp, we could use the Dimension Engineering regulators for that.

For a single servo, use this: http://www.dimensionengineering.com/DE-SWADJ.htm

For multiple servos, or a single high-current servo, use this:
http://www.dimensionengineering.com/DE-SWADJ3.htm

For the second board, are you talking about a board that goes inline on the bus, and transmits the bus wirelessly in both directions? So, for example, you could use this device with my USB Bus board, or a USB2Dynamixel, and have a wireless robot that is PC controlled.

Wouldn't it be easier to use something like an arbotiX (http://www.trossenrobotics.com/p/arbotix-robot-controller.aspx)?

- Jon

darkback2
10-22-2009, 11:46 AM
Wouldn't it be easier to use something like an arbotiX (http://www.trossenrobotics.com/p/arbotix-robot-controller.aspx)?

- Jon

Yeah...but I was hoping to stick with the stock bioloid software. More like just making the bioloid wireless and PC controlled in a meaningful way...probably just being lazy is all.

DB

JonHylands
10-22-2009, 11:50 AM
Okay, so you just want a simple way to make the bus be wireless, with something that you just plug in and it works. That seems reasonable...

- Jon

lnxfergy
10-22-2009, 02:20 PM
Yeah...but I was hoping to stick with the stock bioloid software. More like just making the bioloid wireless and PC controlled in a meaningful way...probably just being lazy is all.

DB

So then... the Bioloid zigbee connector, sold by Robotis/Trossen?

-Fergs

lnxfergy
10-22-2009, 02:22 PM
Okay, so you just want a simple way to make the bus be wireless, with something that you just plug in and it works. That seems reasonable...

- Jon

I think the issue is throughput. DB mentioned an XBEE, but they'll top out at 250k. A wireless adapter that can do 1MBps reliably is probably not gonna be cheap -- hence the idea of either an arbotiX or a zigbee-enabled CM5/510 -- the micro on board that handles some stuff up front is kinda important given the speed you need to control 18+ servos with syncwrites at 30hz (for smooth motion).

-Fergs

lnxfergy
10-22-2009, 02:24 PM
darkback2,

The first one is relatively easy - the only real trick is providing enough power at a low voltage (6.0 volts) for the servo(s) from the bus power (9-12 volts). I'm a big fan of switching regulators, so I'll have to do a bit of research to figure out what to use. Assuming a single servo is going to pull less than an amp, we could use the Dimension Engineering regulators for that.

For a single servo, use this: http://www.dimensionengineering.com/DE-SWADJ.htm

For multiple servos, or a single high-current servo, use this:
http://www.dimensionengineering.com/DE-SWADJ3.htm

- Jon

Jon, I've been thinking about a similar idea actually, just haven't found the time -- but mainly for attaching micro servos to the bioloid bus. Sorta like an SSC-32 (maybe only like 8 servos though) that takes a bioloid bus connection. The lack of serial micro servos is a real pain I find.

-Fergs

JonHylands
10-22-2009, 02:29 PM
Jon, I've been thinking about a similar idea actually, just haven't found the time -- but mainly for attaching micro servos to the bioloid bus. Sorta like an SSC-32 (maybe only like 8 servos though) that takes a bioloid bus connection. The lack of serial micro servos is a real pain I find.

-Fergs

The question is, how much current are 8 micro-servos going to draw? I suppose if you're going to draw more than 3 amps, you could provide your own power supply for those servos.

Any ideas?

- Jon

billyzelsnack
10-22-2009, 02:34 PM
How about the smallest, simplest, and cheapest USB Bus interface possible.

USB is not on-board and works via a FTDI cable that you can purchase separately.
Single AX-12 port.
User provides servo power themselves.
Board is powered by the cable, but you can cut or bend the V+ pin if you want to supply it yourself.
Solder points for servo V+, board V+, and ground.

Is there any disadvantage to doing the tie the TX and RX pins together trick ( maybe weird states on power-up> ) so it could use the most minimal part count?

JonHylands
10-22-2009, 02:45 PM
You can't tie the Tx and Rx pins together when you're using the FTDI chips. The buffer circuitry (tristate and inverter) costs about $1, so its not really a big problem. The two most expensive components on my USB Bus interface board are the FT232 itself (about $5), and the power switch (also about $5).

Using a power connector like I provide isn't optimal if you're running off batteries, but everyone who has a Bioloid kit already has a 12 volt 5 amp power supply that plugs right in.

There's no reason to power the logic part of the board, since it is already powered from the USB connector.

In reality, there just isn't enough volume to justify really low-margin stuff. If I could sell 1000 of them, it would be reasonable to build a board that had a low profit margin, and sell it for $20 or $30. But the reality is, in the almost two years I was selling this stuff, I sold less than 100 USB Bus interface boards.

People who can hack together FT232 cables to build interfaces don't need to buy it - the people who buy these things want something that works right out of the box, with no effort on their part.

- Jon

JonHylands
10-22-2009, 02:49 PM
I think the issue is throughput. DB mentioned an XBEE, but they'll top out at 250k. A wireless adapter that can do 1MBps reliably is probably not gonna be cheap -- hence the idea of either an arbotiX or a zigbee-enabled CM5/510 -- the micro on board that handles some stuff up front is kinda important given the speed you need to control 18+ servos with syncwrites at 30hz (for smooth motion).

-Fergs

Yeah, there isn't a cheap/easy way to do 1 Mbps wirelessly. The only protocol that can handle that reliably is wifi, and there aren't any cheap 1 Mbps wifi modules that can talk to a microcontroller at that speed (or at least, there weren't any when I was looking a few years ago).

- Jon

lnxfergy
10-22-2009, 03:02 PM
Yeah, there isn't a cheap/easy way to do 1 Mbps wirelessly. The only protocol that can handle that reliably is wifi, and there aren't any cheap 1 Mbps wifi modules that can talk to a microcontroller at that speed (or at least, there weren't any when I was looking a few years ago).

- Jon

Yeah, my thougts exactly. Cheapest I've seen is the $70 model from Lantronix (which has it's own issues)

-Fergs

lnxfergy
10-22-2009, 03:07 PM
The question is, how much current are 8 micro-servos going to draw? I suppose if you're going to draw more than 3 amps, you could provide your own power supply for those servos.

Any ideas?

- Jon

Same thoughts here. I was actually looking at the exact same 25W regulator for my setup. 8 maxed out mini servos is probably over 3A, estimated peak current seams to be 500-750mA depending on model (based on web research, no idea how reliable those numbers are).

However, for what I'm trying to do, 3A is probably enough, as my mini servos aren't that maxed out (mainly, as extra DOF for actuating things like eyebrows, lips, eyelids, or the tips of a lightweight tail).

-Fergs

Adrenalynn
10-22-2009, 05:28 PM
The WiFly GSX does serial at 2.5mbit/sec + if their docs are to be believed, and is < $70 in onesie-twosie. No one seems to be rating them above 921kbit/sec though, but according to rovingnetworks (oem), they're 2.7mbit/sec...

JonHylands
10-22-2009, 05:36 PM
The user manual says, on page 12, that you can set the baud rate:

set uart baud <rate> set the UART baud rate. Valid settings are {2400, 4800, 9600, 19200, 38400, 57600, 115200, 230400, 460800, 921600}.
Example : “set u b 9600” sets the baud rate to 9600 baud.
NOTE: the RS232 interface on the RN-134 does not work above 230400.

I don't know what this means, but it seems like anything above 921K is theoretical at this point...

- Jon

Adrenalynn
10-22-2009, 05:39 PM
Somebody's gotta pull the plug and test it... ;)

JonHylands
10-22-2009, 05:58 PM
Yeah, I'll do that. I suspect a board with one of these, and a bioloid bus interface, able to run at 1.0 Mbps, would be a decent seller. I expect the cost to be around $175-$200 for such a board - anyone interested?

- Jon

darkback2
10-22-2009, 06:46 PM
Given how much I've already spent trying to do this...I'm in.

PM me with the details.

DB

JonHylands
10-22-2009, 06:55 PM
I've ordered an eval board, should get it on Monday. I'll play with it, and if it can indeed do 1.0 Mbps, I should be able to have a board done within 2-3 weeks and ready to sell...

- Jon

nagmier
10-22-2009, 07:46 PM
Man thats impressive Jon! Concept to "product" in less than a month you are a monster! you must be a workaholic :D (All the better for the robot builders of the world :D

Adrenalynn
10-22-2009, 08:10 PM
Jon - I [we] appreciate your taking the flyer on the prototype! Nice!

JonHylands
10-23-2009, 09:03 AM
So, here's my first board layout:

http://www.huvrobotics.com/Bioloid-Wifi-01-small.png

Larger Version (http://www.huvrobotics.com/Bioloid-Wifi-01.png)

I don't know if the LEDs are driven from the ground side or the power side, so I've got a surface mount jumper set up for them to do either.

I'm using an ATmega164, because its the cheapest and smallest ATmega I could find with two hardware serial ports...

The big cap is a 100 uF, 25 V electrolytic cap. The regulator is a 1 amp regulator, which unfortunately is going to burn a bit of heat (3.3 volts). I may eventually change to a switching regulator for it. All the parts are on the top side of the board. Board dimensions are 1.5" x 2.25", with a 0.3" overhang for the wifi module antenna on one side.

- Jon

JonHylands
10-23-2009, 12:20 PM
So, as I was out doing some errands at lunch today, I realized that I forgot something - you (the end-customer) have to be able to program the Wifi module to attach to your router, so I added an FT232 and a USB plug, with a jumper to switch between USB mode and Bioloid mode.

http://www.huvrobotics.com/Bioloid-Wifi-02-small.png

Larger Version (http://www.huvrobotics.com/Bioloid-Wifi-02.png)

The FT232 and the USB plug add 1/2" to the overall length, so the board is now 2.75" x 1.5".

- Jon

darkback2
10-23-2009, 12:43 PM
Sorry for seaming dense...

Is this a wifi bridge between a computer and a CM-5? or a replacement for the cm-5?

DB

Suicidal.Banana
10-23-2009, 01:09 PM
Was my idea that bad? :( (seeing how no1 even commented on it)

JonHylands
10-23-2009, 01:14 PM
Sorry for seaming dense...

Is this a wifi bridge between a computer and a CM-5? or a replacement for the cm-5?

DB

This is a wifi interface to the bus, so it replaces the CM-5 with your PC, and provides a wireless full-speed link to the bus.

I think I need to add one more thing to this board - realistically you will need to power the bus from this board, so I will add a terminal block (and a fuse) for battery hookup.

- Jon

JonHylands
10-23-2009, 01:14 PM
Was my idea that bad? :( (seeing how no1 even commented on it)

Honestly, I have no idea what you were asking...

- Jon

lnxfergy
10-23-2009, 01:21 PM
This is a wifi interface to the bus, so it replaces the CM-5 with your PC, and provides a wireless full-speed link to the bus.

I think I need to add one more thing to this board - realistically you will need to power the bus from this board, so I will add a terminal block (and a fuse) for battery hookup.

- Jon

Best way to describe it: It's like a wireless USBDynamixel (although it does serial over ethernet).

Jon, you using the lantronix module? (don't recognize the footprint)

-Fergs

JonHylands
10-23-2009, 01:28 PM
Best way to describe it: It's like a wireless USBDynamixel (although it does serial over ethernet).

Jon, you using the lantronix module? (don't recognize the footprint)

-Fergs

I'm using the module Adrenalynn recommended - the Roving Networks RN-131G.

- Jon

darkback2
10-23-2009, 01:43 PM
Ok...so if I'm understanding correcly you could send it servo positions, but not pages to play?

DB

JonHylands
10-23-2009, 01:46 PM
That's correct. You would no longer use the CM-5 with this setup.

- Jon

Adrenalynn
10-23-2009, 02:17 PM
Just to note: "suggested" more than "recommended" ;) It's a complete flyer you're taking Jon - no one seems certain what the peak baudrate really is, and I've touched one once that was probably already dead at a show table, but never actually had power to one where I could play with it.

It is an awesome notion though! And nice quick work on the board layout!

JonHylands
10-23-2009, 02:48 PM
Just to note: "suggested" more than "recommended" ;) It's a complete flyer you're taking Jon - no one seems certain what the peak baudrate really is, and I've touched one once that was probably already dead at a show table, but never actually had power to one where I could play with it.

It is an awesome notion though! And nice quick work on the board layout!

The data sheet for the RN-131G clearly says 1.0 Mbps on the UART, so I'm hoping (otherwise I wouldn't bother doing all this work). According to the tracking information, I should get the module on Monday, so I'll wire it up on a breadboard and test the baud rate. Once I've confirmed it, I'll order the protoboard and the real module, and see how it all works...

Here's the latest board, with a battery terminal, power switch, and a fuse. It has grown once again to 1.5" x 3"...

http://www.huvrobotics.com/Bioloid-Wifi-03-small.png

Larger Version (http://www.huvrobotics.com/Bioloid-Wifi-03.png)

- Jon

Adam
10-23-2009, 06:08 PM
I'm in for an IMU when they are available again. Lower cost would be good. :D

Any ETA?

Is Trossen still the distributor as stated on your page? I don't see the products listed.

Thanks for creating these products for the community.

JonHylands
10-23-2009, 07:13 PM
IMUs should be available in about two weeks. I haven't set the price yet...

I'm not sure if Trossen is still going to distribute my stuff - its really up to them.

- Jon

Gorbag
10-24-2009, 10:31 AM
Jon: Glad to see you're producing boards again.

What I'd like to see is kind of a generic robotis protocol to D/A A/D board, e.g. with n of each. That would make arbitrary sensor interface simpler. Some ports might just be binary (turn on/off leds or detect switch type sensors). More boards gets you more addresses on the bus to mux more sensors?

Kind of like the phidgets 8/8/8 board, but for the bioloid bus.

Another option would be a bioloid bus to I2C interface or similar. Looking for ways to leverage the bus :-).

JonHylands
10-24-2009, 11:55 AM
Jon: Glad to see you're producing boards again.

What I'd like to see is kind of a generic robotis protocol to D/A A/D board, e.g. with n of each. That would make arbitrary sensor interface simpler. Some ports might just be binary (turn on/off leds or detect switch type sensors). More boards gets you more addresses on the bus to mux more sensors?

Kind of like the phidgets 8/8/8 board, but for the bioloid bus.

Another option would be a bioloid bus to I2C interface or similar. Looking for ways to leverage the bus :-).

My General I/O board has 2 PWM outs, 4 digital I/Os, and 6 A/D inputs. Is that along the lines of what you're looking for?

Interfacing the bus to a protocol like I2C is a little harder, mainly because of the control table metaphor that is used with bus devices. If you have any ideas how to do this in a reasonable way, while still maintaining compatibility with existing bus devices, I'm all ears...

- Jon

Suicidal.Banana
10-24-2009, 12:17 PM
Honestly, I have no idea what you were asking...

- Jon

Oh my bad, well heres a little drawing, bear with my shitty drawing skills

http://forums.trossenrobotics.com/gallery/files/3/5/9/9/servobar.png

JonHylands
10-24-2009, 02:46 PM
I don't see a reason for something like that - you are wasting vast amounts of real estate on the board in pins that will never get used. You don't need or want centralized connections for that - the whole point of a serial bus is to distribute to intelligent devices, which are small specialized things.

- Jon

Suicidal.Banana
10-24-2009, 04:27 PM
okay :P just a idea that popped up hehe

lnxfergy
10-24-2009, 05:09 PM
Ok, a whole bunch of thoughts I've been brewing over the past few days.

RE: Wifi Board
How would this really be *that* useful? With something like the CM5 or arbotiX, poses are stored on the processor, and most importantly the interpolation is done on the processor. Is the idea that this board will have an interpolation engine onboard, or that it will just be a raw pass through? If it's just a pass through, won't that lead to jerky movements: since both the PC and wifi connection are inherently non-deterministic?

RE: Sensors on the Bus
I'm a fan of keeping sensors off the bus. As far as I've found, the individual AX-12+ servos cannot do interpolation for themselves, so you need to send out sync writes at 20-50hz for smooth interpolation. At 18 servos, that's quite a bit of bus bandwidth already. An ADC read on an AVR is quite a bit easier than a read across the Bioloid bus.

RE S.B's Servo Bars
I'm gonna have to agree with Jon here, that generally, a chain is more effective. However, you generally want a chain for each limb -- that's where the little 6-way servo connector board that comes with beginner/comprehensive kits comes in.

-Fergs

Adam
10-24-2009, 05:31 PM
RE: Sensors on the Bus
I'm a fan of keeping sensors off the bus. As far as I've found, the individual AX-12+ servos cannot do interpolation for themselves, so you need to send out sync writes at 20-50hz for smooth interpolation. At 18 servos, that's quite a bit of bus bandwidth already. An ADC read on an AVR is quite a bit easier than a read across the Bioloid bus.


No reason why you can't have a separate bus on another COM port for sensors. IMHO the Bioloid serial bus w/ table data metaphor is strong enough to justify such a solution if someone wanted to go that route.

JonHylands
10-24-2009, 06:46 PM
Ok, a whole bunch of thoughts I've been brewing over the past few days.

RE: Wifi Board
How would this really be *that* useful? With something like the CM5 or arbotiX, poses are stored on the processor, and most importantly the interpolation is done on the processor. Is the idea that this board will have an interpolation engine onboard, or that it will just be a raw pass through? If it's just a pass through, won't that lead to jerky movements: since both the PC and wifi connection are inherently non-deterministic?

The CM-5/arbotix is useful for doing very simple motion. However, to get into anything really complicated, you need a lot more CPU and a lot more memory on the robot, and then you need bigger servos, etc.

This wifi module is intended to give you a low-latency, full speed wireless interface to the bus from a PC. Yes, the PC is not deterministic. Wifi is non-deterministic also. But they're not bad, in general - if you're careful you can keep the latency well below 10 ms.

- Jon

mannyr7
10-25-2009, 04:57 AM
IMHO the Bioloid serial bus w/ table data metaphor is strong enough to justify such a solution if someone wanted to go that route.

If you have completely off-loaded all processing, where is the table data stored? In my experience, using bluetooth to SSC-32, better not to send complex movements, rather than sending simple calls to EEPROM stored sequences on the bot. This sounds like a serial pass-through to the bus, but maybe I'm reading the intent wrong?

Adam
10-25-2009, 09:17 AM
If you have completely off-loaded all processing, where is the table data stored? In my experience, using bluetooth to SSC-32, better not to send complex movements, rather than sending simple calls to EEPROM stored sequences on the bot. This sounds like a serial pass-through to the bus, but maybe I'm reading the intent wrong?

Sorry if I was a little vague. We're talking about Jon's boards, so I made some assumptions. Oh, and I'm coming from the perspective of everything being onboard, but it could work either way.

When I say table data, I'm talking about the Dynamixel Control Table, the structure of which is emulated by the HUV Robotics IO-01 Bioloid I/O Board (which I own) for anything that you can plug into it. This metaphor makes it easy for the programmer to integrate a lot of different types of sensors or other devices through a single API. I'm a skilled programmer, but new to robotics, so this appeals to me.

JonHylands
10-26-2009, 11:44 AM
So, I'm ready to test this thing (wifly), just waiting for fedex to stop by with the part. They are supposed to deliver it today, so here's hoping.

http://forums.trossenrobotics.com/gallery/files/1/3/2/0/bioloid-wifi-test-01_thumb.jpg (http://forums.trossenrobotics.com/gallery/showimage.php?i=1878&c=3)

Here's a picture of my test harness setup - I'm using a 3.3 volt FT232 board, and I'll test and make sure I can get a 1.0 mbps connection to the wifly. The voltage regulator at the top right is a Dimension Engineering AnyVolt Mini, set to output 3.3 volts (they don't make this one anymore, but I have a few in my parts bin).

- Jon

JonHylands
10-26-2009, 06:51 PM
So, after playing around with the wifi module for a while, I spent some time on the phone with the guy from Roving Networks. The module, as it stands today, will only do about 500 kbps. They are working hard on increasing that to 1.0 mbps, and he said they expect to reach that in 2-3 weeks.

The latency may be a problem, but right now I'm having difficultly just getting the thing connected to my router and getting DHCP working.

In any case, it means there won't be a Bioloid wifi board for at least a month or two, assuming they come through on their promise on the UART speed, and the latency is tolerable.

- Jon

Gorbag
10-26-2009, 09:30 PM
My General I/O board has 2 PWM outs, 4 digital I/Os, and 6 A/D inputs. Is that along the lines of what you're looking for?

Interfacing the bus to a protocol like I2C is a little harder, mainly because of the control table metaphor that is used with bus devices. If you have any ideas how to do this in a reasonable way, while still maintaining compatibility with existing bus devices, I'm all ears...

- Jon

Jon,

That sounds serviceable. I didn't see it on huvrobotics so wasn't aware of it.

Control tables and I2C: I think it depends on the specifics of the I2C device, but my interest was sensors. For sensors, I was just thinking that the table would include a matrix of values of each of the I2C devices attached, e.g. representing a sense value specific to the sensor from the I2C packet. For configurable sensors, you might be able to write to the specific cell a value which then gets transmitted as an I2C packet to that device. If you write to multiple table cells it would look like a Group message. Wouldn't work that well for long messages but I suspect sensors are read mostly after some possible initial configuration so the table of sensor values would seem to fit the metaphor. And like with many of the robotis tables, you can specify an offset and value for reading or writing? That would let you focus on a specific or small number of sensors. No need to support all 112 possible addresses of course, and it may make sense to make each entry multiple bytes (at least 2).

Anyway, it was a thought for something that might prove useful; it may be better to just use a processor capable of addressing both USB (to talk to the AX12s) and the I2C independently.

JonHylands
10-26-2009, 09:50 PM
I think another way to do I2C would be to have the last location in the control table be "special". You would specify that location as the start address of the bytes you were passing. Instead of storing those bytes in the table, the device would treat those bytes as an embedded I2C command, complete with the slave address, number of bytes to expect back, etc. I'm not 100&#37; sure it would work, but it might...

- Jon

JonHylands
11-16-2009, 08:51 AM
So, my "new" wifi module (from ConnectOne) has passed both tests - I can hook it up to my router and get an IP address, and I can talk to it over the serial port (using an FT232) at 1 Mbps.

I'm going to start laying out a new board for this, but it looks like the full speed wifi board is going to become a reality.

- Jon

DresnerRobotics
11-16-2009, 01:15 PM
BTW- The USB Board, IO Board, and Foot Pressure Sensor Board (http://blog.trossenrobotics.com/2009/11/16/huv-robotics-bioloid-accessories-are-back/) are now once again available for sale. =)

nagmier
11-16-2009, 02:13 PM
Nice Tybs/Jon Great to see some Bioloid accessories back! now I jest need to get my grubby paws on a comp kit :D

siempre.aprendiendo
11-17-2009, 05:15 PM
In any case, it means there won't be a Bioloid wifi board for at least a month or two, assuming they come through on their promise on the UART speed, and the latency is tolerable.

- Jon




So, my "new" wifi module (from ConnectOne) has passed both tests - I can hook it up to my router and get an IP address, and I can talk to it over the serial port (using an FT232) at 1 Mbps.

I'm going to start laying out a new board for this, but it looks like the full speed wifi board is going to become a reality.

- Jon

Wow. A wifi board for Bioloid at 1 Mbps would be amazing :robotsurprised:

JonHylands
02-26-2010, 12:43 PM
Well, it has taken a great deal of time, mainly because I had a bunch of power issues, and some timing issues.

In any case, here it is - the first production Bioloid wifi board:

1802

I still have to do a bit of software work - I want to add a SYNC_READ command, so you can get back values from the servos in one network read.

These should go on sale within a couple weeks. Price will be $200 USD. You get everything you see in the picture (plus a spare fuse). This gives you a full 1.0 mbps wireless connection to your Bioloid bus robot. Included on this board is a power-in screw terminal block, a power switch, a fuse (same as the one on my USB board), an ATmega164, a pair of bioloid bus sockets, a switching 3.3 volt regulator, the wifi mofule, and an FT232 with a USB plug (mini-B) for configuring your wifi module so it can talk to your local network. The wifi module supports WEP, WPA, and WPA2. I personally use WPA2, and it works great.

Also on the board are a pair of jumpers that allow you to talk to the wifi module from either the bus, or from the FT232/USB.

- Jon