That sounds actually right. Even they explained in the instructions that they used the lightest wood possible and commented that the servos were pretty cheap. I think I may get a UNO starter kit and start playing with various motors.
I have a question though. I know I'll need at least 12 servos, the uno says is has 14added i/o ports, 6 of which are pulse. Will it support 12 servos as well as some sensors?
Uno and Duemilanove boards have 20 I/O lines, and can handle up to 12 servos using the standard servo library, if I'm not mistaken. However, it is a royal pain to connect that many servos to those bds, since servos are 3-pin devices and the usual Arduino bds have 1-pin I/O headers.
If you really want to do a walker with Arduino you might hunt around for bds that can interface servos easier. Others might suggest what is available - I designed my own pcb for my projects. Another possibility is to buy a dedicated servo controller bd where you can plug in the servos directly, then you will have at least 18 I/O lines available on the Arduino.
Of course, doing all of this adds "weight", which is the bane of walker design.
For reference, I am currently working on a quadruped [Poco] that uses Lynxmotion pan-tilt brackets for the legs. These are fairly heavy, and I suggest using something lighter in weight. I started with HS322 Hitec servos, and it walked initially but after adding a host of sensors, pan'n'tilt pod, etc, the servos were no longer powerful enough for walking any longer. So I am currently changing over to MG946R servos.
Both kinds of servos are available for about $10/ea. The 946s came from Hobby King in china. Of course, heavy legs, pan'n'tilt pods, sensors, etc, all add weight so then it needs heavier batteries. Vicious cycle, building walkers.
The below small HTX900 servo has been used in small walking bots with some success. These servos seem to have stronger gearing than similar priced/sized micro servos. Note that with small servos the leg length may need to be shorter to accomidate the lower servo torque.
Those servos are more powerful than the previous mini/micros, but I calculate 2500gm-cm/(28.3gm/oz * 2.54cm/in) = 35 oz-in of torque. That's still pretty darn weak, less than the 44 oz-in of "standard" servos, so you're still not gonna run hog wild using them.
That's still pretty darn weak, less than the 44 oz-in of "standard" servos, so you're still not gonna run hog wild using them.
It is expected that a builder will do the appropriate research if building a DIY bot. Even "standard" servos are inadequate in some versions of "walking" bots.
Haha, man, I really did pick a pain of a project. I think I'm going to get a starter kit with the UNO and play around with motors along with some more research on various servos. Like what most people on here are saying, this is going to be a trial and error project. Good thing this is just for fun, haha!
I am going to assume that the Arduino motor shield won't be the solution. It only supports 2dc motors? Anyone know of a good one that like someone said, supports 18 servos and additional sensors. I'm thinking of two ideas: One would have a motor in the middle that rotates 360 with a distance sensor that follows which direction the robot is moving to prevent it from running into things. The other idea would stick 3 or 4 distance sensors on the thing, but the latter seems to add more weight and take up more I/O ports, so I'm leaning to the first.
I also have an old VEX robot from when I was younger I was thinking of savaging parts from, and it has a camera attached to it with a two pin chord to its microcontroller. It's actually pretty light, but from what I can see, it looks like the camera records and sends the signal wirelessly to a receiver instead of transferring the signal back through the wires, meaning the wires are simply power. Good news for me, I'm thinking of putting that on there as well and using the micro controller just for power. Anyone know how to make the signal wLan instead of a RF signal though?
Building walkers is all about making tradeoffs between weight, servo power, battery power, and cost of multiple servos. All fun. After selecting something long these lines, you might try adding up all the weights and projecting about how long the legs can be.
I am going to assume that the Arduino motor shield won't be the solution. It only supports 2dc motors? Anyone know of a good one that like someone said, supports 18 servos and additional sensors.
Motor shields support D/C motors not R/C servos.
I'm glancing around at servos and the solar servos caught my eye:
Doing some research and math behind the torque, they roughly can hold between 1.5 - 2lbs (680g - 900g) using 3in segments for the legs. That sounds to me like it should hold the microcontroller and a couple sensors with a lightweight container and still move fairly quick. Anyone have experience with them?
Batteries are the heavy part. Have you given any thought to how you will power so many servos ?
Oh yea, didn’t think about that. It draws power from the batteries through the Arduino board right? Hmm… Time for more research on how much power I’ll need!
No, check the two links in my signature for a small intro to driving servos with arduino and powering them both.
It draws power from the batteries through the Arduino board right?
Wrong! You should first strudy the servos that the commercial vendors include in their hexapod kits to see what works. May save you some $$$ and fustration in the long run.
(you asked many posts back, and I believe you did not get an answer:) if you plan on working with servos, you do not need a motor shield. a sensor shield might be something fun to have, to make prototyping easier. you may want to check out robotoshop.com they have a bunch of "robot" arduinos, which have lots of little features which you will find usefull.
something like this: http://www.robotshop.com/productinfo.aspx?pc=RB-Dfr-36&lang=en-US (there are other similar products as well, you may just want to search around.) I also found this: http://www.robotshop.com/productinfo.aspx?pc=RB-Dfr-152&lang=en-US (which I think could be spot on what you need)
Thanks, this should give me a lot to digest. I have a starter kit with a couple basic gizmos and a book to learn the basics with on the way, and I'll work from there with the UNO. I am curious though as to why a sensor shield would work better with a servo than a motor shield?
I am curious though as to why a sensor shield would work better with a servo than a motor shield?
I think what was being said was "you don't need a motor shield to drive a servo (you can do that straight from the Arduino board), but buying a sensor board might be more useful to you because it would allow you to interface to sensors". In truth, you may not need either.
the reason I suggested the sensor shield is that you can directly hook up both servos and sensors to it.
you do not need this board. Its a luxury item. However, I find it saves me a lot of hassle and I enjoy using it for prototyping.
(I mean something like this: http://arduino-info.wikispaces.com/SensorShield)
That shield was actually the one I was looking at getting. I ended up starting a new post, just since I'm working on a different topic now with servos and shields instead of just main microcontrollers.
I'm looking at the legs now, and from a couple replies in my other post, and as fkeel stated, I don't need the shield at all, just eases the process a little. My only concern is powering the servos and I understand the arduino won't do a very good job doing that, much less with 12 servos. I would have to re-route the power to the battery instead of the arduino, would I have to do that with the shield as well? If so, I may not get a sheild if I'm going to have to re-wire everything anyway. Second question is on the servos I picked out. I picked them out because they have the fastest speed for the highest torque and the lightest weight that I can tell. Just curious if anyone had an opinion on these or had another servo they preferred using.
Here are the supplies I'm looking at: Shield: http://emartee.com/product/41953/Sensor%20Shield%20V5.0%20%20Arduino%20Compatible Servo Possibilities: http://www.hobbypartz.com/37gexidised11.html http://www.hobbypartz.com/33p-solarservo-d222.html Pan and Tilt Kit: http://www.robotshop.com/lynxmotion-pan-and-tilt-kit-aluminium-1.html
This is all in the second post as well, but I'll be monitoring both.