Looking for advice

[ta_erog]

Looking for advice (my first post . . . .)

I am looking for a servo based robot (KHR-1/2, Bioloid Comprehensive, Hitec Robonova-1 Kits)

What I am looking for:
A Kit that is decent to start with (I want to teach others that may see this with a steeper learning curve then I) But also Upgradable/configurable enough to keep me happy. Eventually I want to move away from movement macros and simplistic programming to direct wireless control with a closed control loop (this will also include various additional sensors) to match my sims. (FYI I am very into the software and “animation” aspects since I come from the simulation side – I have less experience in rewiring electronics though I can work on PC boards and am not bad craft wise – My day job is midrange Server QA. I handle and modify hardware all of the time but I do not make anything new)

My Take:
The KHR and robonovoa seem to have been out longer and have more (English) support. They also seem to be much easier to add sensors and other add-ons. BUT are limited to a humanoid form unless I can get to a machining shop and make custom brackets and such.
The Bioloid seems very configuration and looks like it has good beginner software . . But because it is so new it seems unsure that are its true limitations, and what mods can be done to add needed sensors and direct control (even some of the wireless mods seem to be allot of work for what should be a built in feature.)

Questions:
-Are the KHR and robonovoa humanoids easily modable into alternate configurations, Is the software flexible? The brackets look rather well made – are they difficult to build with out a full workshop? I know people have done it here.

-Are the KHR and robonovoa humanoids easier to control via a closed control loop? Is ther epremade software (shared) or is everyone reinventing?

-How strong are the Bioloids plastic pieces? I see videos of these falling all over the place they seem rather tough. Can you get more brackets?

-What is the relative use time of a full humanoid moving around (average exercise time to recharge?) I assume all last different periods.

-The Serial setup of the Bioloid seems like a good direction . . do you think this is a trend or a one time deal? It does also seem like this can also be a hindrance with adding sensors, as I see people working on it. Will there be more servos that follow this format?

-In general what do people like and dislike about there kits?


I think I am leaning toward the Bioloid since I expect by the time I get up-to-speed people here would have figured out the electronic mods (which I can't do) to the limitations I see. But I thought to ask a few questions since you are all the experts

PS, Also I do 3D animating. Modifying .x files for new configurations people want should be quite do able.

any responce will help thanks.

[ta_erog]

No responses? allot of looks . .
Really any help would be good. . . or even telling me I am totally off base in my assumptions or right on the money would help.

Anyone?

[JonHylands]

I'll start by stating I am a hardcore Bioloid fan, so you know my bias.

The Bioloid brackets are very strong. You can buy an entirely new frame set for $100 USD.

I personally think the serial bus on the Bioloid is definitely the way to go - I can't imaging why anyone would want to run separate wires to each servo.

I'm building a Wifi interface which will give a full 1.0 Mbps wireless interface to the bus, including all the AX-12 servos and any sensors on the bus. I'm also building some custom bus sensors, including a 6-axis IMU and a foot pressure sensor board. Those will plug into the bus like everything else.

- Jon

[ta_erog]

Ok that is good information, Thanks allot.

I will be Very interested in your projects becouse that is where I eventally want to go.

I do like the serial bus as I am more used to networking . . we use I2c in our servers for enviromentals.


Where are some places you buy it (from the USA)? I am looking for a place with tracking and a good rep and such. I know that you can get it here but I would perfer to get it in the USA if possable. (no offence robosavvy)

[Robo1]

Hi there.

I own a KHR-2 and I find that the brackets aren't that good. I would probably buy a Bioloid after seeing footage of it. I don't like the fact that you can't program it.

Bren[/code]

[JonHylands]

If you're looking for a US supplier, I have bought individual servos from CrustCrawler, and had good after sales support from them.

http://www.crustcrawler.com/products/bi ... hp?prod=59

- Jon

[Pev]

Let me see if I can help a little bit. Firstly I have no experience of the Kondo products so can't say a lot about them. As for the Hitec and Robotis products these are me thoughts:

1. Firstly I believe they are very different products, the Hitec Robonova 1 is a complete biped robot ready to be hacked, but also ready to run out of the box. The interfacing is more standard with Digital I/O lines, I2C, serial comms etc. The controller board is pretty good but there is the issue of the RoboBasic limitations and speed which can be a pain.

2. The Bioloid Kit is more a robotics experimenters kit. The unit will make a Biped but can also be made into whatever you can imagine (Hexapods, quadpods, wheeled robots etc). It comes with some basic but workable sensors (distance, sound etc) and the Robotis philosophy is that robotics is about autonoumous action not remote control. The down sides are that the interface to the controller is a serial interface at halfduplex and so interfacing additional items to the system is more complex but that is being addressed by a number of groups here and I still see the serial bus approach as the way to go. As for software I like the fact the that the CM5 controller holds the motion sequences and they are called by the programs you write but the programing environment leaves a little to be desired if you are used to writing in VB or C as it is very graphically based. An excellent tool however for use in schools for introducing younger pupils to the concepts of robotics.

I really like both products but you have to realise that they are very different and fullfil different requriements. As I have been playing around in robots for 10 years I have enough of an interest in both areas (biped, and experimenter) to see value in both approaches.

Conclusion, if you want a pure biped I'd recommend the Hitech RN1, if you want an experimentation and education platform I'd go Robotis.

To answer a couple of the specific questions:

1. The plastic Bioloid brackets are excellent and have caused me no issues or concerns.

2. Running time wise I get just over an hour with my Bioloid depending on what I am doing (heavy movement testing drops this dramatically but the CM-5 can be run from the mains power supply when programming etc). On my RoboNova I see between 20 and 30 minutes and have invested in a second battery to extend the time. The Robonova cannot be run from its charger when programming (Hitec state this will cause issues and damage the controller board)

Hope this helps

Pev

[ta_erog]

Thank you all for the responses,
I guess I was most worried about longevity and flexibility of the hardware.
I had a (now realized irrational) fear that the Biolod was more Toy (plastic and all) and would have unseen limitations compared to the other kits. (The more I read and research this seems not to be the case.) Also I wanted to keep my options open because each of these kits seems interesting.
I think the point on “robotics experimenter’s kit” vs “biped” has pushed me to decide what I am getting . I am much more interested in changing configurations and experimenting with different motive forms and gates then a pure biped.
I am going to see about ordering one today then, Thanks all who helped . . . I will surely have more questions once I get my unit.

[ta_erog]

I picked the Bioloid . . . I am waiting for delivery . . . Yay!

Any good reads here or other forums on first "steps" while I wait? experienced do's and don'ts

(I can see the excitement growing and the want to jump to "z" 10 min after I open the box . . blowing any good measured approach)