Market Research

[peterwaller]

I have been contemplating going to the next RoboOlympics in the US next June and to help pay for it (Now I am a Pensioner) I am considering producing some antweight controller boards based on the ones I produce for my own use at the moment. I have thought of three options and would like to hear which people think would be best :

1 Just a dual motor control board with mixing
2 As 1 but with built in receiver
3 As 2 but with two sanyo look alike motor gearboxes

I estimate the final market price at 30, 50, 75 pounds respectively.

I have a prototype of option 3 which weighs 36 gms and consists of a board 55mm x 55mm including a 4 mm border with 8 x 2.5mm holes for mounting the armour. This board would effectively be the chassis.
I have a layout as yet untried for option 1 which is 23mm x 36mm.

[EpicentrE]

I would say the Dual Motor Driver, simply because I prefer having more smaller components than less large ones (such as 2x Scorpion Mini's rather than a Sozbots, for example) as it makes it easier to find places to put things.

Edit: Pete, if you could put up rough CAD Drawings maybe of all 3 options with dimensions, it would help people understand. I'd change my vote to number 2 if the board with the RX in wasnt that much bigger than the one without.

[josh]

i would like to see the second one as i often get fed up of soldering all those reciever cables back on and would defenitely buy one if u made them

[BeligerAnt]

Pete

Option 1 gives the builder most flexibility so I would think it would be the biggest seller. At ?30 it's quite competitive, with the advantage of coming from a reliable source!
Option 2 could be viable if it has significant size/weight savings over the usual discrete solutions. But what crystals would it use? And are they readily available on a good range of freqs?
Option 3 is a neat solution, but as Scott says it is a very "fixed" solution so might have limited application. On the other hand it is a "complete" solution. Unfortunately at ?75 it might be a bit expensive for a beginner who is most likely to benefit from a "turnkey" solution.

I assume you have included a BEC in all the designs?

Personally I would be most likely to go for option 1 I think.

HTH

[leo-rcc]

For me, option 2 sounds like the best option, although i can see te reasoning behind the choice for option 1.

but having the receiver built in on the board, and provided there is still a channel left to control a 3d servo or ESC, would eleminate another point of failure, the link between receiver and controller.

just my 2 eurocents.

Prices sound very reasonable by the way.

[ewan]

I'd go for option 2, but only if you can really exploit the size and weight advantages of having an-inbuilt Rx (and crystal compatibility would certainly be an issue). Otherwise option 1 would be a great solution for the suggested price.

Mixing is a useful feature that some of the US controllers lack, that would be a good addition. I'm assuming failsafing, regen braking and the like would be covered as can be reasonably easily incorporated?

In how many speed/frequency "steps" to you aim to drive the mosfets of a dual motor board with?

I'd be interested to hear how you progress, and could offer to retail the units eventually if that would be something you're interested in!

[peterwaller]

Here is an image of the two sides of the option3 board.
.
On the left image you can see the motor gearboxes, receiver, power plug and third channel output connector.
The thing I like about option three which no one else seems to like is the lack of wires and weight.
Ewan the processor has 256 speed control in both directions but really anything of about 20 is over kill.
To sell the comercially I would need EMC testing which can cost 1000's so I would probably just sell to people I know.

[leo-rcc]

Is that also the case if the components you use have been tested individually before?

[EpicentrE]

Pete, if i'm totally honest, my biggest problem with option 3 is you could basically add 2 wheels and bolt a sheet of polycarb on it and have a simple robot - and I'll be the first to express my utter dislike at robot kits (see the millions of VDD kit-clones in the states, aswell as team inertia labs pre-made chassis', aswell as other things) providing parts together is one thing, providing everything done for you in my opinion not only leads to un-origional robots, but also its less likely to make new people stay for the long term, as they wont learn nearly as much while building as you would if you had to do everything yourself. Personally I would hate to see a swarm of robots with the chassis you just showed us from option 3 with just a servo stuck on top with a wedge on it or something, which is what these robot kits normally spawn. I realise I'm exagerating to a degree, but this is something I personally feel strongly about.

However, I'm just one person, and that's just my 2 cents.

[Simon Windisch]

I'd go for option 2 because it gives me more choices, there are three different 'sanyo copy motor' speeds out there.

I agree with Scott, but reading his post another way makes me think that option 3 would be the most popular, amongst new roboteers.

36g for option 3 sounds extraordinary. As I recall, a barello board weighs 11g, and a receiver weighs about 10 grams, and two sanyo copies weigh 9g each, which totals to 39g, plus a few grams for all the extra wires and leads.

I take it that option 2 would be option three but simply without the motors, in which case it wouldn't bo too hard to offer both.

Simon