Making custom sensors: What would be the demand?

[inxt-generation]

I've been recently delving into the world of MicroControllers, and I've got a couple ideas for NXT and LEGO accessories. BTW, Please don't steal them! I'm just running them by anyone here to see if they would be interested.

Idea #1: A automatic pneumatic system. It would be like a one piece block that you plugged into your power, your pump motor, and an air pipe for reading the pressure. It could use a simple button and LED system to set the air pressure you desire. You could program it to keep the air pressure in between a certain pressure. i.e. between 40 and 45 psi, 50 and 60, 10 and 50, etc. It would allow much finer control than you get from a mechanical system, and doesn't use any ports on your NXT!

Idea #2: A serial adapter. Not much to it, a micro that receives I2C signals from the NXT, and translates them to serial commands, to control something like a large LCD display, for instance.

I've got some more, but don't want to give them away just yet. So, do any of you think that these are good ideas? That I could perhaps sell them on Ebay if I made them? I don't really have anything planned, just throwing out some ideas.

P.S. Like my new sig?

[mattallen37]

I have already made many NXT sensors. I like to post the code and any schematics online for other people (sort of open-source them I guess).

There already is an I2C to UART adapter specifically for the NXT; the HT IR Link. If you need higher speeds or don't want to deal with IR, and if you only need half duplex, you can use the RS485 that's already built into the NXT.

Something I have been thinking about a lot recently is a UART "mux" for the NXT. It would be similar to what you seem to be talking about, but far more advanced. It would ideally have 3-4 UART ports (full-duplex rx and tx), and interface to the NXT using I2C and/or RS485. I actually would have started on this project before now, but I have some other projects I'm working on ATM (other Arduino programs and libraries). I thought I would use an Arduino 2560, because it supports 4 UART ports (rx and tx). I also figured on 128+ byte buffers, to match or exceed that of the NXT's RS485 port, and individually selectable baud rates.

It seems that most people who would take advantage of a serial adapter as you described could just make their own (I've done it a time or two myself).

If you make a good product, I'm sure you could sell it on eBay. I have seen many home-brew mindstorms sensors sell on there. I really don't know how big the market would be though. You'd have to take it slow at first to see how things go.

You also need to remember that in order to sell mindstorms sensors/interfaces, people will want full support for the device in the programming language they are using. That means full drivers for at least 3 programming languages, and maybe more than 5.

In re. to you sig:

[h-g-t]

There are already I2C LCD driver chips available at low cost - http://www.ebay.co.uk/itm/I2C-LCD-Displ ... 48394e3436

The all-in-one pneumatic system sounds bit more promisig though.

[inxt-generation]

I have already made many NXT sensors. I like to post the code and any schematics online for other people (sort of open-source them I guess).

There already is an I2C to UART adapter specifically for the NXT; the HT IR Link. If you need higher speeds or don't want to deal with IR, and if you only need half duplex, you can use the RS485 that's already built into the NXT.

Something I have been thinking about a lot recently is a UART "mux" for the NXT. It would be similar to what you seem to be talking about, but far more advanced. It would ideally have 3-4 UART ports (full-duplex rx and tx), and interface to the NXT using I2C and/or RS485. I actually would have started on this project before now, but I have some other projects I'm working on ATM (other Arduino programs and libraries). I thought I would use an Arduino 2560, because it supports 4 UART ports (rx and tx). I also figured on 128+ byte buffers, to match or exceed that of the NXT's RS485 port, and individually selectable baud rates.

It seems that most people who would take advantage of a serial adapter as you described could just make their own (I've done it a time or two myself).

If you make a good product, I'm sure you could sell it on eBay. I have seen many home-brew mindstorms sensors sell on there. I really don't know how big the market would be though. You'd have to take it slow at first to see how things go.

You also need to remember that in order to sell mindstorms sensors/interfaces, people will want full support for the device in the programming language they are using. That means full drivers for at least 3 programming languages, and maybe more than 5.

In re. to you sig:

I'm not saying I wouldn't go OS, but it would probably easier for most people to just buy one, instead of having to source all the parts, and equipment to program the Micros.

Many people in my robotics club have 3D printers, and they were talking about making a prototyping service with them for the club. Wouldn't it be cool if I made custom-printed cases for the sensors?

I didn't know that the IR-Link could be used like that. Well, actually I guess I did, just didn't think of it. I guess my idea is already redundant then, isn't it?

Once I learn more about I2C, I could probably make drivers for RobotC, NXT-G, and NXC, which probably wouldn't be hard to learn now.

[mattallen37]

Being able to build (print) cases would be very cool!

An I2C to UART adapter that supports full-duplex and a nice assortment of BAUD rates could be very useful, I just don't know if there would be much of a market.

But I agree with h-g-t, an all-in-one pneumatic supply could be a big hit.

[inxt-generation]

Being able to build (print) cases would be very cool!

An I2C to UART adapter that supports full-duplex and a nice assortment of BAUD rates could be very useful, I just don't know if there would be much of a market.

But I agree with h-g-t, an all-in-one pneumatic supply could be a big hit.

The clubs meeting tomorrow. Maybe I'll ask about it.

Hmm. I'll think about that.

Yes, YES, YEEEEEESSSSS!!!! BWUAHAHAHAHA!!!! I'm already thinking of how to build it. Hmm again. BrickWorld Chicago is in about two months, and I'm going. Maybe I'll have a prototype by then. It would be interesting to see what the MindStorms guys there think about it.

[HaWe]

Making custom sensors: What would be the demand?

for me the demand would simply be:
anything that makes it possible to connect 15 NXT-motors and 40 NXT-sensors to 1 NXT...

[mattallen37]

Helmut, such expanders are already available. All you need to do is program for them to work with your specific application.

[HaWe]

wot?
Expanders for 40 NXT sensors (Lego, MS, HT, Dexter, homebrewed) at 1 NXT plus 15 NXT encoder motors???

never saw them
can you show me please??

[inxt-generation]

It's easy (albeit exspensive) to attach many motors to one NXT. Just look at doc222's work. But I'm with the Doc, Matt. Where's the 40 sensor ports?