Could someone answer a few questions for an ultraneophyte? Please. I need some

Ken Pierce

Hi thanks for reading.

1) CAVEAT: I have no idea what I'm doing but I'm trying to do about 10 things at once.

2) PROBLEM: I need to develop a product which can accept keyboard entry, read an id from a card, query a database on a machine within ~100-200 ft, get back a few values and then transmit a line of data based on the feedback and a few entered values. I believe it could work with a 16-key keypad and an LCD with a few lines. I'm a statistics/data guy not an electronics guy. For example, say I was doing botanical research, I might transmit a code for a plant family and receive back a list of species from that family in a database on a computer, I could then arrow through the species then put a count in another field and send the line of data <user_code, family_code, species_code, count, time_stamp> to a database. At any one time there could be several hundred collectors in the local area transmitting.

3) QUESTION: I am willing to dive into getting a starter kit, etc. If I figure out how to get something working with a Basic Stamp module, would I implement it eventually with a different cheaper product such as the SX chip? OR should I start with a different product if I need to develop a commercial design? OR should I just get the thing working in any way possible for demonstration then get someone who actually knows what they're doing to develop a final product, assuming I get funding? (That last part might be for someone with patent experience)

4) THANKS: I sincerely appreciate any responses anyone takes the time to provide. My electronics experience doesn't go very far beyond the computer I built which I realize really means I have no experience at all since its just plugging things together. Most of my programming experience is in high level languages like S-plus which I guess is really a scripting language at best. Minimal visual basic, fortran and C. Eventually I hope to need help that will be accompanied by at worst minimal compensation and at best reasonably lucrative compensation.

Thanks,

Ken Pierce

Bongo

At a glance the fasted path to implimentation would be palm pc's.· Would I be right in guessing you are looking for a cheaper option?



bongo

Ken Pierce

Definitely cheaper and it needs to be a dedicated module. I would hope ultimately in the $50-100 per unit cost. It actually couldn't be something like a palm I don't think. Thanks for the input.

Gadgetman

I assume we're talking about some sort of wireless 'terminal' for accessing a database, right?

And that the database is on a centally placed computer?

What you're thinking of should be doable, if you pick the right transmitter/receiver equipment and have a decent database front-end.

In that case all the stamp needs to do is to ask the server for a list of options, maybe the textstrings to be output, then wait for the user to make a selection, relay that back, get a new list and so on. (A little bit more than that)
What is important is that the Stamps doesn't have a buffered serialport, and that all transmissions needs to be started by the stamp.

As the system will have to receive a lot of text I suggest that you look at some of the multibank type BS2s, the BS2p/pe/px as they all not only have a Scratchpad RAM, but also allows you to buffer incoming serial data in the Scratchpad RAM. (This area is 128Bytes on these models)

If you need some 'responsiveness' in the system, you may want to have it store frequently used strings in EEPROM (The BS2p and BS2px has 8 Program slots of 2KB each, so you can spare a bit of room, and the BS2pe have another 8 DATA slots, giving you a whopping 32KB in total to play with) and let the server transmit codes instead of those strings.

One thing to consider is using fixed-length strings. They may waste EEPROM or Scratchpad space, but they will make it much easier to store and retrieve data.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Don't visit my new website...

Ken Pierce

Good info. Yes and Yes to your first to assumptions. I expect the return list of options to the user to be limited to about 10-15, 16 character strings. The codes would work because the total options at any particular module will be a small subset of the overall dataset. Thanks.

pjv

Hi Ken;

A couple of observations based on my experience in these matters...........

Firstly, you mention patent experience, so I suspect you are wanting to commercialize this product. Based on that, $50 to $100 per unit is unrealistic unless you expect to produce thousands of units. In small quantities the raw compnents and modules will exceed this, or at least nearly so.

As a commercial product, your customers will expect a finished design, not a "shoebox stuffed with goodies", and that will probably entail a custom plastic housing. That alone will eat up the best part of $25,000 to $50,000 in mold design and fabrication. After that, the plastic pieces will cost a few bucks each. Hence the need for larger volumes.

If you can live with some enclosure that is already available on the market, then that mold cost can be saved, but then the design will be constrained by that enclosure.

If RF links will be used you might need FCC approval, and that could be another scary adventure of $50,000 for consulting and application fees.

Patents may be another $25,000.

For a "shoebox full of goodies" proof of concept prototype, the material costs should be less than $200, and a development effort in the order of a few hundred hours, perhaps less for someone who has a bunch of the harware and software modules "on the shelf"

A "finished" commercially acceptable product in an existing plastic eclosure would roughly double that development effort, as it would include documentation, schematics labeling, shipping/display boxes etc.

And a finished custom product designed for volume production would likely double the latter again.

So, I'm not suggesting you shy away from the project, but be forewarned that it takes a lot of effort to get to someting commercial.


Just some pointers for thought......

Cheers,

Peter (pjv)

Ken Pierce

Thanks. Those numbers make sense, I have no frame of reference to estimate those costs but they sound reasonable. I envision thousands being produced if it works. The prototype investment also seems within my expectations. What are the options for wireless transmission of data? I hadn't considered the FCC issue. The utility of my idea is collecting the data easily. The application and selling point is the subsequent analysis. The "shoebox" is just to see what data I can get transmitted so I can make the analysis software as robust as possible.

PJAllen

ultraneophyte

Thanks for not using newbie, noob, or the like.

Kevin Wood

ken,

I don't know that the BS2 would be right for this, but for proof of concept, you could get a Discovery Kit or RadioShack What's a Microcontroller Kit, a database, and a programming environment for your pc, and build a serial or USB version. Once you get that part working, add the wireless and off you go.

If you need recommendations for any of this, post or pm, and I'll share what I can.