Have an idea - but where to start?

[JugglingReferee]

Hi.

I have an idea of a very simple device I'd like to create, but I don't know where to start.

I gave my particulars to a design company and their quote was $10,000 to design the device and produce a prototype.

My device is nothing except a countdown timer, that sits in a box that clips to your belt.

So, I need a power button, a start button, a stop button, a reset button, and a simple LCD display, and a battery of some type (hopefully a AAA or CR2032).

If I learn electronics, can I get this idea to work using a breadboard?

What are good sources for parts? DigiKey? Newark? Others?

Thanks for reading.

 

[seadolphine2000]

If I learn electronics, can I get this idea to work using a breadboard?

It depends on how good you are in electronics.

Here you are some websites where you can get some ideas.
http://www.electronic-circuits-diagrams.com/
http://www.discovercircuits.com/list.htm

What are good sources for parts? DigiKey? Newark? Others?

You can buy your components from:
www.Microchip.com
www.TI.com
www.Maxim.com
As well as Digikey, NewArk, Mouser...etc.


Good Luck.

 

[Chipmunk]

I gave my particulars to a design company and their quote was $10,000 to design the device and produce a prototype.

My device is nothing except a countdown timer, that sits in a box that clips to your belt.

So, I need a power button, a start button, a stop button, a reset button, and a simple LCD display, and a battery of some type (hopefully a AAA or CR2032).

It's hard to believe these guys wanted to charge you $10,000 just to implement such a trivial functionality. The only challenge I can see here is to design a cute looking plastic case for this gadget so that it will appeal to the customers. Just building the electronic prototype should be trivial.

Or do you have any special requirements regarding power consumption ? How long do you want the battery to last ?

EDIT: Please forget my ranting about $10,000 beeing too much, I mistakenly thought that any old countdown timer would do, but see JugglingReferee's detailed spec below from Feb 22, 3:49

 

[karesz]

..Or do you have any special requirements regarding power consumption ? How long do you want the battery to last ?

Hi,
think pls an battery CR2032, how high is their current capability?...
K.

Added after 3 minutes:

@JugglingReferee,
What are you thinking to display an the LCD pls, is it a 16x1 or 2 lines type or some others?

Added after 2 minutes:

Dont forget pls, if you will begin to learn electronics, it needs some moths/years & you must know not only software but some hardware too...
I think thes was reason for the 10K$ bill.

 

[biff44]

The correct way would be to design this with a simple microcontroller. How involved the design gets depends on how fancy you want the design. If you are happy with a chip with wires hanging off the board, it will probably cost you a few hundred for a standard prototype board from a manufacturer (like Microchip), and another few hundred for some hobbyist to program it for you.

If you want a consumer product that can be mass produced for a profit in large volume, $10K is cheap, since you want a really good design using the lowest cost components, best packaging design, and a truly manufacturable design. Cut a few thousand out of the design cost, and you will literally lose millions of dollars in production over a few years due to much higher manufacturing costs and poor reliability.

So, the call is yours. Exactly how DID you specifiy the tasks to be designed.

 

[JugglingReferee]

@seadolphine2000: Thanks for those additional links.

@Chipmunk: I don't believe that I have any special requirements for power consumption. If I used a CR2032 battery, I would expect that this battery lasted for a very long time. The CR2032 in my Timex Ironman watch typically lasts 3+ years. My device seems to suck less power than a watch.

@karesz: The display I want is further down...

@biff44, and others: I sent specs to a few companies like this:

Housing: small plastic case (like a TV wireless microphone transmitter) with belt clip
Power: Either CR2032 or AA or AAA (that can use re-chargable AA or AAA batteries)
Display: LCD (since LED requires much more power), with a minimum of two (2) 7-segment characters, although a 1-line, 16-character display would be nice
Switch:
a) Countdown/Set
Buttons:
a) Power (responds only to holding for 4-seconds, much like the PC power off operation)
b) Start/+
    i) Start in Countdown mode
    ii) +1 in Set mode
c) Stop/-
    i) Stop in Countdown mode
    ii) -1 in Set mode
d) Reset Timer 1/Active Timer 1
    i) Resets Timer 1 to user-defined value in Countdown mode, and also stops the countdown if it was running when pressed
    ii) Sets the display to show the value of countdown timer 1 and change according to + and - buttons
e) Reset Timer 2/Active Timer 2
    i) Resets Timer 2 to user-defined value in Countdown mode, and also stops the countdown if it was running when pressed
    ii) Sets the display to show the value of countdown timer 2 and change according to + and - buttons

The device would be marketed to the public, although to only a smart market. Customers in both the USA and Canada would be interested. The device isn't a necessity for customer, but is a nice to have. I expect that people would pay as much as $50, but I think $30-$35 would really make the device take off.

If my idea is a pipe dream, then so be it. But I won't give up easily.

Added after 1 hours 54 minutes:

I'll be watching these videos this week.

**broken link removed**

 

[Chipmunk]

I think what biff44 wrote on Feb 21 17:53 makes much sense. If you want to have a prototype just to demonstrate the concept, find some hobbyist to do it for you. Learning to do it yourself might be fun. But from an economic point of view it doesn't make sense to invest many months in learning if you are only interested in getting this one prototype to work. And there is absolutely no chance that you on your own could come up with a design that is optimised for manufacturing.

One more thing: If you decide to ask somebody to design the "real" product for you (i.e. one that is ready to be produced in volume), make sure to include you expectation that the battery should last for a very long time in your specification. If it's not absolutely necessary include it as "nice to have", then the designers can discuss with you how much effort they should spend on reducing power consumption.

Also spend some time thinking about any other features that you haven't included in your spec because you just take them for granted, e.g: Temperature range ? How robust does it have to be ? Make sure your spec is as complete as possible. Do not assume that anything comes "naturally" just because you have seen it in every other product. There is nothing worse than discovering that a crucial feature is missing when your gadget is already in production...

Best luck with your product !

 

[JugglingReferee]

Thanks for your advice Chipmunk.

That's what I am going to do.

 

[hussong1555]

IMHO $10,000 is spot on just for the plastic enclosure to be designed and prototyped, sounds like your getting the rest of your project for free.

I read through your requirements and I must say that this sounds like an expensive dressed down rather large stopwatch; am I missing something from your description?

 

[ProEngineer]

I used to work for a consulting firm that specialized in just this sort of project, and I can tell you that $10,000 is a ridiculously high sum to develop this. No legitimate firm would recommend you design your own case or LCD for someone like yourself - you'd never sell enough of them to recoup your investment. It should take an experienced engineer no more than a week to do the circuit design, layout the PCB and write the software. It would certainly take me less time than that!

There are plenty of enclosure manufacturers that sell stock cases that could be used, and there are off-the-shelf LCD's that meet your requirements.

Look for a design firm that’s experienced working with small companies – many of them will put together a prototype using parts they have from other projects. If you use a hobbyist beware that you get what you pay for. If your project fails, is it because the hobbyist lacked the knowledge to design your product? If it works and you then want to go into production, can they support you through the whole process?

 

[FvM]

What you presented doesn't involve more than a µP, a watch crystal, the LCD (most likely a non-mux 7 segment due to low power requirements), the battery, a bypass capacitor and the keys.

CR2032 operation for 3 or 5 years seems at the edge of standard uPs power consumption (it's capacity is about 5 µA@5 years), but seems possible though, e.g. with a "nanoWatt" PIC or MSP430. Consider, that standard products like a digital watch are using specialized ASICs rather than general µPs, which allows for considerably lower power consumption.

Standard LCD have 1 uA/cm² current consumption however, so the display size would be limited to a small area. All standard character displays are using multiplex drivers with resistive dividers and a "huge" power consumption - related to the available battery capacity.

 

[JugglingReferee]

@ hussong1555 and ProEngineer: My previous employer was a company that worked in the high tech industry. I had enough systems development knowledge to know that $10k - or even $8k (imho) is high to develop this item. You should see what I could get done for $5k for a device that handles HD television!

hussong1555, you're right that my device will be "expensive dressed down rather large [countdown timer]". It's the nature of small specific devices, though.

ProEngineer and FvM, I hear what you're saying. I'm going to contact two local colleges and see if they have an electronic design class to see if an upper-year hobbyist wants to help develop a real-world device. I'd be willing to give him/her a good portion of the profits as payment for developing the device.

The maximum countdown value is 99 seconds. Hence, I'd need at least a 2-digit 7-segment display. When the user sets the value of the countdown timer for their purpose, having the display say "Set T1-xx" where xx is the value from 01 to 99 seconds would be ideal. That's 9 characters, so a 16-character display would work nicely.

 

[JugglingReferee]

I was wondering if someone could confirm for me what steps are necessary to complete this project.

1. I need to have the circuit designed, correct? So that means drawing a circuit on paper with all the resistors, chips, components, etc.

2. Test the design with a breadboard and the other parts. Revise if necessary.

3. When it works, transform the layout to a PCB software package so that a PCB mfg company can manufacture the prototype board.

4. Test the prototype. Revise if necessary.

5. Once it works, plug it into a belt-clip case of some kind.


Am I missing any general steps? Please feel free to treat me like a beginner, because I are one!

 

[hussong1555]

I recommend downloading KiCAD or eagle to do your schematic design (KiCAD is free and eagle has a free version).

once you are happy with it, you can also use them to do the pcb layout.

once layout is complete you will need to build gerber files for board house.

then once your ready for production, you will need a cm to populate the boards.

make sure when you design the board you include test points, i would also build a test fixture to supply the cm for testing before assembly.

 

[hameeds01]

it sounds just like a piece of cake
u can built such device under 6$ if u r in pakistan

 

[doraemon]

Hello!

it sounds just like a piece of cake
u can built such device under 6$ if u r in pakistan

It depends on how far you go in the design.
If you take a microcontroller that sleeps in a drawer, other parts that you
already have, then you can even build it for nothing, even here in japan.

Now, if you think about your time, you will need, say, half a day do do the
hardware, half a day for programming, debug, test, then you cannot do it
for 6 USD, even in Pakistan I guess.

Then, if you are a company:
- You have to pay one engineer to do it;
- Usually, if you are asked for a design, you have to first choose components
that are optimal for the application, think about the cost in case of mass
production, etc… In this case, 1 cent is important, and this phase takes time.
It also means that you will have to design a PCB, a case, documentation,
etc…
If you add everything, I would say about 1 man month.

How much is one man month for a company? Probably cheaper in Pakistan than
in Japan, indeed. In Japan, it's about 1 million yen (pay the engineer, pay the
office space expenses, heater or cooler depending on the season, electricity,
taxes, social insurance, development tools (oscilloscope, etc…).
1 million yen is 11000 USD. So I think 10 000 is not overestimated even for that
kind fo small product.

Dora.