Cheapest object sensor on the market!

embed

or

perfect

for

it

Ok, give me you best shot. Can not pass more the $0.80 total cost, can yo do it?

Sure. But I can't take on any additional designs right now. You'll have to experiment a lot to make sure it's sensitive enough but not too much. Look at what you already have on the card. You could use a CMOS inverter as an amplifier or take a comparator. Make sure the output is a clean digital signal.

--
Regards, Joerg

http://www.analogconsultants.com/

under

last

Avr169

dollar.

what

last

these

cheaper

But would it handle the voltage and supply a good brightness to the LCD t for a readable under a bright large rooms during day time. If so, this wil help me on having a thinner product.

The main question would be: $1 at what quantity per month? I assume you mean parts cost without PCB but with SMT assembly (in China).

--
Regards, Joerg

http://www.analogconsultants.com/

Did you ever deduce the motion detector in the supermarket cooler, Lewin?

Michael

are

i

selling

of

100,000 for 1st. order, see the estimated break down;

0.25 for PCB with assembly in china most probably.

0.25 for LCD 0.50 for Micro 0.25 Coin Batteries or Solar cell.

Now the challenge is to find a solution for $0.80. I need some one t simplify the idea as its a disposable item.

They have mentioned "down below" to drive the glass directly from th MSP430 in order not to use a sensor module to cut down the cost!!

do you have another option?

The only way to meet this price point is to build a mixed signal ASIC and mount it directly on the LCD. $100,000 would be a good starting point. We would need a DC-DC booster regulator, switching AC sequencer, analog to digital converter, segment decoder and analog operational amplifier.

does

fingers,

hours

you

Thank you on your great alternatives, you guys have mentions som alternatives:

1. Using uC and PIR (Infra-Red heat sensor).

1. Drive the glass directly from the MSP430 (Texas Instruments
   formatting link

2. Build a mixed signal ASIC and mount it directly on the LCD adding DC-D booster regulator, switching AC sequencer, analog to digital converter segment decoder and analog operational amplifier.

1. How about the crude touch-sensors from early tv remote controls (thin

70th)? These where just two electrodes (part of the PCB) and a very simpl analog transistor-circuit that measures the presence of body-resistance.

1. LED modulated with 40KHz. A photo-detector can detect the reflection o the 40Khz. The strength of the reflection can give a simple distance.

2. Send pulses to an LED. Detected the reflected light. Ignore light tha doesn't appear to be from the LED pulses.

1. "Digital theremin". With essentially a schmitt-trigger inverter and few passives, can create a capacitance-based circuit to detect th approach of a body part. But it is fairly difficult to design one that i tolerant of voltage and temperature changes. Easier with a micro. TI als has an app note on doing this with an MSP430F20xx,

2. Devices like motion-sensitive toys use an AC-coupled circuit triggered by a change in the resistance of a CdS cell.

Let me refresh the requirements:

1. I will be using this device in a constant light ambient so turnin lights on & off would not occur.

1. During detection the device will be fixed, will not be moving.

2. Will run on batteries and Needs around 30 to 100 hrs. of power on.

1. 1 digit display; either LCD or LED which ever cheaper & less powe consuming.

2. the LCD or LED will be counting till 9 votes max. accumulative.

1. Detection distance can be 10mm to 100mm. does not matter, as hands o fingers will pass very close to the device.

2. No one will be using gloves.

1. No memory needed. I need to accumulate votes when it is on "it loses all its memory when it shuts off" and when it turns on again will start counting all over again.

Any one can tell me, initially, what seems the cheapest solution? Can you breakdown the cost on the best alternative?

Regards, Raed

Patently not true. There are bunches of other options - micros for the electronic toy market, for instance.

High for a single-sided board

High unless it's bigger than watch size

High unless you insist on a powerful micro

Very high for a coin cell

Sounds like the sensor is your issue.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Now that's finally a good number.

That can be improved. Hint: Consider moving from FR4 to phenolic. I don't think you'd have any other option for this project.

For 1-2 numbers? Sounds high. I thought you only need to count to 9.

You can also do better here but only with Asian products and most likely you'll have to enter the not-so-pleasant world of 4-bit coding.

Renegotiate that. $0.25 for a coin battery is way too high a price at

100K qties.

Even at Digikey you can get 4000qty 3V cells for under $0.15:

?name=P189-ND

Honestly I think your first avenue would be to negotiate a bit tougher.

The MSP430 would likely be too expensive. You can try but it'll be hard to get one that can drive glass for under $0.70. Look around for other uC. As I mentioned before one option would be to use on-chip comparators and a kind of "touch key". You'd set the touch key to VCC/2 via two 18M resistors or thereabouts, then from there to the comp input via another

1M (for some ESD protection). The other input of the comp would get an RC filtered PWM from the timer. You'll have to detect excursions in either direction because you won't know the charge polarity of an approaching hand. Just a thought, try it out.

--
Regards, Joerg

http://www.analogconsultants.com/

We are constantly looking. No reasonable LCD uC under $1 @100Ku.

However, using custom ASIC with power managements for LCD drivings, there are plenty of uC for 25 cents. ASIC would cost another 25 cents.

Don't forget assembly costs as well, which are going up. Inflation in China is even worse than the US.

Ok, but somebody must be building the chips for those radios that have an LCD and retail for $0.99 at times. Not that they are any good IMHO but they must turn a profit.

True. That's why one has to take a hard look at the materials. For example consider phenolic board.

--
Regards, Joerg

http://www.analogconsultants.com/

LCD uCs are for wimps. Use a regular uC and some resistors. Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Yes, they are most likely ASICs. LCD driving circuit does not cost much to make. But if you want it in a general purpose uC, they will jack up pin counts, memories, cost, etc.

and only 10% of the battery.

I haven't priced out 4-bitters with small LCD capabilities like the Panasonic 1500 series lately. Their 150404 maybe? Has the comparator on the chips AFAIR.

Or talk to companies in the wrist watch business:

[...]

Regards, Joerg

Spehro's idea might actually be good if the uC has enouigh pins. AFAIR the OP was mentioning a 30hr runtime. That's not a whole lot for a 3V coin cell.

--
Regards, Joerg

http://www.analogconsultants.com/

Might work for static LCDs. Multiplexed (x4) LCDs are very tight in voltage tolerance. It becomes unreadable below 2.5V. You can't get constant brightness without proper voltage regulations.