Official circuit design help post :/

Hey everyone. It's been a while since I have posted here, I have had some family issues and random stuff pop up and take me away from my rgb build. It's sitting in a box under my bed :/vim still trying to figure out the pcb there since I was scammed, and can't really put more money into it.

Where I'm looking for help is with a project that I have diverted most of my time to now. It is a ring I'm hoping to build for my gf before she comes home. I can't really talk about where she has been but I have a bit more time to try to finish this project.

I'm hoping to use the ltc3105 chip to convert thermal energy into just enough voltage to lite a few LEDs. Does anyone have experience with this chip? Or would anyone like to help design the circuit? As a side project with reimbursement...

And I was wondering why somewhere doesn't have a place that can produce customizable partner chips for microchips. Think, resistor array, except with different values, capacitors, inductors and other passive components all in a single chip.

monster.com

Something this special, for a fiance's ring, warrants proper funding. Not low-balling hobbyists.

Get the right help and get it done.

Have you even done any research on that specifically named chip? Looking at the specs on the website it looks like a typical step-up regulator, not something that could harvest thermal energy. Anyway, that thermal energy harvesting idea is a crapshoot for those low temperatures: you need large temperature differentials for usable power. You'll never get that in a ring unless you attach it to a radiator or something. Why don't you just use a battery?

Hell, if you want the ring to light up, go fix a tritium vial in it. It'll last for 30 years and requires no electronics... if that costs too much you need to GET A (better paying) JOB like Pschlosser is implying. Your GF will be more happy with that than some ring.

As for your last pondering: because manufacturing custom components is expensive, and reconfigurable components are expensive too. Think of the orders of magnitude price difference between even a potentiometer and a resistor. Whereas a cheap pot may go for $1 in bulk, resistors cost $0.01-0.05 a piece. Same thing for making computer chips (I got to make an integrated circuit for a mere $1000, and all it does is ADD TWO BITS!), vehicles, artwork, or basically anything in the world.

Just as a reference, it costs about $10,000 to make a custom design 1x1cm chip on silicon. That'd be why

The $1000 figure I quoted above was for 1x1mm^2 so 1/100th the size -- and that was still a good deal for that process (normally it's $1200/mm^2 if you buy a bunch at a time), and included the workshop too.

pschlosser: this is a special project, but not that special...

im sorry to have misled you guys, im not proposing yet..haha but it is for a special occasion. at which i am not at liberty to discuss publicly.

i feel like the hobbiest community contains professionals well suited for the needs of this project. as i am not an EE, i have trouble understanding most of it, but i know there are plenty of people out there that can throw this (design the board) together in an hour.

bionicbadger: this dc/dc converter will be used in conjunction with two HV56 eTEG modules by micropelt, they will be putting out .6V at 60mA each at 50K delta. the average i am predicting is 30K which is about 25mW total. all insidde a 3mmx3mm package. im using ultra low power leds, i will test a few different ones, the 1mA 1.9V up to the 5mA 2.9V to see which work best for diffused lighting. i know its not going to be that bright.

the body of the ring i will be using is not cheap. i have two ceramics, the first, which i may use is zirconia stabilized alumina. the other is a yittria toughened zirconia alumina.
if these end up too brittle i will fall back to a nice titanium. i have a machine shop that can work the tolerances. i just need the circuit to put in it.

i came across moses, they do custom dies apparently? but i just think it would be a decent investment, if TI came out with a chip, that was going to sell in the millions, and made a companion chip that contained all the resistors and capacitors the chip needed externally. saving on size and complexity of the final product