SparkFun Forums 

Where electronics enthusiasts find answers.

Questions relating to designing PCBs
By seraph44
#161379
Hello everyone,

My coworker and I are looking for someone who has experience in designing simple circuit boards; micro circuit boards to be exact. We don't have knowledge on building them, but have a very thorough idea of the functions we want it to perform.

To give you the basics we need a board about 1 in. width and 2-3 in length. We need a micro temp sensor with accuracy of 1 degree C or greater. We need an activation button which activates the first light, if the temp sensor reaches a certain degree it activates a timer and a 2nd light. If the temp does not come back to normal range within a set time it activates a 3rd light. If it does come back to normal range it maintains the second light. We only want the circuit to operate for 72 hours upon activation.

If anybody can help with this project please contact us. This is an idea for a business with great potential, we are looking for a business partner or someone that wants to get paid for the work. This idea needs validation which will be handled on our side.

Thank you,

Serafin
Hamton Roads, Virginia USA
By skimask
#161395
One PIC12F1822 (or 1840, doesn't matter), one tact switch, one coin cell, 3 LEDs, like 20 lines of code, PCB maybe 1/2" square at most, and it'll likely operate for a month after initial activation.
I'll do it for $10K up front, $25 per module up front, plus 10% of any sales after.
By TopHatHacker
#161396
With Thermal Mass, that pic just isn't going to keep up with the temperature fast enough. I'd say use a TMP36 sensor that's DESIGNED to read temperature, not something of an afterthought they threw into a uC. Also, $10k is a bit steep. I'd do it for $5k upfront, $25 per module up front and 10% of sales.
By skimask
#161412
TopHatHacker wrote:With Thermal Mass, that pic just isn't going to keep up with the temperature fast enough. I'd say use a TMP36 sensor that's DESIGNED to read temperature, not something of an afterthought they threw into a uC. Also, $10k is a bit steep. I'd do it for $5k upfront, $25 per module up front and 10% of sales.
Oh yeah?
Well, I'll go as low as $4K in advance (after the initial prototype, but before full scale production), $24 per module in quanity, and 9% of sales after an initial batch.
User avatar
By unexpectedly
#163104
skimask wrote:One PIC12F1822 (or 1840, doesn't matter), one tact switch, one coin cell, 3 LEDs, like 20 lines of code, PCB maybe 1/2" square at most, and it'll likely operate for a month after initial activation.
I'll do it for $10K up front, $25 per module up front, plus 10% of any sales after.
I promise I'm not being a smart-aleck, but is this a serious quote?

Just curious...
Chris
User avatar
By Ross Robotics
#163122
If you read all the posts, you would not be asking that question.

On a serious note, most engineers get paid in upword of 6 digits. But freelancers like to charge by the hour, around $50+.
By UNTEngineer
#163136
codlink wrote:If you read all the posts, you would not be asking that question.

On a serious note, most engineers get paid in upword of 6 digits. But freelancers like to charge by the hour, around $50+.
This sounds like a super simple project...the up front charges sounds like price gauging to me (or a joke).

$50/hr sounds good, but as Im busy with other projects, I'd have to decline.
By lyndon
#163151
UNTEngineer wrote:
codlink wrote:If you read all the posts, you would not be asking that question.

On a serious note, most engineers get paid in upword of 6 digits. But freelancers like to charge by the hour, around $50+.
This sounds like a super simple project...the up front charges sounds like price gauging to me (or a joke).

$50/hr sounds good, but as Im busy with other projects, I'd have to decline.
Really? Let's take a look, shall we.

The description is that it is a circuit that has three input buttons, some timing features, and a temperature sensor and the basic operation is stated. So the basic requirements are already known. A HUGE plus in my experience.

So assume we don't need too much work to get the exact requirements that everyone can agree on. Let's assume that this takes a grand total of two hours of writing and reading email, documenting the requirements and making sure that everyone agrees.
Stage 1= 2 hours

So how big can this PC board be? Does it have to fit into an enclosure? Is it battery or A/C powered or both? What happens on battery/AC failover? What size battery? What kind of power cord? Is he going to build 1, 10, or 10k? Let's say this takes (being generous) about 3 hours to resolve.

Stage 2= 3 hours

We now know the operational requirements and physical limitations and what production quantity to design for. The basic circuit design and initial firmware will likely take under 2 hours. Assume you've done something similar so component selection, including checking for price breaks at the expected periodic quantities and parts not on limited lifetime and ordering will take less than 2 hours. PC board layout another four (remember to capture your time on everything from starting up the tool to getting verified, good Gerber files out) if you're fast.

Documenting your design decisions, parts, generating BOM, let's say 2 hours.

Stage 3= 2 + 4 + 2 + 2 = 10 hours

Board comes back from vendor, components are in hand. Time to assemble it
Everything goes well, first board powers up in 1 hour. Firmware loaded. Runs the "go path." Create a set of tests to test the edge conditions and any likely failure modes. Create a board test fixture. Perhaps add self-test feature if it can be supported. Being optimistic, this takes about 6 hours.

Stage 4= 6 + 1 = 7 hours

Collect all documentation and make sure it's in a "shippable" state. Make sure you have filed all information away for the inevitable change requests and when you're asked an obscure question about this design a year later. Let's say that all gets done in 1 hour.

Stage 5= 1 hour

So for all stages, assuming nothing goes wrong, and applying some little measure of engineering and manufacturing discipline it took us 23 hours to conceptualize and ship this little temperature timer board. At a low-average rate here in the Midwest of $100/hour, that's $2,300 for the first unit.

That price is unrealistically low, but I don't think it's too far off. Remember, in all cases, I assumed everything goes perfectly right and this project is "cake." That's an easy way to screw yourself over.
User avatar
By Ross Robotics
#163152
I agree with your assessment, but I tend to work in volume. I will make the same profit over, let's say 3 projects. But, I have 3 times the chance that they will come back with more challenging work, which adds to the price. Besides, it keeps me busy.
By lyndon
#163158
That works great if they keep coming back. After hearing enough times, "do this and it will mean tons of more work" followed by, "sorry, we didn't get the contract" or "my client cancelled the project," I just started pricing more realistically. No more Mister Nice Guy ;-)
User avatar
By Ross Robotics
#163161
Ah, yes, I've heard all those and then some. But I think area has something to do with it as well, I'm about 3 hours from a major city. I would imagine that if I were in a large city, my pricing would change.
By UNTEngineer
#163222
Thats a great breakdown, I never really knew thats what it took. So for something so simple, it can end up costing 2-4k for development of it. That doesnt even cover overhead costs... Cool.

I take back my "price gauging" statement. ;)