Kickstarter for scriptable USB RGB LED(kickstarter.com)
kickstarter.com
Kickstarter for scriptable USB RGB LED
http://www.kickstarter.com/projects/thingm/blink1-the-usb-rgb-led
4 comments
I would rather sell to a million home users at a realistic price than to a thousand "industrial" users for some obscene amount for a small part like that.
Or maybe: please do sell this for $30+ so that we can buy it on deal extreme a month later for $5 :-)
Or maybe: please do sell this for $30+ so that we can buy it on deal extreme a month later for $5 :-)
There is a software-only USB stack for AVR microcontrollers, and it has a example section with LED and Display projects:
http://www.obdev.at/products/vusb/prjdisplay.html
Especially this project is almost identical, but has a different form-factor:
http://www.vandenbrande.com/wp/converting-a-led-cube-into-an...
(look at the schematics, and keep in mind that the ULN2003 is not needed for driving lower-power LEDs, and there are microcontrollers running the VUSB stack that can use their tunable internal RC oscillator, so the XTAL can go away, too...)
For the cost, it's difficult to say what's a reasonable amount: I like the idea to retrofit a general purpose notification LEDs to computers, but have no immediate personal use. For one-off projects every "maker"/hacker/tinkerer/electronics-enthusiast I know has maybe 10 unused boards of various kinds laying around where one could retrofit such an indicator (in a ugly way, I admit) in 10 minutes.
The only way this could be made accessible to the normal non-savy home-user would be to have an assortment of ready-made plugins for all major email-clients, instant-messaging, twitter, skype or long-running PC tasks (has this upload already finished?)... Maybe Outlook notifications? And creating all those will be a pain in the...
http://www.obdev.at/products/vusb/prjdisplay.html
Especially this project is almost identical, but has a different form-factor:
http://www.vandenbrande.com/wp/converting-a-led-cube-into-an...
(look at the schematics, and keep in mind that the ULN2003 is not needed for driving lower-power LEDs, and there are microcontrollers running the VUSB stack that can use their tunable internal RC oscillator, so the XTAL can go away, too...)
For the cost, it's difficult to say what's a reasonable amount: I like the idea to retrofit a general purpose notification LEDs to computers, but have no immediate personal use. For one-off projects every "maker"/hacker/tinkerer/electronics-enthusiast I know has maybe 10 unused boards of various kinds laying around where one could retrofit such an indicator (in a ugly way, I admit) in 10 minutes.
The only way this could be made accessible to the normal non-savy home-user would be to have an assortment of ready-made plugins for all major email-clients, instant-messaging, twitter, skype or long-running PC tasks (has this upload already finished?)... Maybe Outlook notifications? And creating all those will be a pain in the...
Yeah, ok. I completely agree with everybody else. The BOM is basically a LED, a USB connector, and an ATTiny chip. It looks fantastically useful, but at their price point, just knowing who easy it is to build makes me feel ripped off.
Super cool but WAY too expensive for what it is.
That thing should be $5
That thing should be $5
Given the recent criticisms of other hardware Kickstarters that were perceived (quite likely correctly) as being unprofitable and a risk it is kind of refreshing that they've staked out a profitable price point to start.
They do open source the plans so I suppose it is possible to build one yourself for less.
They do open source the plans so I suppose it is possible to build one yourself for less.
Something like this shouldn't cost more than $0.50 in parts to make, and it shouldn't cost more than a computer mouse, which already has not only an LED but a relatively expensive optical recognition system built in.
It shouldn't cost more than $0.50 when you're building millions of units monthly overseas. This group is obviously not there yet. Do you understand the economies of scale and geography at work here?
When people say X should cost Y, what they really mean is that they only value X at Y.
You can't get them to increase their valuation by arguing that X can't be made profitably for price Y.
You can't get them to increase their valuation by arguing that X can't be made profitably for price Y.
Good point.
So why is the perceived value dropped so low? Is it the plethora of teardown sites that show you "oh, SuperWidget Tab4 only has two chips and an LCD in it. How expensive can that be"? Or is it the deluge of cheap commodity stuff on Monoprice and NewEgg that has inoculated us as how this stuff needs multiple generations of devices to cost-reduce the toys down to a price that is finally acceptable?
So why is the perceived value dropped so low? Is it the plethora of teardown sites that show you "oh, SuperWidget Tab4 only has two chips and an LCD in it. How expensive can that be"? Or is it the deluge of cheap commodity stuff on Monoprice and NewEgg that has inoculated us as how this stuff needs multiple generations of devices to cost-reduce the toys down to a price that is finally acceptable?
The amount of electronics required to connect a commodity LED to a commodity USB connector is so minimal. The very same sort of component goes into every single computer mouse sold today, many of these costing less than $1 to produce.
$30 is extremely expensive for such a toy. Do they think I'm spending $3,000 to make a pretty blinky-light server rack?
Anyone with a credit card can get a custom anything manufactured if they're resourceful enough and know how to communicate with their suppliers.
I think the makers of this product should do more research in how to cut costs, find a part that does 95% of what they need and come up with a way to bridge that gap in software.
$30 is extremely expensive for such a toy. Do they think I'm spending $3,000 to make a pretty blinky-light server rack?
Anyone with a credit card can get a custom anything manufactured if they're resourceful enough and know how to communicate with their suppliers.
I think the makers of this product should do more research in how to cut costs, find a part that does 95% of what they need and come up with a way to bridge that gap in software.
Okay, and I'm paraphrasing Louis CK here, you go make one if it's so sucky and imperfect.
Seriously, how hard can it be? Get together a BOM for the $5 in parts, send it through DigiKey and have it in a couple of days. That's like what, an hour of time?
Then layout a PCB in your favorite schematic capture and send it off to a board house, quantity one. Two hours, tops. Never mind you're wasting 75% of the minimum PCB size.
While DigiKey and PCB Express are cooking, time to write some firmware. If you chose a nice chip like a Cypress part and have some experience, you could be done in another 2 hours. You do have a development kit on-hand, right?
Then take all the parts, solder them up in your reflow, burn the code, test, and assemble. Another hour? Sure. (Hope those boards were right the first time and didn't come in upside-down) Whoops, don't forget the drivers. Let's do a quick linux one.
So $20 in parts and board, and six to eight hours of your time. What's your time worth? $2 an hour? Cool, you're golden.
I tend to charge a bit more for my time, so it puts me in the range where $30 is a great deal. But then again, I'm using the old ScrollLock LED trick as my message indicator. It's free.
Seriously, how hard can it be? Get together a BOM for the $5 in parts, send it through DigiKey and have it in a couple of days. That's like what, an hour of time?
Then layout a PCB in your favorite schematic capture and send it off to a board house, quantity one. Two hours, tops. Never mind you're wasting 75% of the minimum PCB size.
While DigiKey and PCB Express are cooking, time to write some firmware. If you chose a nice chip like a Cypress part and have some experience, you could be done in another 2 hours. You do have a development kit on-hand, right?
Then take all the parts, solder them up in your reflow, burn the code, test, and assemble. Another hour? Sure. (Hope those boards were right the first time and didn't come in upside-down) Whoops, don't forget the drivers. Let's do a quick linux one.
So $20 in parts and board, and six to eight hours of your time. What's your time worth? $2 an hour? Cool, you're golden.
I tend to charge a bit more for my time, so it puts me in the range where $30 is a great deal. But then again, I'm using the old ScrollLock LED trick as my message indicator. It's free.
The same group is making a very similar product, minus housing, for a $15 unit cost. Clearly there's some serious slack in this Kickstarter price.
Probably not as much as you think. Remember that Kickstarter takes 5% and Amazon Payments takes 2%. Then the group has to pay federal taxes on that, which could be another 20%. Now we're close to $20 remaining for the device.
Oh yeah, and I didn't even get into the costs of tooling and shooting a plastic housing. That's something a little more time-consuming and pricey than firing off an order to DigiKey and PCBExp on your credit card.
Oh yeah, and I didn't even get into the costs of tooling and shooting a plastic housing. That's something a little more time-consuming and pricey than firing off an order to DigiKey and PCBExp on your credit card.
I think that pretty much everybody sees a single RGP LED as a super cheapr component. Even if they have no domain knowledge.
That's exactly what I meant!
I understand how hard it is though for the founders of efforts like this one - they are valuing all the work and effort it went to making this a reality. Thus they feel that a $30/unit cost covers all that effort.
The problem is they can often forget that the outsider will value the end product alone and typically at far less than the creators.
They have a mock up image of a bunch of servers in some racks. thats a hefty price for a bunch of servers with a fairly questionable value for the utility.
I understand how hard it is though for the founders of efforts like this one - they are valuing all the work and effort it went to making this a reality. Thus they feel that a $30/unit cost covers all that effort.
The problem is they can often forget that the outsider will value the end product alone and typically at far less than the creators.
They have a mock up image of a bunch of servers in some racks. thats a hefty price for a bunch of servers with a fairly questionable value for the utility.
Or less. You could buy an entire raspberry pi for what they're asking!
Home user/hacker: this thing only cost $0.50 to make, I won't pay more than $5 for it.
Industrial user: I need an indicator that I can hang off the USB port of the computer and light up when a new work order comes in so the guy at the other end of the workcell knows to start working on it right away without having to be sitting at the PC. $150 each? Cool, it'll pay for itself inside a week. Send me 20!