Simple 5 volt power cape for Beagle Bone.

Hello List;

I have developed a simple 5V 2 amp cape for one of my own projects.

The cape accepts anywhere from about 16VDC to 9VDC and provides about 2 amps @ 5VDC.

(There might be a 3 amp version of the switcher available.)

The board is fused on the supply side (9-15V) and reverse polarity protected.

If anyone is interested in purchasing a built and tested board, please respond to this post.

The price would probably be $125.00 USD or so, depending on how much interest there is.

(Blank PCB prices drop as quantity goes up)

Bill

Late breaking news !

The switcher I was using, has been replaced by a 28 volt DC input version !

So by changing the voltage rating of the input cap, I can now offer 7 – 28VDC input capes J

I have developed a simple 5V 2 amp cape for one of my own projects.

The cape accepts anywhere from about 16VDC to 9VDC and provides about
2 amps @ 5VDC.

<snip>

The price would probably be $125.00 USD or so, depending on how much
interest there is.

$125 per cape?

-Andrew

$125/cape!?? as simple as this is and as little as this cape does I’d likely be interested at somewhere between $12.50 and $25/cape but not $125. Heck the battery capes are only $48.95 shipping included! I figure I can put a power supply cape on a beagle proto cape for under $25 including the proto cape and all the parts!

Eric

Good luck Eric;

The switcher alone is $23.00 unless you would like to buy 1000, in which case it drops to a mere $13.00 !

Go to Digikey and type: PT78HT205V-ND For the 2 amp version or PT78HT305V-ND for the 3 amp version.

BTW That’s all the info you get for free !

Bill

Don't buy a module, use an IC with internal switches and something like
TI's WEBENCH Designer tool (see ti.com, it's in the lower right corner)
to design the passives needed. It'll even give you layouts.

You'll need Adobe Flash to run it, but it'll give you schematics, a BOM,
layout recommendations and a cost estimate.

I'd think a wide input range (6 to 64 V DC, for example) cape
shouldn't cost more than $30 to $40, retail.

-Andrew

There is a reason I ALWAYS use modular switchers.
The NRE is already done for me and so is the CSA, UL, etc approvals.

Not to mention the fact that I have been on enough switching supply design
teams to know
what a sink hole for $$$$ designing your own can be.

I'm not a "hobbyist" (no insult intended, just stating the facts) and I
don't work for free.
Texas Instruments has spent hundreds if not thousands of hours ($$) tweaking
and testing the design.

With enough orders, I could probably get it under the $100 mark, but never
under the $50.00 mark.
All I can say is: How many Beagle Bones would you like to smoke before $100
sounds cheap ??

(Bill)

We’re simply giving you an idea of what we feel the market (being this list) will bear. Most of us here don’t need the CSA, UL, etc. approvals. Many if not most work with the beagle platform(s) because it’s something we like doing not because someone pays us to do so (gumstix fits that bill, beagle is not intended as a commercial product) Ultimately, the output of something like TI webbench is good enough and I trust that TI has a reputation to protect in providing good tools that output good designs which I then verify with the testing I deem appropriate for my needs. Sure there are advantages and places where using a module is a good idea, but within the community that is beagle, I can’t see the doubling of cost being justified.

Eric

There is a reason I ALWAYS use modular switchers.
The NRE is already done for me and so is the CSA, UL, etc approvals.

I don't believe anyone who's buying capes expects any kind of
certifications. At least not me. FCC compliance might be nice, but
even that seems optional for beagle accessories not put out by
CircuitCo.

Not to mention the fact that I have been on enough switching supply
design teams to know
what a sink hole for $$$$ designing your own can be.

True, but you can't get low cost without designing your own. That's
more of what I'm getting at.

I'm not a "hobbyist" (no insult intended, just stating the facts) and
I don't work for free.
Texas Instruments has spent hundreds if not thousands of hours ($$)
tweaking and testing the design.

They've also spent quite a lot of time with the WEBENCH tool (which was
a National tool before TI bought National). The power supply guys at
TI / National / Unitrode know what they're doing. They want to sell
chips, they won't give you bad designs. Although you can go cheaper if
you don't follow all their recommendations.

With enough orders, I could probably get it under the $100 mark, but
never under the $50.00 mark.
All I can say is: How many Beagle Bones would you like to smoke
before $100 sounds cheap ??

I think you're going to have a hard time selling these at that kind of
price.

To compare, most people are buying $10 to $15 wall warts to power their
beagles. Trying to sell a power supply at $100 isn't going to get many
customers. I'm not trying to dissuade you from making this, just that
your price target is much too high and that I think if you didn't use a
module that you'd hit a much nicer price target.

Regards,
Andrew

Eric;

Thanks for the input / advice.

I appreciate it.

As the saying goes: “It’s free to ask.” J

Now I know …

Bill, you replied only to me, I'm replying to you and the beagle list.

You Guys are right.

You could probably stick a 7805 and a REALLY BIG heat-sink on a Beagle
Protoboard.
Just remember: 12V - 5V = 7V across the regulator. 7V x 2A = 14
Watts!

I suppose I could put together a circuit diagram for a cheap cape the
list users could build.
There are a number of cheaper 5V and 3.3V switchers out there as long
as you are willing to live with 1 amp.

Using TI's WEBENCH tool, the first design specs a BOM cost of $2.44 for
a switcher (decent efficiency) using 462 mm^2. That's pretty decent.
I ran the tool with voltage input from 7 to 28 V DC and output current
of 2 A at 5 V. The first result has a range of efficiencies from about
85 to 91% at 1 A output current.

BOM wise, that's a heck of a lot cheaper than your module. It's just a
straight buck regulator but that's all that's needed at this low power
level. With that kind of BOM cost, you'd be in the range of a low
volume $30 to $40 retail cape with a profit margin after paying for
assembly and PCB.

I believe WEBENCH assumes a volume order of about 1000 units for the BOM
cost, so if you're building less, multiply that by about 3 to get in
the ball-park of a 50 unit order price, at least for the BOM.
Then check out some of the lower cost 2 and 4 layer PCB manufacturing
services, like OSHPark or SparkFun's service, at least for prototyping.
This would be a really cool project to sell on tindie!

Regards,
Andrew

OOPS!

Hopefully this will go to the list now :frowning:

Actually I found a drop in module with the same pin out as a 7805 for about
$12.00 Qty 10 pricing.
But I'll also give the TI workbench a try too.

Hey, keep working at it. It’s a good idea, just gotta be in the cost range that people will buy. Just a thought… A large lithium pack for the bone would rock!

Eric

Yes, do keep working at it and feel free to bounce ideas off me and the
list. We'd love to help if we can. More capes == good! :slight_smile:

Regarding the lithium battery pack idea, it's a good one except for one
little issue with the bone where a rework is required in order to power
down into TPS65217 OFF or SLEEP modes when a battery is connected due
to SYS voltage staying present and the way 3.3V_EXP is generated. So
long as you don't want to power off with the battery still connected
there's no issue, though.

The existing battery cape doesn't have this problem as it's 4 AA cells
with a regulator and that feeds into the 5V line on the expansion
connectors.

-Andrew

Yeah Eric that would be cool.

There is one worry I have when selling into the “hobby” market, perhaps you, Andrew and Gerald have some thoughts on this.

The national sport in the US seems to be suing people for ‘damages’.

Whether it is someone suing Mac Donald’s over hot coffee or something else.

That is why I was going with CSA/UL modules. You see if I design a product that doesn’t have thermal shutdown or short circuit protection,

and someone uses it wrong, and it catches fire. Guess who will get sued ?

Same with a Lithium pack ….

There is one worry I have when selling into the "hobby" market,
perhaps you, Andrew and Gerald have some thoughts on this.

If you're selling to the "maker" market, I think that's less likely to
happen. People are going to mod your design and break it and it might
catch fire, but that's the risk when you are buying a product that's
without case and meant for modding.

The national sport in the US seems to be suing people for 'damages'.

Whether it is someone suing Mac Donald's over hot coffee or something
else.

You've chosen a bad example, but yes, getting sued is a concern. Do
your design assuming you would be able to pass UL but getting UL itself
doesn't buy you any protection from being sued. I could sue you even
though I have no idea who you are (granted I'd probably lose, but you
still might need to pay for a lawyer).

Regarding hot coffee, read more about that story. There's more to it
than it seems.

That is why I was going with CSA/UL modules. You see if I design a
product that doesn't have thermal shutdown or short circuit
protection,

and someone uses it wrong, and it catches fire. Guess who will get
sued ?

Even with UL mark, you are the one who gets sued. Regardless of if UL
will help you, you're on the hook still, at least till you prove it
wasn't your fault. It's a risk of being in business. Create an LLC or
other corporate entity to shield yourself and you will be fine. So
long as you don't intentionally make a product that will hurt people,
you're fine.

Same with a Lithium pack ..

There's more concerns with a battery, yes. But designing to pass UL
would be definitely recommended here. Paying for UL isn't worth the
money though, unless you're selling retail finished product to
customers who care about UL mark. Your customers won't care about UL
mark.

If you're concerned about getting sued, create a company that shields
you personally from liability as much as possible and buy some liability
insurance for your company.

-Andrew

Actually I have my own consulting company, so that is not so much of an
issue.
I am tempted to simply post a schematic, and let the 'builders' have at it.
If it turns out to be real popular, I can then consider kits or finished
product .

If you want low risk, offer a kit with a PCB and a bag of parts. Then
you're off the hook for liability (for the most part) too since the
user has to assemble it and thus takes on responsibility of when it
catches on fire :slight_smile:

If you just post a schematic, I don't think it'll go anywhere. If you
post up manufacturing files for the PCB, it might get some people
interested as they could simply use one of the low cost proto houses
for the PCB. Best bet is to get some bare boards made and offer those
up for sale with the schematic.

Good luck! :slight_smile:

-Andrew

Good ideas.
When I order boards for my commercial build, I may just order some extras :slight_smile:

Hey, keep working at it. It's a good idea, just gotta be in the cost
range that people will buy. Just a thought.... A large lithium pack
for the bone would rock!

I agree, I'd love a large lithium pack for the Beaglebone.

Yes, do keep working at it and feel free to bounce ideas off me and the
list. We'd love to help if we can. More capes == good! :slight_smile:

Regarding the lithium battery pack idea, it's a good one except for one
little issue with the bone where a rework is required in order to power
down into TPS65217 OFF or SLEEP modes when a battery is connected due
to SYS voltage staying present and the way 3.3V_EXP is generated. So
long as you don't want to power off with the battery still connected
there's no issue, though.

So I can't put a lithium battery powered Beaglebone to sleep unless I
disconnect it from the battery?

- Grant