Re: [PDXHS] Power supply project: Anybody interested?

[Peter Torelli]

Good idea for a project.  The USB interface is interesting.  I've built simple switching supplies (single phase VRs & converters), and lots of regulator chips have an oversupply current detect/shutoff through a very low value resistor (like 0.15ohm).  Since I often need measure power, I'd like to have a USB output that samples the current and voltage AT the regulator so that I could compute power more accurately and log it over time.  Usually I just add a low value resistor to the supply line and probe it with a scope, but having the info out of the supply's monitor would be much more accurate. 

However, making the USB stack work is way outside my knowledge, and seems like such a HUGE learning curve that I always just move on to something else.   There's nothing on "instructables.com" showing how to do this, so if you're taking requests....

On Sun, Oct 7, 2012 at 10:38 PM, Ben Porter <xenov...@gmail.com> wrote:

Hello Hardware Hackers,

The BrainSilo hackerspace is a bit short on variable power supplies, so I'm endeavoring to build one (or more).  I figured I'd ping the community to see if there are others interested in building a power supply for themselves (to keep or donate to the hackerspace).  I could obtain parts & host a build session.

Regarding functionality, here's what I'm thinking:

* 0-24V DC variable

* 0-2 amps (adjustable current limiting)

* 7 seg/LCD display?

* computer connectivity (USB)?

* multiple outputs?

* high voltage outputs (muahahaha...) ?

Chime in if you're interested or have ideas on what else would be useful in a power supply.

- Ben

[Ben Porter]

USB is pretty easy, I've done several USB enabled projects before.  One simple route is an Atmel USB enabled MCU (I.E, at90usb162) + the LUFA USB framework.  Connecting the power supply to a PC enables some cool use cases - not only power tracking, but also control of the supply to make it into a low-frequency function generator or programmed shut-off in certain conditions (I.E, when charging batteries).

[Michael Shiloh]

this could also be a universal charger and programmable load. wish i
were in pdx...

[Michael Shiloh]

if you make the microcontroller an Atmel ATmega328 (or other
appropriate), you can use the Arduino IDE to program and take advantage
of all those libraries (like PWM).

On 10/09/2012 02:21 PM, Pavel wrote:
> I would be interested in building a couple of these.
>
> A feature I'd like to see is a merging of certain function generator
> abilities. Often I need pulsed or PWM'd power, so it would be great to have
> that built in.
>
> What sort of data can we send over USB? It would be interesting to have the
> ability to log things like current draw over time, etc.
>

[Peter Torelli]

Not to hijack this thread, but this brings up another opportunity for a project:

I'd pay a fee for a course on the software architecture for connecting an Arduino to Windows via USB.

I haven't written a win32 app since 1997, and attempting to use CygWin and windows hardware makes me convulse.  I figure someone must have solved this for CygWin (or MS Studio, which I also haven't used since 1997).  I ended up using the Arduino IDE interface and serial comm + ctrl-v/ctrl-c to download a 1Mb flash project I did.  It worked but it sucked.

Things I'd like to learn:

- How to scan USB to tell what kind of arduino is there

- Serial TX/RX programming as a term (AT x commands)

- Other interfacing protocols: eg. USB2.0 commands for data transfer

- Simple windows "shell" / "demo" code to implement these interfaces

- How to build them under Windows or CygWin (using MS Studio, Cygwin, or Eclipse... probably Eclipse since it works on Linux and Windows)

- a test app to verify tx/rx round-trip between the app and the target arduino

The web is not friendly to this activity for noobs... I need some handholding and would pay a nominal fee for a half-day hands-on transfer of knowledge.

-p-

[Aaron Eiche]

I'll throw my hat in. I'd like to do one as well.

On Oct 9, 2012 6:10 PM, "Noor" <noororber....@gmail.com> wrote:

I'm interested in any project I can learn something from.

If anything is going on I'd love to be a fly on the wall and learn whatever I can.

I learn best hands on, so I'll be there to assist in anyway I can.

Keep me posted.

-Noor

[Ben Porter]

Here's some of the parts I've scoped out for the power supply:

TL494CDG4: pwm controller

Si8234AB-C-IS: mosfet driver

MAX9918ASA+: current sense amplifier (for current feedback)

NTD5867NLT4G: power mosfet

Specifications I'm shooting for: 3 amps continuously @ < 10 volts, 1 amp continuously up to 24 volts.  The reduction in current is a limitation of the transformer, and if a beefy enough transformer were used, it would be possible to do 3 amps at 24 volts.  Transformers of this power are pretty spendy, but can be salvaged from old electronics for cheap.

With the transfomer (120 volts AC to 24 volts AC), some passives & other various parts and the price for power electronics alone will be around $20-25.  A USB enabled microcontroller & supporting HW will probably be another $5-10.  Add in 7-seg displays, a fancy case, knobs & cables for ~ another $20.

I'm designing the supply so the microcontroller can be added later optionally (it's only needed if you want to monitor/control the supply via USB or power some cool display).

So does the price point & specs sound good?  I can work on lowering the cost if that's an issue.  As far as timeline goes, I should have something designed by next weekend, prototyped next week, and ready for others to put together around the end of October.  Also, if anyone wants to be involved in the design/prototyping work, let me know!  I'll probably be at the space this Thursday hacking something together.

- Ben

[Justin Dormandy]

I just wanted to say I would be really interested in doing this with a group of people. Power supplies are essential for molecular biology work and right now I'm limited to an electrophoresis gel box with preset voltages. Depending on the application with electrophoresis you might want a voltage of 0.25 volts/cm to 5 volts/cm. So if the electrodes are 20cm apart you might need 5V to 100V. Usually it would probably be closer to 100 for 20cm. You also generally want to be able to choose between either constant amperage or constant voltage. I'm not sure if you would ever need to go above about 30mA.

Ok so in that example (you'll have to forgive if any of this doesn't make any sense. Its been quite a while since I've dabbled with electronics and electricity) that would be 3 watts at 100V. So I would want to give myself room for other applications that might require much higher amperage. So would it be doable to make a 100 watt power supply with lets say a max output of 300V?

Another reason I would be very interested in doing this is because in biotech you use a lot of equipment but don't understand how it works. A lot of people might not even understand the principles behind their equipment and which I think something is lost there. I've also had an arduino controller for too long and would be very interested in working on a project that used it - or any micro controller for that matter.

Would anyone be interested in giving a class or discussion on the basic concepts behind every step we might take to make a power supply, instead of just instructions on how to build it. I could definitely use a refresher on basic electronics and electricity - just to make sure I'm remembering general electrical concepts right. Then anything on basic electronic components and circuits would definitely help me out as I have forgotten a lot there. Then if instead of just copying and pasting the code into the micro crontroller I think it would be incredibly beneficial if anyone was interested in going over the basics of how to program it.

What do you all think?

-Justin

[Ben Porter]

I think it would be a great idea to work some electronics theory/math into the project.  It's like that one guy said (Feynman): "What I cannot create, I do not understand" or whatever.  Of course, if you're just creating from a script you won't really understand it either - so there should be some instruction as to how things are actually working.

I do envision this thing being built as a group project(s), so I'll remember to throw in some of the concepts behind the design too.  If anybody's up for it, I can even give a short talk on electronics this Thursday at the space (so maybe it will sink in by the time we actually get to building the power supply :)).

[Aaron Eiche]

I think I'm likewise interested in "A talk on electronics"

On Tue, Oct 16, 2012 at 9:05 PM, BioDude(Justin)
<justin....@gmail.com> wrote:
>> If anybody's up for it, I can even give a short talk on electronics this
>> Thursday at the space (so maybe it will sink in by the time we actually get
>> to building the power supply :)).
>
>

> That would be great! I also have an electronics learning kit in my locker at
> BrainSilo that anyone who wants to could make some circuits with. You would
> know better than me, but it might be a good way to demo some of the
> electronics concepts. For example it comes with a small transformer. Anyone
> who wants to experiment making some circuits with it would be more than
> welcome. I have several electrical and electronics books that have been
> collecting dust for years, but maybe I could make copies of pages that would
> be most relevant to your discussion. That way people would have something to
> take home with them they could actually keep as reference and study for
> understanding.