Post by thread:[EE] tapping power from PC supply

'[EE] tapping power from PC supply'

Appologies in advance for a very basic EE question...

What issues should I be aware of if I want to tap power directly from

the my PC's power supply to drive my breadboard(s)?  The PC provides a

very convenient source of regulated +5v and +12v, and is easily

obtained through one of the 4-pin drive connectors.  A cable could be

very easily made from sacrificed fan taps and 120v line power cable

(16-18 ga.)

I imagine the distance (approx 5 feet) from the regulator might have

some issues, but are they serious, and can they be overcome with a

large decoupling capacitor for the entire breadboard?  If so what

values would be apropriate?

-p.

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

> What issues should I be aware of if I want to tap power directly from

> the my PC's power supply to drive my breadboard(s)?  The PC provides a

> very convenient source of regulated +5v and +12v, and is easily

> obtained through one of the 4-pin drive connectors.  A cable could be

> very easily made from sacrificed fan taps and 120v line power cable

> (16-18 ga.)

The biggest problem will likely be when you accidentally short circuit it,

and it reboots yout computer.

Much better to get a separate power supply for prototyping.

Cheerful regards,

Bob

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

>

> Appologies in advance for a very basic EE question...

>

> What issues should I be aware of if I want to tap power directly from

> the my PC's power supply to drive my breadboard(s)?  The PC provides a

> very convenient source of regulated +5v and +12v, and is easily

> obtained through one of the 4-pin drive connectors.  A cable could be

> very easily made from sacrificed fan taps and 120v line power cable

> (16-18 ga.)

>

> I imagine the distance (approx 5 feet) from the regulator might have

> some issues, but are they serious, and can they be overcome with a

> large decoupling capacitor for the entire breadboard?  If so what

> values would be apropriate?

>

> -p.

Many PC switching supplies require a minimum load on +5V, +12V or sometimes 

both for the switcher to start "switching".  The information is sometimes 

found on a sticker on the supply or if you are lucky and it is a brand-name 

supply you can find a datasheet with Google.

The simple way to make the minimum load is with large power resistors across 

the appropriate rails.  Just remember P=I*I*R or P=(V*V)/R and size the 

resistors appropriately.  Also, don't touch them either because they will be 

hot.

Also, most PC supplies aren't that well regulated.  +/-5% to +/-10% and 

possibly worse.  You will also get a lot of switching noise and its 

amplitude and possibly its frequency will depend on how you load the supply.

I would recommend that you keep your eyes open for a good deal on a nice 

linear bench supply (+5V @ 1 to 5A, +15V @ 1A adjustable and -15V @ 1A 

adjustable is nice, plus some kind of current limiting control is wonderful) 

but in the short term you can use a PC supply.  Just remember it won't be 

the greatest and you may be able to toast a bagel on your dummy load 

resistors...

Rob Young

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

> > What issues should I be aware of if I want to tap power 

> directly from

> > the my PC's power supply to drive my breadboard(s)? 

*if* you want to do this take the 12V and run it trough an 78(L)05. Or

take the USB.

Wouter van Ooijen

-- -------------------------------------------

Van Ooijen Technische Informatica: http://www.voti.nl

consultancy, development, PICmicro products

docent Hogeschool van Utrecht: http://www.voti.nl/hvu

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

I wrote:

> >What issues should I be aware of if I want to tap power directly from

> >the my PC's power supply to drive my breadboard(s)?

Rob Young replies:

> Many PC switching supplies require a minimum load on +5V, +12V or sometimes 

> both for the switcher to start "switching".  The information is sometimes 

> found on a sticker on the supply or if you are lucky and it is a brand-name 

> supply you can find a datasheet with Google.

I'm aware of this, but it shouldn't be an issue since I'm already

driving the rest of the computer with the same supply.

> Also, most PC supplies aren't that well regulated.  +/-5% to +/-10% and 

> possibly worse.  You will also get a lot of switching noise and its 

> amplitude and possibly its frequency will depend on how you load the supply.

I measured the +5v line and I'm getting 5.09V so in this particular

case I think I'll be ok.  I would think that even a relatively small

capacitor should be able to filter out any ripple.

Bob Blick hit an issue that I completely missed: a short on my

breadboard will likely take the whole PC down.  That's probably enough

reason not to do this right there without some current-limiting facility.

> I would recommend that you keep your eyes open for a good deal on a nice 

> linear bench supply (+5V @ 1 to 5A, +15V @ 1A adjustable and -15V @ 1A 

> adjustable is nice, plus some kind of current limiting control is 

> wonderful) but in the short term you can use a PC supply.  Just remember it 

> won't be the greatest and you may be able to toast a bagel on your dummy 

> load resistors...

Yes, a benchtop power supply with multiple fixed and variable

voltages, adjustable current limiting, and current/voltage readout on

every channel sure would be nice, but something like that is *way* out

of my price range... or would it?

The biggest cost here would be the meters, but I'd be willing to bet

that I could use a PIC for all the voltage and current monitoring and

then send the output via serial to a PC.  Since the PIC is monitoring

current, it could probably be used as a cheap and easy current

limiter.  (Would this be fast enough?)  Variable output voltages might

even be dialed in right from the PC.  Hmmm... ponder, ponder...

If I were to build such a device, would there be any downside to using

a switching supply as the unregulated input?  That is, use the 12V

line from the PC supply to generate a clean, nicely regulated supply

4-11V, and the 5V line for voltages below 4V.  It would be silly to

purchase a PC power supply for such a purpose, but then again it would

be equally silly to purchase a big transformer, bridge, and filter

caps if I don't *really* need them.  I'd probably use a 6V halogen

bulb (or a pair of 6V and 12V bulbs, if necessary) as my dummy load,

mounted on a gooseneck as a worklight.

Thoughts?  Comments?  Suggestions?

-p.

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

-- snip ---

> Yes, a benchtop power supply with multiple fixed and variable

> voltages, adjustable current limiting, and current/voltage readout on

> every channel sure would be nice, but something like that is *way* out

> of my price range... or would it?

-- snip --

>

> Thoughts?  Comments?  Suggestions?

>

There was an issue of Circuit Cellar with plans for a multi-channel current 

limited supply.  A few years back.  Google may help or maybe the Circuit 

Cellar web site has a search feature.

As to "price range", consider how much you would be using it.  Good quality 

tools are ALWAYS worth what you pay for them.  You could buy a pretty nice 

B&K brand supply or some other non-name brand new for a reasonable amount. 

Used HP & Kepco supplies appear on Ebay regularly.

Rob Young

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

> There was an issue of Circuit Cellar with plans for a multi-channel current

> limited supply.  A few years back.  Google may help or maybe the Circuit

> Cellar web site has a search feature.

February 2002, Issue 139

"A Tracking Lab Power Supply" by R. Lacoste.

A good place to start, actually.  He starts from 220 instead of 110, but that's

easily remedied by a different transformer.

Two tracking 0-20V 3A sections and 1 0-20V 2A section.

Of course, the heretic didn't put a PIC in anywhere.  You'll have to add that,

as well as the Wifi and IEEE.1394.  ;-)

Mike H.

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

On Sep 24, 2004, at 3:28 PM, Peter Johansson wrote:

> Yes, a benchtop power supply with multiple fixed and variable

> voltages, adjustable current limiting, and current/voltage readout on

> every channel sure would be nice, but something like that is *way* out

> of my price range... or would it?

>

> The biggest cost here would be the meters, but I'd be willing to bet

> that I could use a PIC for all the voltage and current monitoring and

> then send the output via serial to a PC.

Or a cheap surplus LCD.  I've been contemplating a project like this;

but it's unlikely that it'll creep up on my to-do list - I already

have  a lab power supply.

What IS your budget?  Cheap VOMs are well under $10, so a wall-wart 

(free)

a couple 317s (cheap) and two dedicated meters would run you well under

$50.  But then, commercial units start around $60

     http://www.mpja.com/productview.asp?product=14600+PS

I like your idea of using the PC as a display.  I can see converting an

old (ie free or near free) PC into a whole electronics experimentor's

dream.  Power supply  (0-12V) mounted and accessible from a drive bay,

sound card used as a scope and signal generator, parallel port based

logic analyzer, and of course device programmers "everywhere", plus

assorted IDEs and layout SW.  Most of this already exists; putting

it together into an extensible package wouldn't be TOO hard (sorta

the "winavr" of hardware?)  Hmm...

BillW

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

Mike Hord writes:

> > There was an issue of Circuit Cellar with plans for a multi-channel current

> > limited supply.  A few years back.  Google may help or maybe the Circuit

> > Cellar web site has a search feature.

> 

> February 2002, Issue 139

> 

> "A Tracking Lab Power Supply" by R. Lacoste.

> 

> A good place to start, actually.  He starts from 220 instead of 110, but that's

> easily remedied by a different transformer.

> 

> Two tracking 0-20V 3A sections and 1 0-20V 2A section.

> 

> Of course, the heretic didn't put a PIC in anywhere.  You'll have to add that,

> as well as the Wifi and IEEE.1394.  ;-)

Then you should look at the follow-on to that article which appeared

in the December issue of that year.  Although it uses an Amtel MCU and

not a PIC.  ;-)

http://www.circuitcellar.com/library/print/1202/millier149/

This circuit is really much more what I'm looking for: multiple

outputs all at relatively low current.  While in this case the author

chose to replace multiple digital displays and multi-turn pots of

Lacoste's design with a rotary switch, MCU and single LCD, my idea

would be to take it a step further removing all display from the front

panel and moving it to a window on my PC.  It saves money, but that's

really only a small part.  The supply could then then can be small an

unobtrusive, or turned into a little piece of desk art, as I think I

might do.  I purged a lot of junk a while back, and one of the things

that went was some old stereo gear with some absolutely gorgeous heat

sinks.  I was *so* tempted to save them, but alas they went.  But I

bet it wouldn't be too hard to find more.  Anyways, polish them to a

mirror finish and mount fins-up on a small black box.  Get some retro

cloth-insulated wire for supply leads (they still make the stuff for

lamp restoration) for the finishing touch.  But if it had to have a

display, it would definitely have nixie tubes.  ;-)

Ok, pardon my mind wandering there... Where was I?  I'm not sure how

useful having all those displays active all the time really is.  It is

a good interface that allows us to easyly notice when there is a

problem, however we must poll it for information.  With software on a

PC, it would be trivial to add a little slider to set min and max safe

values on every parameter, and generate an audible alarm whenever one

of those conditions is exceeded.

Excessive in it's own sense, perhaps, but it's an interesting design

exploration, and a very good project to learn power supply design and

MCU interfacing all in one go.

But in the meantime, I'm happily tapping into my PCs power, regulating

off the 12v line with an old 7805...

-p.

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

On Fri, 24 Sep 2004 18:28:39 -0400, Peter Johansson wrote:

> Bob Blick hit an issue that I completely missed: a short on my

> breadboard will likely take the whole PC down.  That's probably enough

> reason not to do this right there without some current-limiting facility.

Indeed, because even a fairly modest PC PSU will be rated for about 30A on the +5V line, so your accidental 

short could be dissipating 150W without upsetting the power supply!  That's pretty bad news for your long 

wires, PCB tracks, and the little cloud of metal vapour that used to be the wire that caused the short!

Better to go for one of Wouter's ideas:  USB will supply up to 500mA at 5V, or use the +12V and regulate it 

yourself with a 7805.  The latter has overload protection so will stop any embarassing smoke-releasing...

Cheers,

Howard Winter

St.Albans, England

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

On Fri, 24 Sep 2004, Peter Johansson wrote:

> Appologies in advance for a very basic EE question...

>

> What issues should I be aware of if I want to tap power directly from

> the my PC's power supply to drive my breadboard(s)?  The PC provides a

> very convenient source of regulated +5v and +12v, and is easily

> obtained through one of the 4-pin drive connectors.  A cable could be

> very easily made from sacrificed fan taps and 120v line power cable

> (16-18 ga.)

>

> I imagine the distance (approx 5 feet) from the regulator might have

> some issues, but are they serious, and can they be overcome with a

> large decoupling capacitor for the entire breadboard?  If so what

> values would be apropriate?

Be very careful of wires and connectors and never leave it run unattended. 

The PC psu gives enough amps to be able to set cables on fire in minutes 

under the right conditions, although it is short circuit and overload 

proof.

Peter

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

> Bob Blick hit an issue that I completely missed: a short on my

> breadboard will likely take the whole PC down.  That's probably enough

> reason not to do this right there without some current-limiting facility.

Those PC power supplies go for a few bucks, and even cheaper if you take

them out of old decommissioned PCs. There's really no need to tap the

supply of the computer you're working with, right?

> Yes, a benchtop power supply with multiple fixed and variable

> voltages, adjustable current limiting, and current/voltage readout on

> every channel sure would be nice, but something like that is *way* out

> of my price range... or would it?

I once bought a bench top PS in a surplus store for IIRC $30. Not multiple,

but one good output (adjustable voltage and current limit and meters for

both) and a good place to start.

Gerhard

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

Ebay has bench PSUs on sale all the time. You can get a nice current and 

voltage adjustable one with a display for under $50 US

I bought one from this place, and its fine for what I need: 

http://www.mpja.com/listitems.asp?dept=2&main=1

You almost never need more than an amp for electronics/uC work. The one 

I got has 18V/3A, both adjustable. The only complaint I have is that the 

pots they used don't have very good adjustability. A fine-tune-knob 

adjustment for both voltage and current would be useful. Also, dual +- 

output is occasionally useful, but can be arranged from a single supply 

pretty easily.

Regards,

 Bob Monsen

Gerhard Fiedler wrote:

{Quote hidden}>>Bob Blick hit an issue that I completely missed: a short on my

>>breadboard will likely take the whole PC down.  That's probably enough

>>reason not to do this right there without some current-limiting facility.

>>    

>>

>

>Those PC power supplies go for a few bucks, and even cheaper if you take

>them out of old decommissioned PCs. There's really no need to tap the

>supply of the computer you're working with, right?

>

>  

>

>>Yes, a benchtop power supply with multiple fixed and variable

>>voltages, adjustable current limiting, and current/voltage readout on

>>every channel sure would be nice, but something like that is *way* out

>>of my price range... or would it?

>>    

>>

>

>I once bought a bench top PS in a surplus store for IIRC $30. Not multiple,

>but one good output (adjustable voltage and current limit and meters for

>both) and a good place to start.

>

>Gerhard

>_______________________________________________

>http://www.piclist.com

>View/change your membership options at

>http://mailman.mit.edu/mailman/listinfo/piclist

>

>  

>

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

PC supplies can be had very cheaply.

Tiger Direct has them for $15.

I have a box of them at my home.

I use them for all sorts of things.

Getting rid of noise can be accomplished

as mentioned earlier by dropping 12V to

5V.  You can easily put front ends on the

output to track or current limit or whatever.

Harbor Freight has sales where DMMs

are sold for $3-4.  Those could be used for

for voltage or current monitoring.

All controlled by a PIC of course.

http://www.tigerdirect.com/applications/SearchTools/item-details.asp? 

EdpNo=935375&CatId=1076

http://harborfreight.com/

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

Verrrrry interesting Idea.

Worthy of far more thought

----- Original Message ----- 

From: "William Chops Westfield" <spam_OUTwestfwTakeThisOuTmac.com>

To: "Microcontroller discussion list - Public." <.....piclistKILLspam@spam@mit.edu>

Sent: Saturday, September 25, 2004 3:56 PM

Subject: Re: [EE] tapping power from PC supply

{Quote hidden}> On Sep 24, 2004, at 3:28 PM, Peter Johansson wrote:

> 

> > Yes, a benchtop power supply with multiple fixed and variable

> > voltages, adjustable current limiting, and current/voltage readout on

> > every channel sure would be nice, but something like that is *way* out

> > of my price range... or would it?

> >

> > The biggest cost here would be the meters, but I'd be willing to bet

> > that I could use a PIC for all the voltage and current monitoring and

> > then send the output via serial to a PC.

> 

> Or a cheap surplus LCD.  I've been contemplating a project like this;

> but it's unlikely that it'll creep up on my to-do list - I already

> have  a lab power supply.

> 

> What IS your budget?  Cheap VOMs are well under $10, so a wall-wart 

> (free)

> a couple 317s (cheap) and two dedicated meters would run you well under

> $50.  But then, commercial units start around $60

> 

>      www.mpja.com/productview.asp?product=14600+PS

> 

> I like your idea of using the PC as a display.  I can see converting an

> old (ie free or near free) PC into a whole electronics experimentor's

> dream.  Power supply  (0-12V) mounted and accessible from a drive bay,

> sound card used as a scope and signal generator, parallel port based

> logic analyzer, and of course device programmers "everywhere", plus

> assorted IDEs and layout SW.  Most of this already exists; putting

> it together into an extensible package wouldn't be TOO hard (sorta

> the "winavr" of hardware?)  Hmm...

> 

> BillW

> 

> _______________________________________________

> http://www.piclist.com

> View/change your membership options at

> mailman.mit.edu/mailman/listinfo/piclist

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

Do it everyday...well....using a PC switcher for

general 12V, 3.3V, 5V....

Need to short across the enable pin and ground so the

supply will turn on.

__________________________________________________

Do You Yahoo!?

Tired of spam?  Yahoo! Mail has the best spam protection around 

http://mail.yahoo.com 

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist