| Author |
Message |
|
|
Posted: Feb 26, 2010 - 02:16 PM |
|

Joined: Feb 26, 2010
Posts: 10
|
|
Hi,
I understand that we can provide 5V power to the AVR chip by using e.g. a 9V wall supply through a 7805 chip and supporting capacitors etc.
However, if I have a high quality handphone charger, e.g. original Samsung or Motorola chargers which states itself as 5V, do I still need the 7805 circuitry?
If possible, how do I test if the voltage is indeed "safe" for the avr chip? |
|
|
| |
|
|
|
|
|
Posted: Feb 26, 2010 - 02:19 PM |
|

Joined: Nov 02, 2009
Posts: 530
Location: Zelenograd, Russia
|
|
Just measure the output voltage of your "high quality handphone charger" BEFORE applying it to your AVR - I bet you'll get surprised  |
|
|
| |
|
|
|
|
|
Posted: Feb 26, 2010 - 02:33 PM |
|

Joined: Jul 13, 2008
Posts: 204
Location: Pennsylvania
|
|
|
|
|
|
|
Posted: Feb 26, 2010 - 02:41 PM |
|


Joined: Nov 01, 2005
Posts: 4906
Location: Hilversum - the Netherlands
|
|
|
|
|
|
|
Posted: Feb 26, 2010 - 04:06 PM |
|

Joined: Feb 26, 2010
Posts: 10
|
|
Thanks!
Which means it'd still be advisable to get a linear voltage regulator and not cut corners.. hmm
However if in the case I'm running a series of 17 normals LEDs each drawing 30mA for a total of 510mA, and they're drawing power from the AVR chip, will it overheat the (regular) 7805 chip too much with an input voltage of 9V?
I am actually trying out a binary clock, and will hope it to run 24/7.
Sorry if it sounds off-topic, I am pretty new to this and if required I will promptly move this topic elsewhere.
Thanks!
P.S. If using the 7805 and AVR to run everything is not feasible, are there any obvious solutions? |
|
|
| |
|
|
|
|
|
Posted: Feb 26, 2010 - 04:17 PM |
|

Joined: Dec 15, 2005
Posts: 532
Location: Bristol, UK
|
|
Although new to this, you need to get to be able to read and understand data sheets...
Power dissipated in regulator = 4 x 0.51 watts = 2.04 watts - yes it will get extreeemely hot without a good heatsink. Look at the 7805 data sheet for the thermal characteristics of a 7805 (a little bit complex to understand - Google can probably find somewhere where someone explains how to deal with the thermal characteristics of a regulator...)
510ma through the AVR will almost certainly destroy it - see the Electrical Characteristics section for your AVR. There is normally a limit to:
- amount of current for the whole chip
- amount of current for each output pin
I'll leave others to describe alternative solutions... |
|
|
| |
|
|
|
|
|
Posted: Feb 26, 2010 - 07:47 PM |
|

Joined: Jul 13, 2008
Posts: 204
Location: Pennsylvania
|
|
|
|
|
|
|
Posted: Feb 26, 2010 - 09:48 PM |
|


Joined: Nov 01, 2005
Posts: 4906
Location: Hilversum - the Netherlands
|
|
If you check the outputvoltage with a voltmeter as suggested by MBedder, and the result is OK, you *can* use that mobile phone charger for your final, enclosed clock, but with a big BUT during development: when everything is open on the table, you need to be aware of the leakage current mentioned in the thread posted earlier. Also note the voltages that come as a bonus with that leakage.
To get rid of those, connect the groundlead (or negative) of the secundary side of the powersupply, to safety ground of the mains.
If you are unsure about what I wrote, do use a traditional transformer-based wallwart .... at least during development. |
_________________ Problems with Parallel Port Programmers ? Read this:
http://www.aplomb.nl/TechStuff/PPPD/PPPD%20English.html
|
| |
|
|
|
|
|
Posted: Feb 26, 2010 - 10:05 PM |
|


Joined: Feb 19, 2001
Posts: 18830
Location: Wisconsin USA
|
|
|
Quote:
If you check the outputvoltage with a voltmeter as suggested by MBedder, and the result is OK,
And be sure to check it with no-load, minimal app load (e.g., AVR in reset and no power suckers active), normal load, and max load. A lightly-loaded unregulated wallwart can have way-high voltages.
For demo rigs and the like, we keep a small stock of >>regulated<< wall-warts on hand in common voltages such as 5, 12, and 24 VDC.
Lee |
|
|
| |
|
|
|
|
|
Posted: Feb 28, 2010 - 12:09 PM |
|

Joined: Feb 26, 2010
Posts: 10
|
|
|
|
|
|
|