EEVblog Electronics Community Forum
EEVblog => EEVblog Specific => Topic started by: Russel on March 30, 2011, 09:15:42 pm
-
EEVblog #158 - AVR ISP MK2 + LM317 Regulator Tutorial
Great tutorial Dave!
I'm no electrical engineer. I've worked repairing electrical control systems for hydraulic machinery, but I've never spent much time experimenting with electronics, although I find it fascinating. In any case, the EEVblog has inspired me to experiment and learn more, if even just for the fun of it.
After watching the blog I got to thinking about some stuff I had laying around. A while back the 12V 'wallwart' power supply for my ethernet switch died. Since then I have been using a 12 volt linear power supply (12A) to feed the switch, a little overkill. I remembered that I had a 12V 700mA unregulated 'wallwart' power supply laying around somewhere and got to thinking that I might be able to add a regulator and use it to power my network switch.
The wallwart power supply measured 18V no load, and 14V with a 35 ohm resistor for load, it might work. So, after watching the blog, I bought a few parts and made up the circuit on a breadboard. The biggest question for me was whether the wall wart voltage would drop too low under load for the regulator to work effectively. The dropout voltage was going to be close, if it worked at all. I hooked up two multimeters and both channels of my DS1052E DSO to measure the input and output voltage and used a small variable lab power supply to feed the circuit.
I brought the circuit input voltage up to about 15V and adjusted the pot on the adj. leg of the regulator for 12V output. Then, ramping up the input voltage, the output begins to directly follow the input voltage from about ~9V (input) and appears to begins to regulate when the input voltage reaches ~13.5V for a dropout voltage of about 1.5V at room temp, no load. (I enjoyed using the math function on the Rigol to really show when the circuit began to regulate.)
OK, time to test the wallwart powering the circuit. Not too bad with no load, but the real question is what the dropout voltage is like with a load. So I use a 35 ohm 25W resistor to load the output... Woe! What the Heck! Both the input and output have a nice sawtooth waveform shape with an amplitude of about a volt. (At this point I forgot all about measuring the dropout.) I should have known from the first second I saw it that it was the wallwart, duh, the input waveform was just as sawtoothed as the output! But, seeing as I'm no EE and have to figure things out the hard way I replaced the 1µF cap across the output with a 100µF...no change. So, I loaded the wallwart directly with the 35 ohm resistor with the DSO connected and observed the nice sawtooth waveform. That is, of course, the point that I remembered seeing the sawtooth waveform on the circuit input and slapped my forehead with my hand. So much for using that wallwart for powering a network switch! Ah...well, maybe I can scrounge up another unregulated wall wart an d try again. Or, maybe just buy a transformer and bridge rectifier and build the whole power supply.
Well, I had lots of fun experimenting, even if it wasn't for anything important. One of the main points that I wanted to make is that although I'm not an EE, I really enjoyed the Blog and learned from it. Dave, there are plenty of us amateurs out there watching that appreciate this type of blog. Keep up the good work!
[Edit] I'm not sure why the ohm symbol came out as a question mark. I changed it to the word ohm.[end edit]
-
It's strange, I got the same original issue on my JTAG adapter. The JTAG Key takes the Vref from the target board to power it's input buffer and comparators, but for some boards, I don't have this available easily so I wanted the possibility of "fixing" it to a given value.
So I came up with this :
(http://www.246tnt.com/files/jtag_vref.jpg)
It takes the 3.3v regulated from inside the adapter and provides a vref.
The first switches selects between 3.3v 2.8v 1.8v.
And the second switches select between:
- Off (Vref supplied by the board)
- Full regulator (so, the voltage selected above)
- Diode Drop (basically the voltage above -0.3v, using a schottky diode drop to get 2.5v or 1.5v)
The PCB was made with the DorkPDX group order and costed me 2.5$ for 3 boards :) (it's 0.5 sq inch, 1in x 0.5in)
I currently omitted the protection polyfuse because I didn't order them yet ...
-
Thanks for the idea, Dave. I finally got around to modifying my AVR ISP II. I used an LDO 3.3V regulator and added indicators for the different modes. I documented the hack on my site: http://joesacher.com/electronics/enhancing-the-avr-isp-ii/ (http://joesacher.com/electronics/enhancing-the-avr-isp-ii/)