5
\$\begingroup\$

Assumption: I think it's generally true that commercially available AC/DC wall-worts can be expected to be pretty noisy (e.g. you can't count on much better than 5% stability).

Background: I'm dorking around with an RF application right now, and I hypothesize that some significant subset of my headaches is coming from powering my RF receiver with 5VDC coming directly from an AC/DC wall-wart (i.e. no RF choke, no regulator, nothing).

Question: What's the simplest way (e.g. fewest parts) and cheapest mechanism for converting the "angry" 5VDC off a wall-wort into a "happy" 5VDC for my RF receiver? I think the other subsystems are more tolerant to noisy power.

Aside: I wish I could give more definitive characterization of the noise coming out of the wall-wart, but sadly I am sans o-scope.

Edit1: In response to the request for more detail on parts being used:

\$\endgroup\$
  • 1
    \$\begingroup\$ the amount of time you might end up spending trying to figure out a noisy power supply might offset the cost of picking up an adjustable power supply, or even a decent single voltage Meanwell (or similar) 5VDC power supply. That said, maybe an LM6805 or LP2954 will do the trick? \$\endgroup\$ – Dave Dec 29 '10 at 15:27
  • \$\begingroup\$ @Dave i want to be able to take this "out of the lab" and put it in my living room though :) \$\endgroup\$ – vicatcu Dec 29 '10 at 15:29
  • \$\begingroup\$ What are the characteristics of your "headaches"? Can you give us a better idea of what parts you're using? \$\endgroup\$ – W5VO Dec 29 '10 at 15:47
  • 2
    \$\begingroup\$ That module will work down to 3.5V. So maybe you try putting an LDO regulator in and making 4.0V. \$\endgroup\$ – markrages Dec 29 '10 at 18:09
  • 1
    \$\begingroup\$ Have you put probes on your power line to see what your actually voltage is as well as the noise on it? An unregulated 5v supply isn't usually actually 5v after you hook it up. It would probably be good to actually see whats going on. I realize you don't have a scope, but a DMM is better then nothing. \$\endgroup\$ – Kellenjb Dec 29 '10 at 18:15
5
\$\begingroup\$

I have never had noise problems with wall warts, but then again, I never trust power supplies either and always add at least minimal filtering on my boards.

I usually add a 10 - 47uF electrolytic in parallel with a 0.1uF cap at the power inlet "just in case." This has never failed to clean up any ripple or spikiness from wallwarts.

Is the power supply regulated? Have you verified that the voltage is really 5V?

\$\endgroup\$
  • \$\begingroup\$ I added a link to the datasheet for the wall-wart in the question. The datasheet says "Regulated output", but also says you can expect a 200mV Pk-Pk ripple (if you use 10uF + 0.1uF filter caps). \$\endgroup\$ – vicatcu Dec 29 '10 at 16:49
6
\$\begingroup\$

The way to do this with the fewest parts - the only way to get 5Vdc from less than 5V - is to use a switching regulator. There are a few options, based on how much current you need. All of these can have inputs from ~1V-5.5V:

  1. TI REG711: 1MHz; 40mVpp ripple; 50mA; regulation not listed.
    REG711 layout
  2. Linear LTC1754: 600kHz; 65mVpp ripple; 50mA; 4% regulation. Linear LTC1754 layout
  3. Maxim MAX1759: 1.5MHz; 60mVpp ripple; 100mA; regulation not listed.
    alt text
  4. National LM2750: 1.7MHz; 4mVpp-15mVpp ripple; 120mA; 4% regulation.
    alt text
  5. Microchip MCP1252/1253: 650kHz/1MHz; 50mVpp ripple; 120mA; 2% regulation.
    alt text

There are others that will give 140mA and more, but the linked RF widget doesn't need it, these were pretty cheap, simple because they're just charge pumps, take up little space and are entirely surface mount due to the high switch frequency. Those with MHz switch frequencies are less likely to interfere with your 315MHz-434MHz project.

Easiest way to do it, in my opinion, is to get a different wall-wart!

\$\endgroup\$
  • \$\begingroup\$ @tyblu get a different wallwart is a perfectly valid recommendation! Thanks for the detailed answer :). \$\endgroup\$ – vicatcu Dec 29 '10 at 17:35
  • \$\begingroup\$ "the only way to get 5Vdc from less than 5V - is to use a switching regulator" not precisely true, you could use a charge pump. \$\endgroup\$ – markrages Dec 29 '10 at 18:20
  • \$\begingroup\$ @markrages, I would call that a switcher, too... \$\endgroup\$ – tyblu Dec 29 '10 at 18:28
  • \$\begingroup\$ @tyblu how about a light bulb and solar cell arrangement, would you call that a switcher? \$\endgroup\$ – markrages Dec 29 '10 at 18:32
  • \$\begingroup\$ @markrages, ... I don't get it. Does that count as a charge pump? \$\endgroup\$ – tyblu Dec 29 '10 at 18:35
4
\$\begingroup\$

It's a switching wall wart. The easiest solution may be to just get a linear wall wart, or even just a transformer and regulate the voltage at your board. It's bulky, but 50/60Hz won't give you any RF problems.

Switching wall warts can be noisy, and there is little filtering you can do about it in terms of RF. All component filtering will do is filter conducted noise, not radiated, which is what could be your problem.

Also consider other sources of interference. I have an RF remote that won't work anywhere in my house when I run my dishwasher.

\$\endgroup\$
  • \$\begingroup\$ that's hilarious. I had a problem where TV channel 7 dropped out whenever the kitchen light was on. Then I realized that fluorescent tubes look like conducting rods when they're on, and they were close, and parallel to the rods of the antenna. Moved the antenna a little & problem solved. \$\endgroup\$ – Eric Nelson Sep 2 '18 at 5:31
2
\$\begingroup\$

An an alternative, turn an ATX power supply from an old computer into a benchtop power supply. I can't speak for how clean the power is (I imagine quite good) but it's a cool project!

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.