Taming a wallwart (plugin adaptor)

[PDubya]

Howdy!

I've got a couple 9v wall adaptors that don't appear to be regulated. Would a circuilt like the one I've sketched below work to get a regulated output of 5v if I replace R2 with something like a 740ohm fixed resistor?

Thanks!

 

[DirtyLude]

Well, a 9v wall-wart would give generally 9v. There used to be a decent page that described them and why you got really high voltages with no load, and such, but I can't find it now.

Anyway, if you specificly want 5v, you would be better off getting a fixed regulator like a 7805, rather than the adjustable regulator in the diagram. The there would be no need for the resistors.

EDIT: also, check the datasheets as these regulators usually require decoupling capacitors before and after them.

 

[philba]

well, do the math as given by the 317 datasheet. ditto on the 7805 and caps comments.

note that unregulated wall warts generally will produce the rated voltage at the rated current. This means that the unloaded voltage will be well above the rated voltage. Though there is some wishful thinking on the ratings.

 

[PDubya]

Here's the datasheet I'm using: https://www.electro-tech-online.com/custompdfs/2006/05/LM117-1.pdf

The "Typical" example has caps on the front and back ends, but they note they are only required "*Needed if device is more than 6 inches from filter capacitors." or "†Optional—improves transient response."

This is all new to me, so I could be completely reading this wrong.

While for this specific application I'm wanting 5v, if I'm ordering them, I'd rather just nab a handful and use them where needed. I assumed an adjustable would be more flexible over a variety of projects. Plus, I've got so many adaptors laying around, I figured this type of circuit would lend itself to being able to just nab an extra laying about and be able to use it.

Any suggestions would be appreciated!

 

[audioguru]

You are using a 240 ohm resistor that is used with a higher cost LM117. It has a lower operating current than an LM317. An LM317 with the 240 ohm resistor could have its output voltage rise when the circuit doesn't have a heavy load. Use a 120 ohm resistor instead with an LM317.

 

[PDubya]

Finally got my EWB cd's...

Using the 120, 360 would be the compliment for 5v right? Thanks again!

 

[audioguru]

The LM317 with the H suffix is in a small TO-39 case and will overheat and shutdown if its dissipation exceeds about 700mW. Therefore if your power adapter has an output voltage of 11V then the LM317H overheats with only 117mA of load current.

 

[Hero999]

The question is, how much current do you intend to draw?

100mA is fine for many low power applications.

 

[PDubya]

Audioguru >> The demo circuit has the AH package, but the part I was looking at Digikey was the "LM317TFS-ND". Unfortunately EWB didn't have anything that seemed similar to that part.

Hero999 >> The plan is to use most of them in color-cycling LED nightlights that can plug in using a wall adaptor. The reason I was wanting to go with the variable was that some of the light setups vary in the # of LED's, while still allowing the possiblity to use them in simple 5v low-power circuits. Just seemed silly not to order 25 with the price break.

Does the LM317TFS-ND from Digikey seem like the appropriate regulator for this type of use and flexibility?


Thanks again!!!

 

[audioguru]

The LM317TFS-ND is fine. If it gets hot then you can bolt a heatsink to it.
Your last schematic is fine.

 

[PDubya]

Thanks Audioguru!

Since you mention heat, maybe I'm going about this the wrong way. Since these will be kid-oriented nightlights, the last thing I want to worry about is something getting hot and catching fire or what not. Would you have a more elegant solution to powering a 4-6 LED light, with the occasional 555 to blink lights?

Thanks again, you all have been a great help!

 

[DirtyLude]

Why don't you just get a 5v DC adapter instead of the 9v one? There are plenty available.

 

[philba]

as long as it's regulated.

don't expect an unregulated "5V" WW to work. expecially if powering any kind of digital logic. but even just leds will need fussing. this is because the unloaded voltage with be well above 5V and, with load, will drop based on the current draw. If you assume that the WW is producing 5V and then figure out your resistors based on that, you may well be way off. For example, if the WW is producing 8V and your led has a Vf of 2V then you would be doubling the current through the LED from what you expected.

an unregulated WW (often called "pulsed DC") typically has a symbol on it - a dashed line above an unbroken one. It's just a transformer, rectifier and filter cap. Regulated WWs will say "regulated" on them, it it doesn't, presume it isn't.

To run a 5V regulator, I'd make sure I have at least 7V if the load is reasonably light and 8V otherwise. That's why you see a lot of 9V WWs out there.

 

[PDubya]

DirtyLude >> the idea is that I could just nab various 9-12v WW that I have laying about and be able to re-purpose them for little hobby projects.

Where do you even go to get a regulated wallwort that isn't considerably more expensive than an unregulated + the regulator circuit?

 

[Hero999]

as long as it's regulated.
an unregulated WW (often called "pulsed DC") typically has a symbol on it - a dashed line above an unbroken one. It's just a transformer, rectifier and filter cap. Regulated WWs will say "regulated" on them, it it doesn't, presume it isn't.

A pulsed DC adaptor doesn't have a filter capacitor, it's just a rectifier (often only half wave) and transformer but an unregulated DC adaptor does.

You can power an LED of an unregulated supply providing you calculate the series resistor using the unloaded voltage and I believe an astable 555 will also work but a monostable could have problems with false triggering.

 

[PDubya]

That was a big issue I had with one of the first circuits I had built. It was a variety of 555's blinking at varying speeds and offsets - but trying to use the adaptor just made everything blink, for lack of a better term, crazy.

I'm not sure how it would have handled just LED's, but anything timing related - it didn't seem to foot the bill.

 

[Hero999]

It sounds to me like the adaptor doesn't have a filter capacitor, either that or it's far too small, this can be a problem even if you're using a regulator. I recommend adding a 470uf to 4700uf capacitor before the regulator, the value will depend on the current you plan to draw.

 

[audioguru]

If the LEDs are white then they probably operate at 3.5V plus some voltage for a current-limiting resistor. Each LED might operate at 25mA so the total current with 6 LEDs is 150mA.
If the WW gives 11V then the LM317T will dissipate 0.9W and will be fairly warm, or adding a small heatsink will make it cooler.

 

[philba]

if it's just the white LEDs (i.e. no other elctronics to be powered) then I'd do 3 parallel strings of 2 each (7V drop), set the 317 for, say, 8V and 3 resistors for current limit (40 ohm). The regulator will run cooler for 2 reasons: 75mA current and about a 3V drop at the regulator (for .225 mW). No heat sink needed and you can use 1/8 W resistors, even...

In general, running multiple leds in series is a good idea if you have the voltage for it.

 

[philba]

A pulsed DC adaptor doesn't have a filter capacitor, it's just a rectifier (often only half wave) and transformer but an unregulated DC adaptor does.

You can power an LED of an unregulated supply providing you calculate the series resistor using the unloaded voltage and I believe an astable 555 will also work but a monostable could have problems with false triggering.

out of curriosity, where are you getting this information? I'm not saying you are wrong but have heard differently. I have also looked at a number of WWs output with a scope. The pulsed DC one I looked at clearly had some sort of filter cap in it and looked like it used a bridge rectifier (ripple was 120 hz).

I think using an unregulated WW can work but you have to factor in the actual load (i.e. know the droop). Also, I wouldn't count on temperature stability of the voltage. I think a VReg is a very reasonable and cheap solution to gain certainty. Seems like good engineering practice to me...