Improving a wall wart's performance

Thread Starter

cmartinez

Joined Jan 17, 2007
8,218
I like to use ordinary 12VDC wall warts for most of my projects. Such as this one, which is capable of delivering up to 2.5 amps:

eliminador-regulado-de-3-a-12-vcc-2-5-a-con-puntas-intercambiables.jpg
What I like to do is to open it, extract its main board, and incorporate it into my designs. That way things not only look more elegant, but one ends up without having to plug the cumbersome thing into a power outlet, avoiding the nuisance of its getting in the way of neighboring receptacles.

e78db801-eac0-4532-a983-59960f435109.jpg

This thing is, however, quite prone to transients. So what I normally do is add an inductor at its output, and a big fat cap, as shown in the following schematic.


1583949141115.png
When I started adding said components to my circuits, glitches like sporadic resets and unstable input signals from encoders and sensors almost went away. So I decided to add yet another 4,700 µF cap in parallel to the previous one to see if things improved even further. And sure enough, they did. (As a side note, the average circuit draws about 3/4 of an amp at most).

Well, a few days ago, I used this same circuit in two boards that were simultaneously connected to a single 2.5A wall wart. I figured things would be perfectly safe, since together they wouldn't be drawing more than 1.5A at most. I was wrong.

When I powered up the wall wart, both circuits turned on for a couple of seconds, and then went back off again... then on again... off again... etc... Strangely enough, if I connected one board first, waited a couple of seconds, and then connected the second one, things would work just fine!

After a little thinking, here's what I concluded. Each board has one 50 µH inductors and two 4,700 µF caps in it. So the wall wart was trying to charge four fat caps totaling 18,800 µF when it was powered up ... My guess is that the wall wart has overcurrent protection in it, and the poor thing was struggling too much to charge those caps up during startup. And I confirmed this when I powered the circuits from a bench top power supply capable of delivering up to 30 amps without a hitch.

And now for my question: What would be the best way to have the wall wart charge those caps during startup without affecting the overall output voltage?



 

Attachments

Thread Starter

cmartinez

Joined Jan 17, 2007
8,218
A varistor? (https://en.wikipedia.org/wiki/Varistor). Resistance increases as voltage across it increases. Thus, during charging, it would have a higher resistance than after the caps were charged.

I have never designed a circuit with one, however.
That's a possibility ... although I was thinking more in terms of current... something that will allow a basic amount of current to flow through it (with as little resistance as possible), but will then limit it if exceeded. The goal being to use as much of the wall wart's current capacity during normal operation, and only limit it during startup.

Perhaps something similar to this?:

3d4709815f1dc108b201412655135bc8.jpg
I wouldn't need it to be precise, though...
 

KeithWalker

Joined Jul 10, 2017
3,063
Why did you choose such a large value for the filter capacitors? If the main intent is to suppress transients then surely much smaller ones with low ESR would do a better job and eliminate your charging problem.
Regards,
Keith
 

MrAl

Joined Jun 17, 2014
11,389
You do realize than when you take a wall wart apart like that you void the warranty right? (ha ha).

If you can dissect the circuit you may be able to add a slow start mechanism and it might be very easy such as adding a small capacitor. This is the way modern circuits work.
Can you get or create a schematic of your wall wart circuit?

Other than that, you could short out an added series resistor after maybe a second using a transistor or maybe even a relay if you dont mind that.
Note that an LM317 regulator for current limit could work, but it will drop some voltage so it wont be 12v anymore it might be only 10v on the output unless of course you mod the wall wart regulator circuit.
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
Is this a one-off or a least just a few?
What i have done over the years for this type of application, is to seek out mains transformer wall-warts, if you can find them, (they are getting scarce).
Usually flea market sales etc.
Remove the guts from the case and then board mount them, they generally come un-regulated, so a suitable 3 pin regulator is needed.
Max.
 

Thread Starter

cmartinez

Joined Jan 17, 2007
8,218
If the main intent is to suppress transients then surely much smaller ones with low ESR would do a better job and eliminate your charging problem
I always use 100 nF bypass caps in every chip in every one of my circuits. Plus I also add a few 10 µF here and there to make sure that transients within the circuit are taken care of. But sometimes my circuit's power supplies (in this case, the wall warts) are connected to, say, the same power line to which a 5 HP motor is also connected to. This causes very serious transients during motor startup, and those fat caps have been working great ever since I started using them for this purpose.


Note that an LM317 regulator for current limit could work, but it will drop some voltage so it wont be 12v anymore it might be only 10v on the output unless of course you mod the wall wart regulator circuit.
Sorry, but a 2V drop would be unacceptable. I use 12V inductive sensors connected to most of my circuits, and downgrading their supply to 10V would affect their perfromance.


Is this a one-off or a least just a few?
This is definitely not a one-off. This is something that I intend to implement in most of my projects in the future. Reliability is very important to me, especially since some of my customers reside overseas.
 
Last edited:

MaxHeadRoom

Joined Jul 18, 2013
28,617
Have you dismissed the idea of a linear supply mounted in the board itself?
For higher quantity designs I have used the small board mount power transformers from DigiKey e.g, for a linear supply,
Reliability was important to me also.
Max.
 

Thread Starter

cmartinez

Joined Jan 17, 2007
8,218
For higher quantity designs I have used the small board mount power transformers from DigiKey e.g, for a linear supply,
A small favor, Max. Can you point me to an example of one of those small transformers so that I can have a better understanding of what you're saying? Remember, all I need is 2 to 3 amps in the worst case scenario, tops.
 

MrAl

Joined Jun 17, 2014
11,389
Sorry, but a 2V drop would be unacceptable. I use 12V inductive sensors connected to most of my circuits, and downgrading their supply to 10V would affect their perfromance.
Yes, that it why i mentioned it. It was not my idea in the first place :)
 
Top