• Please review our updated Terms and Rules here

Smoothing a cheap universal power supply

RickNel

Veteran Member
Joined
Apr 24, 2009
Messages
641
Location
Canberra, Australia
I bought a cheap "universal laptop power supply" online, to provide 24vdc to an Apple Studio Monitor through the DIY DVI-to-ADC adapter I have built. The video signal comes from a PowerMac G4 with DVI output.

It works but is electrically very noisy. Spurious emissions drown out AM broadcast completely for a receiver in the room. The PSU is a sealed unit and I don't want to bust it open.

The adapter uses a common ground for all signals, grounds and sheilds including the DC power that is fed to the monitor through the ADC cable along with the video and USB signals. In the cable itself, each signal uses separate twisted pair, and 24v power and return is separately shielded.

Question - should I put a smoothing capacitor between 24v+ and ground on the adapter? What value?

Maybe related: video passed through this adapter using the noisy PSU gives a completely clean picture on a 15" Studio Monitor, but on a 17" Studio Monitor I see individual pixel errors that are stable for a given screen, often spaced at regular intervals. They are not monitor pixel failures because they change with the signal pattern.

I'm wondering if this might be power-related, because the 17" will draw more power from the PSU. Or clock-related? If the adapter was introducing some interference to the TDMS because the adapter wires connecting the DVI to ADC make a ratsnest without pair-twist, I would expect to see the fault on both monitors.

Advice and/or experience?

Rick
 
Buy a "more expensive" universal laptop power supply. Just going through the pain of EMC testing with a product here, and I'm pretty sure that psu wouldn't pass here!

....or use a linear psu.
 
You could try putting the lines cord, psu, and power cord under a grounded metal bucket to see if the signal is leaking through the cabling, or propagating through the mains.
patscc
 
Buy a "more expensive" universal laptop power supply.

Even those can be very RF-noisy. My dad bought one of those to replace his Dell laptop's dead power supply, and it ruins the AM band on any radio in the house with buzzing and static when it's in use. He bought it in a big-name electronics store, too, so it wasn't a cheap knockoff Chinese eBay item.
 
The big-name electronics store PSU probably came from the same factory in China.
Is the DYI adapter in a shielded enclosure of any sort ?
Do you have access to a linear 24 VDC source, or two car batteries in series, just to test out if it is in fact the PSU that's causing the problem ?
patscc
 
Those generic Chinese power bricks are not all created equal. Some of them are built with similar quality as the name brand ones, while others... aren't so much. Sounds like you have one of the crappy ones.

The RF interference is probably due to a complete lack of any filtering of any kind. On the input side, there should usually be a choke, an X cap, and two Y caps. Your adapter probably has unpopulated spaces on the PCB instead. (And as a side note, if it's that cheap, it probably also has a fuse on the neutral side, no inrush current limiter or MOV, and a completely inadequate heatsink on the switching transistors and output rectifier... so don't ever leave it plugged in unattended.)

On the output side, you probably have one dreadfully undersized and crap quality filter cap and... well, that's about all really. But again, you probably have unpopulated spaces for a filter choke and a second filter cap.

The reason I know this, is because I have rebuilt several of those things. I have a stash of parts salvaged out of other PSUs, so I have a good supply of the components to make the necessary improvements with little or no cash out of pocket. Installing/upgrading the relevant components fixes the interference, and also gains you a much more stable output voltage. Of course, there's no way to fix it without actually getting inside the brick, but fortunately that's pretty easy to do, if you're inclined to tackle such a project. Most of 'em aren't actually sealed, they usually have a couple screws hidden under either the label or rubber feet. But the ones that are glued, are rarely glued very well... throw it in the freezer for awhile, then spend a few minutes squeezing around the perimeter at the seam, and they'll usually crack apart with little effort.
 
My reason for asking is that if you want an RFI-free PSU, get a linear supply. If you know the output requirements, it makes it easier to find one.
 
Old Thrashbarg said
But the ones that are glued, are rarely glued very well... throw it in the freezer for awhile, then spend a few minutes squeezing around the perimeter at the seam, and they'll usually crack apart with little effort.
Great tip. I can't wait to get home and try it out on some unsuspecting PSU. I have quite a few that have offended in me in one way or another...heh...
patscc
 
Nominal power requirement for the Studio Display is 25vdc 40w.

I've made up a small linear supply before, and maybe that's what I should do again. It will cost somewhat more than the laptop supply did. Correcting the deficiencies of the cheapy seems a little outside my technical knowledge. This one has selector for multiple output voltages 12-24vdc, so could be a bit more complex inside. It was sold as a laptop supply, and probably on the assumption that laptops have their own DC input filters. It works fine for laptops. That's why I was hoping I could filter the input to the DVI-ADC adapter rather than breaking into the PSU without knowing what is in there.

I've had another PSU with lots of RF emission, that came with an external USB HDD enclosure.

The noise is propagating through the cables as well as emitting from the box. AM interference gets louder when the DVI cables are plugged in. It's the noise in the cables that bothers me more than the radio interference, because I suspect it is affecting the image on the 17" monitor. My DVI > ADC adapter is completely unshielded, so far, but I plan to shield it once circuits are confirmed.

Rick
 
Is this something that cropped up only with the new brick ?
If not, you should try shielding the adapter, at least temporarily, since unshielded PCB's can throw off a lot of RF, just to rule out that it's not the adapter.
patscc
 
Linear is the way to go if you want a clean supply. Too bad you're not in the USA; there are still some 24V linear tabletop supplies available fairly cheaply.

If the EMI is bad through the supply cable, you might try looping the DC cable a few times through a ferrite toroid, if you have any kicking around.
 
If it's interference on AM ( MW ), wouldn't it have to be a pretty good-sized toroid ?
Anyone have any idea what switching frequency most cheap bricks these days use ?
patscc
 
Teardown of a $5 eBay counterfeit "Canon" power supply:

www.pcbheaven.com said:
Some years ago i bought myself a Canon PowerShot camera for my digital photos, which eventually became the official pcbheaven camera.

The camera needs 2 AA batteries to operate. I use of course rechargeable batteries. But then i decided to buy a power supply to have my camera always ready on the fly. As a matter of fact, i wanted to make one power supply instead (big deal), but the DC jack of the camera was very small and i did not really knew which one to get, so i decided to buy one instead.

First stop - ebay. I immediately found one for this camera, and bought it for 4.99 USD including shipping. This is ridiculously cheap!

When i received the item, it looked as a normal power supply. I immediately plugged it in. At first, i did not notice anything strange, only that some times when i was moving the jack the camera would turn off. It was not very annoying. But after some time, things became really strange.

Watch the following video:

 
If it's interference on AM ( MW ), wouldn't it have to be a pretty good-sized toroid ?
Anyone have any idea what switching frequency most cheap bricks these days use ?
patscc

Lotsa turns--nice way to take up the slack in a cord.

Usually somewhere around 30-40KHz--just out of hearing range of your dog.

I wonder if the Canon PSU is a genuine Canon product or if it's a Chinese counterfeit knockoff.
 
Last edited:
Okay, that's kinda what I thought, so if we're looking at AM (MW) roughly 0.5 Mhz to 1.6 Mhz, we're looking at like the 12th and higher harmonic of the switching frequency ? Really ? It's that cheap to be that noisy above the 12th harmonic ? Do they truly have no input or output filtering ? Anyone have any pics of one pulled apart ?
patscc
 
Okay, that's kinda what I thought, so if we're looking at AM (MW) roughly 0.5 Mhz to 1.6 Mhz, we're looking at like the 12th and higher harmonic of the switching frequency ? Really ? It's that cheap to be that noisy above the 12th harmonic ? Do they truly have no input or output filtering ? Anyone have any pics of one pulled apart ?
patscc
Just open a compact fluorescent that's on its way out; great AM radio noisemakers.
 
Back
Top