I've heard about this in many places, did some tests and the only problem I saw was that in some cases the retro console gets some interference, possibly from the switching circuits.
But better be safe than sorry, so I'm asking here: Is it true that switching mode power supplies can cause damage on retro consoles?
Some vendors around say that yes, using these kind of power supply can quickly damage retro consoles.
But the reasons, like more "pure" and stable power, free of spikes and fluctuations, just didn't convinced me, specially because these guys were trying to sell me a really overpriced non SMPS one.
Please, let me know your thoughts and considerations about this.
Thanks in advance.
In general, switching power supplies should be better in pretty much every aspect: Less power consumption during operation, less power consumption in standby, and more constant output voltage, and better precision on actually outputting the specified voltage.
The classic power supplies in the 8bit era were rather crappy and failed on all those aspects: They wasted energy, and I think even the DC supplies did have 50Hz/60Hz noise on them and and having a "9V power supply" in those days usually meant getting "around 7...11 volts".
Those old supplies worked only because the console contained internal 7805 voltage regulators for converting the dirty 9V to stable 5V. On the other hand, switching supplies are so good that you can remove/bypass the 7805 regulator, and directly use a 5V switching supply (that works fine, as long as you can make sure that the console is used only with the good new 5V supply and that nobody connects one of the old crappy 9V supplies to it).
It might depend on quality though. The 5V from desktop PC's is usually pretty good. I haven't yet tried to use the modern 5V usb battery chargers for driving 5V circuits directly, theoretically they might be optimal for that purpose, but I've heard that they don't work so well in all cases, maybe the cheaper ones are lacking proper capacitors. Or if you were trying to draw 2A from an 1A supply then they might output stuttering voltage because they can never fully charge their capacitors.
If you have a voltage meter (and best also oscilloscope) then you can check if the supply does really output what it should (best check voltage in both standby and operation).
The interference that you have noticied... I can't say if it's dangerous or merly cosmetic. Maybe you can solve that with extra capacitors, if it's noise leaking on the supply output? Or is it leaking elsewhere - eg. when you use two consoles, one with switching supply and one with older supply - do you see the interference on both? Or try difference switching supplies, maybe some are working better.
Btw. what is your setup? A NES console and/or other retro consoles? And are you using 5V directly? Or are you passing something like 9V or 12V through the consoles's built-in 7805 regulator?
The cheap ones are really bad for the most part and have potential to actually damage your console(s), with very dirty output that has spikes going way above the supposed output voltage. This all happens due to insufficient filtering of the output in the power supply (capacitors alone will not help, inductance is needed too) and it is high enough frequency that various parts along the line are not able to overcome it. Maybe only if you coiled the cable around a ferrite block with a handful of turns you get more positive results. 7805 and equiavlent are the only things preventing destruction and they have very poor PSRR in the higher frequencies where all the crap is on a SMPS and that's also why noise in video and/or sound is fairly common from my experience. PSRR is okay only at the low frequencies like 50/60Hz, perfect for old AC transformer and you only ever see issues if the bulk capacitor in the brick has failed (the big 1000µF+ one). You need a high quality power supply to get decent results, *original* PS2 slim brick for example is a good bet.
I have designed switching supplies (SMPSs) into several of our aircraft products where they had to be really quiet to avoid getting switching noise into the radios and into the entertainment electronics. Getting such good performance out of them is beyond the capabilities of mere electronics hobbyists; but you don't need anything nearly that quiet for a game console where 25 or 50 or 100mV of power-supply noise isn't going to hurt a thing. I have not tried many commercially available SMPSs, but I don't think you'll find 100mV peak-to-peak of noise on them. Since others here have thought it could be a problem, I'd check it on my oscilloscope, but I don't expect to see that much. If you check though, check under load; because SMPSs have a narrower range of output current over which they're well behaved. If you check without a load, you might not get a valid indication of what it's going to do with your normal load.
As far as USB power goes (since nocash mentioned it), I will comment that in the design of a recent product, I found that trying to get even 500mA out of it pulled the voltage from 5V down to hardly over 3.5V because of the resistance in a common, cheap, skinny, 36" USB cable and its connectors, and the 5V microcontroller I had planned to run on it went into brown-out reset when the rest of the circuit took so much current.
SMPSs can fail and damage other parts. Obviously that's not normal, but it becomes way more likely if lower-quality cheaper components are built into it. It's not a console, but I've repaired a lot of a particular Z80-based industrial control where the SMPS fails rather violently, because the earliest ones used a variety of capacitor brands (and temperature ratings). It might run 24/7 for 10-15 years, then bam. Maybe 5% if the time it will damage other parts of the circuit, but it's usually contained to the power supply. And by failing violently, there's a TO220 FET that can heat up enough to leave the board black and charred through the board, exposing the fiberglass layer. One case where it self-heated enough to desolder itself by gravity, just fell off the board. I guess that's just the failure mode that particular chip had if the caps are bad. (For electrolytic, Nichicon HE and Panasonic FC are what I usually go to first)
Not trying it make it a complete horror story though, haha. I think a good quality SMPS would be fine, but if I was selecting among random unknown power modules, I'd try not to go by price. You might find some bad ones that are marked up, but you're less likely to find a good one that's marked down.
There's a wide range of switching frequencies, the older ones in the khz range and the modern ones in the mhz range. I'm sure that would affect the type of interference, I'm curious if there's a range that would be better, or worse.
The commodore 64 PSU is a Linear regulator and they are death bricks, the regulator breaks, and then shorts sending 12-18V straight into the motherboard.. the 128 fixed this with a switching power supply they fail and don't (usually) take the 128 with it.. finding a 128 psu is hard as all the 64 users nicked them over the years.. so I would say ditch the old PSUs and use modern Switching PSUs whenever possible, a cheap one from today is going to out do a "good" one from back then easily.
Thanks guys.
That was more or less what I imagined.
It's great to have this confirmed by more experienced peole.
nocash wrote:
The interference that you have noticied... I can't say if it's dangerous or merly cosmetic.
I think the interference is merely cosmetic, since it only appears when the console is plugged through RF cable, wich gets very close to the said power supply. Altough it's annoying it's possible to play and maybe even live with it.
Also SMPSs are replacing the linear ones, wich proves that they're not the future, but the present.
As always the rule of thumb is: good quality is always the way to go.
This is a good article to read on the matter :
http://sound-au.com/articles/external-psu.htm