I recently spent hours reading about and researching the Amiga PSU situation. First, I firmly believe that Jens has written a lot of valuable posts about the current state of affairs and he overall comes closest to accurately describing potential issues. Love him or hate him, you can't accuse him of grifting or taking advantage of people's ignorance. He's an electrical engineer by trade, has designed tons of excellent hardware for the Amiga, his forums are open for everyone to read, he goes out of his way not only to help people but also the community. Finally, his engineering mindset, hypothesis forming and root-cause analysis skillset is top-notch. The man is a hardcore engineer at heart. As a disclaimer, I don't have any special relationship with him besides being a repeat customer, what I wrote should become obvious to anyone that simply spends 20-30 minutes reading posts in his english support
forum at icomp.de.
Now that said, here are the main issues with most widely sold MeanWell-based Amiga PSUs (Keelog, Electroware/c64psu, Retropower, A-Power etc) today:
1. There's excessive ripple (typically 80-100-120mV, Commodore specifies 50mV max). MeanWell RT-65B used to specify 50mV peak-to-peak ripple, but that changed as of a few years ago, and now none of their compact PSUs that I know of specify ripple below 80mV. Some folks at
eab.abime.net have come up with additional LC (inductor and capacitor) filtering to bring the ripple in-spec, but that's not something easily done by most people.
2. There's no automatic voltage drop compensation which can lead to undervoltage that can damage modern 3.3V ICs. Some MeanWell supplies allow the user to manually tweak the 5V output but that's not a good solution either: It's static and doesn't change with load but it can also throw the 12V rail out of spec. There are some custom modifications that add load to the 12V rail but again that's not something easily done by most people.
3. The 12V rail is not regulated on any MeanWell compact supply typically used for Amiga PSUs. This means that voltage fluctuations can be wild, leading to component damage.
4. Modern ATX PSUs do all their regulation on the 12V rail, so if there's minimal load there which is always going to be the case with an Amiga, the 5V rail output will not be stable. Older AT (and maybe older ATX) PSUs are better in that regard, but given their age, would need to be reworked.
Which brings us to the vendors of these so-called "Amiga PSUs". The most charitable characterization for them IMV would be tinkerers that are looking to make a quick buck. Maybe a more accurate characterization would be charlatans or fraudsters that are taking advantage of people's ignorance. All one has to do is read the misleading marketing that they put out on their websites: None of them admit that they're simply taking a dirt-cheap off-the-shelf MeanWell supply that doesn't even meet the Commodore specifications and wrap it in a plastic case with cables whilst charging 3-4 times the price. Besides the ludicrous margins, there's zero engineering here.
Yet Google search consistently puts them on the front page of results and people keep buying their products. Best case, if you're running an unexpanded Amiga with old ICs that are less sensitive to ripple and undervoltage, maybe you're willing to take the risk of using one of these supplies. But personally, I've spent thousands of dollars building Amigas with modern accelerators and other components and falling for this outright grift and putting my systems at risk is not something I'm willing to do.
If you don't want to buy Jens's CA-PSU (that some might see as over-engineered, I own several and couldn't be happier), Jens himself has said that recapping an original Commodore PSU is a good alternative. Moreover, investigating DC-DC solutions like PicoPSU could also be another alternative.
I'm mainly writing this to summarize what I put together after hours of research and help others. Most folks have the wrong idea re: evaluating a PSU, it's not simply a matter of using a multimeter to measure output. Even if you measure voltages while the Amiga is running (e.g. on the floppy connector) that doesn't show ripple and it doesn't show how the PSU behaves dynamically (e.g. when you insert a floppy, spinup a hard-drive or a CD, plug in a USB stick).
Some references:
[1] Jen's
slides for "Designing a better Amiga PSU" and video presentation
[2] Video
presentation of the above
[3]
Commodore PSU Specification
[4]
Commodore vs Electroware (ripple measurement under load)
[5]
Ian Stedman ATX PSU Guide (look at "minimum load" section)
