Well, their tests at full load look okay.

Only 11mV peak-to-peak ripple on the 5 volt rail at full load, 34 on the 12 volt rail at full load, 14 for the -5 volt rail.
The pictured unit looks like it has fairly large filter caps which is a plus.
The specification lists + or - 2% on the 5v rail, which looks good, although it's +/-6% on the 12v which isn't so important anyway.

I also found another model: https://www.meanwell-web.com/en-gb/closed-fra … -output-rq--65b
This one has both -12v and -5v rails and has similar specs.

p166s wrote:

Well, their tests at full load look okay. […]
Show full quote

Well, their tests at full load look okay.

Only 11mV peak-to-peak ripple on the 5 volt rail at full load, 34 on the 12 volt rail at full load, 14 for the -5 volt rail.
The pictured unit looks like it has fairly large filter caps which is a plus.
The specification lists + or - 2% on the 5v rail, which looks good, although it's +/-6% on the 12v which isn't so important anyway.

I also found another model: https://www.meanwell-web.com/en-gb/closed-fra … -output-rq--65b
This one has both -12v and -5v rails and has similar specs.

In other words. Meanwell makes good PSU's for something like Amiga that require well regulated 5volt. Perhaps installed in BBC micro's as well. I have been looking at both the RT-65b and the RT-125b too.

IME, MeanWell supplies are generally well built. 2% and 6% regulation quoted above is well beyond the ATX requirement of 5% and 10%. There's always a compromise between design goals like transient response, stability, overcurrent protection, etc., so going too far on super tight regulation requires a trade-off in other areas -- which may not yield a better supply.

I'm not familiar with Amiga or any special design considerations, but I can't imagine it's too much different than any other computer that uses the 5V rail to supply the logic ICs directly (vs. stepping down the 12V rail, or stepping 5V down to 3.3V.) It may not tolerate poor regulation, but unless there's something really unique about them, I would think 2% is more than consistent enough.

SirNickity wrote:

IME, MeanWell supplies are generally well built. 2% and 6% regulation quoted above is well beyond the ATX requirement of 5% and 10%. There's always a compromise between design goals like transient response, stability, overcurrent protection, etc., so going too far on super tight regulation requires a trade-off in other areas -- which may not yield a better supply.

I'm not familiar with Amiga or any special design considerations, but I can't imagine it's too much different than any other computer that uses the 5V rail to supply the logic ICs directly (vs. stepping down the 12V rail, or stepping 5V down to 3.3V.) It may not tolerate poor regulation, but unless there's something really unique about them, I would think 2% is more than consistent enough.

Thanks....
This is good to know, especially when I am looking for replacements for old 8/16-bit non-PC computers. You know. Those that are all-in-one-keyboard machines. They are really sensitive to voltage levels, so I am after something that can deliver as close to the ratings as possible. Even under load. Now they do not really use that much power, so I might be happy enough with the 65 model.

I am using a RT-125A to power my arcade boards and I am really pleased with it , I haven't had any issues so far
https://www.tme.eu/en/details/rt-125a/built-i … lies/mean-well/

keropi wrote:

I am using a RT-125A to power my arcade boards and I am really pleased with it , I haven't had any issues so far
https://www.tme.eu/en/details/rt-125a/built-i … lies/mean-well/

Cool... Another thumb up for MeanWell. I like it....

Yes it is really good and it stays cool.
I was using a DELTA psu before that (to it's defense it was only 9A at 5v and meant to be used with a heavy-duty unix-based ethernet switch) and it was getting warm after a while, the meanwell on the other hand stays cool and has provision to fix it to an extra heatsink if needed.
The electrolytic caps on the RT125-A are Nippon-Chemicon (input ones) and the rest Rubycon.

keropi wrote:

Yes it is really good and it stays cool.
I was using a DELTA psu before that (to it's defense it was only 9A at 5v and meant to be used with a heavy-duty unix-based ethernet switch) and it was getting warm after a while, the meanwell on the other hand stays cool and has provision to fix it to an extra heatsink if needed.
The electrolytic caps on the RT125-A are Nippon-Chemicon (input ones) and the rest Rubycon.

In other words. They will last a long time under heavy use, and even longer if only powered on a couple of times each month?

I guess so.
I have also thought about using them to power amigas (maybe inside the original psu shell) they really are good psus in my book.