Morning all,
In my latest ST-related video, I repurpose a Mean Well industrial PSU for my shabby old STFM motherboard.
This means I've covered all three obvious PSU upgrades available to us. Recapping an old PSU, installing a new-build custom PSU and adapting an generic unit.
The playlist for these PSU escapades can be found below, featuring none other than our own Exxos supply in the second.
Cheers,
BW.
[Video] More ST PSU shenanigans
-
Badwolf
- Site sponsor

- Posts: 3043
- Joined: 19 Nov 2019 12:09
[Video] More ST PSU shenanigans
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
-
coonsgm
- Posts: 451
- Joined: 30 Jan 2021 01:30
Re: [Video] More ST PSU shenanigans
Great video! Watched and commented there already as well. I find the RD-35a is a better fit size-wise and also matches the original PSU output specs very well 5V @ 4A and 12V @ 1A. Thanks for your videos! I wish I'd had it as a reference when I started.
-
PhilC
- Moderator

- Posts: 7440
- Joined: 23 Mar 2018 20:22
Re: [Video] More ST PSU shenanigans
I've been wondering about having a single output 5V psu and then stepping it up to 12V with a cheap buck converter.
If it ain't broke, test it to Destruction.
-
exxos
- Site Admin

- Posts: 28344
- Joined: 16 Aug 2017 23:19
- Location: UK
Re: [Video] More ST PSU shenanigans
:bravo: :goodpost:
So about 50mV to 100mV isn't to bad for a "off-the-shelf" PSU really. The original Atari ones were more like 200mV area.
Such swithmodes don't really have detectable ripple because the actual transformer acts as a huge "choke" which basically eliminates ripple. But as you witness they become very spiky because that huge choke has a huge inductance where it its flux cannot change very quickly. Switching like 300VDC into a huge inductance is not what I would really call a good idea. They tend to give very poor "instantaneous power" demand regulation also. That was one of my pet hates with the original supplies that when the floppy drive turned on, the voltage would drop enough to cause the video display to dim fractionally. Of course there are the usual general motherboard problems which do not help either.
In terms of my power supply it likely has higher ripple but still practically nothing. But the instantaneous power is practically instant because it uses very small inductors. The regulation of course is a order of magnitude "tighter" because of the small inductors and faster operating frequencies etc. While the inductors are physically large, these are rated at 30A with basically zero resistance which helps with efficiency and regulation no end. But of course another pet hate is a lot of these inductors are only linear up to a certain amperage. In this case they are rated to 10A.. Which is still plenty because we are never going to be really drawing more than 2-4Amps from the 5V rail anyway. Most people who design stuff are totally oblivious to the actual "real" current rating of inductors. Just because it says 10uH 5A does not mean you can run it at that. Realistically after 0.5A the inductance starts dropping and basically becomes a short circuit, IE 0uH. This is why filtering progressively gets worse along with regulation because of inadequate inductors.. Don't get me started. :)
It's all swings and roundabouts really. I went with a huge toroidal for main side because you basically have a huge choke for mains filtering "built in". Also everything on the board is below 20VDC. In my old job I used to repair switch modes and high-voltage switch modes were notorious for failing and exploding. There was only one power supply whichever lasted and that was when they used a transformer primary side as I do. Again this is another reason I went with a lower voltage system. I mean yeah toroidal is pretty big and bulky but I think all the advantages it brings are certainly more than worth it.
So about 50mV to 100mV isn't to bad for a "off-the-shelf" PSU really. The original Atari ones were more like 200mV area.
Such swithmodes don't really have detectable ripple because the actual transformer acts as a huge "choke" which basically eliminates ripple. But as you witness they become very spiky because that huge choke has a huge inductance where it its flux cannot change very quickly. Switching like 300VDC into a huge inductance is not what I would really call a good idea. They tend to give very poor "instantaneous power" demand regulation also. That was one of my pet hates with the original supplies that when the floppy drive turned on, the voltage would drop enough to cause the video display to dim fractionally. Of course there are the usual general motherboard problems which do not help either.
In terms of my power supply it likely has higher ripple but still practically nothing. But the instantaneous power is practically instant because it uses very small inductors. The regulation of course is a order of magnitude "tighter" because of the small inductors and faster operating frequencies etc. While the inductors are physically large, these are rated at 30A with basically zero resistance which helps with efficiency and regulation no end. But of course another pet hate is a lot of these inductors are only linear up to a certain amperage. In this case they are rated to 10A.. Which is still plenty because we are never going to be really drawing more than 2-4Amps from the 5V rail anyway. Most people who design stuff are totally oblivious to the actual "real" current rating of inductors. Just because it says 10uH 5A does not mean you can run it at that. Realistically after 0.5A the inductance starts dropping and basically becomes a short circuit, IE 0uH. This is why filtering progressively gets worse along with regulation because of inadequate inductors.. Don't get me started. :)
It's all swings and roundabouts really. I went with a huge toroidal for main side because you basically have a huge choke for mains filtering "built in". Also everything on the board is below 20VDC. In my old job I used to repair switch modes and high-voltage switch modes were notorious for failing and exploding. There was only one power supply whichever lasted and that was when they used a transformer primary side as I do. Again this is another reason I went with a lower voltage system. I mean yeah toroidal is pretty big and bulky but I think all the advantages it brings are certainly more than worth it.
-
exxos
- Site Admin

- Posts: 28344
- Joined: 16 Aug 2017 23:19
- Location: UK
Re: [Video] More ST PSU shenanigans
We did that on the H5 :lol: :chairsmack:PhilC wrote: 13 Sep 2022 13:35 I've been wondering about having a single output 5V psu and then stepping it up to 12V with a cheap buck converter.
https://www.exxosforum.co.uk/atari/store2/#0223
-
PhilC
- Moderator

- Posts: 7440
- Joined: 23 Mar 2018 20:22
Re: [Video] More ST PSU shenanigans
Yes but I never got round to actually doing it. I might do now :Pexxos wrote: 13 Sep 2022 13:43We did that on the H5 :lol: :chairsmack:PhilC wrote: 13 Sep 2022 13:35 I've been wondering about having a single output 5V psu and then stepping it up to 12V with a cheap buck converter.
https://www.exxosforum.co.uk/atari/store2/#0223
If it ain't broke, test it to Destruction.
-
Badwolf
- Site sponsor

- Posts: 3043
- Joined: 19 Nov 2019 12:09
Re: [Video] More ST PSU shenanigans
Voodoo, if you ask me! I do like it when design choices are explained clearly and with reasoning, though. :-)exxos wrote: 13 Sep 2022 13:42 In terms of my power supply it likely has higher ripple but still practically nothing. But the instantaneous power is practically instant because it uses very small inductors. The regulation of course is a order of magnitude "tighter" because of the small inductors and faster operating frequencies etc. While the inductors are physically large, these are rated at 30A with basically zero resistance which helps with efficiency and regulation no end. But of course another pet hate is a lot of these inductors are only linear up to a certain amperage. In this case they are rated to 10A.. Which is still plenty because we are never going to be really drawing more than 2-4Amps from the 5V rail anyway. Most people who design stuff are totally oblivious to the actual "real" current rating of inductors. Just because it says 10uH 5A does not mean you can run it at that. Realistically after 0.5A the inductance starts dropping and basically becomes a short circuit, IE 0uH. This is why filtering progressively gets worse along with regulation because of inadequate inductors.. Don't get me started. :)
BW
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
-
Badwolf
- Site sponsor

- Posts: 3043
- Joined: 19 Nov 2019 12:09
Re: [Video] More ST PSU shenanigans
Yes, that looks ideal. This one kind of just 'fell into my lap' so went with what I had. :)coonsgm wrote: 13 Sep 2022 13:18 Great video! Watched and commented there already as well. I find the RD-35a is a better fit size-wise and also matches the original PSU output specs very well 5V @ 4A and 12V @ 1A.
Many thanks!Thanks for your videos! I wish I'd had it as a reference when I started.
BW
DFB1 Open source 50MHz 030 and TT-RAM accelerator for the Falcon
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
Smalliermouse ST-optimised USB mouse adapter based on SmallyMouse2
FrontBench The Frontier: Elite 2 intro as a benchmark
-
exxos
- Site Admin

- Posts: 28344
- Joined: 16 Aug 2017 23:19
- Location: UK
Re: [Video] More ST PSU shenanigans
:lol:Badwolf wrote: 13 Sep 2022 13:57 Voodoo, if you ask me! I do like it when design choices are explained clearly and with reasoning, though. :-)
Big retro toroidal = Good. Not much really to say I think :lol:
I mean you have a oldskool toroidal transformer which indeed classifies as retro these days. Mixed in with modern switch mode power supply designs. It is like mixing the old and new in a very elegant way... Or crossing the streams.. Or putting hydraulic suspension on a pushbike.. Jury is out I guess :lol:
-
PhilC
- Moderator

- Posts: 7440
- Joined: 23 Mar 2018 20:22
Re: [Video] More ST PSU shenanigans
And makes them much heavier and therefore much more useful for throwing through somebodies window when a half brick cannot be found.exxos wrote: 13 Sep 2022 14:03
Big retro toroidal = Good. Not much really to say I think :lol:
I mean you have a oldskool toroidal transformer which indeed classifies as retro these days. Mixed in with modern switch mode power supply designs. It is like mixing the old and new in a very elegant way... Or crossing the streams.. Or putting hydraulic suspension on a pushbike.. Jury is out I guess :lol:
If it ain't broke, test it to Destruction.
Who is online
Users browsing this forum: ClaudeBot and 8 guests