Thanks folks! I have a spare step-up transformer. I just wasn't sure if providing the UK Atari STE a 240V / 60 Hz power source would be a problem as it's probably designed for a 240V / 50 Hz source. On Sun, May 31, 2026 at 10:31 Wil Birkmaier <wil@birkmaier.org> wrote:

I have one I just pulled out the psu and rebuilt it with a modern 100 to 240 or something. The hz I believe is set by the tos.

Been a while don't remember but if you want I can tell you the PSU used, for some reason meanwell comes to mind.

On May 31, 2026 09:23, John Heritage via vcf-midatlantic < vcf-midatlantic@lists.vcfed.org> wrote:

Hi folks,

I'm importing an Atari STE from the UK, that requires 240V. I have a 120V to 240V transformer, and was wondering if that would work, or would there be a problem because the UK's grid runs at 50 Hz while the US grid runs at 60 Hz.

I plan to replace the internal power supply with a modern 120V solution, but wanted to test the STE out of the box before I replaced the PSU.

Thanks! John

On 5/31/26 11:20, David Wade via vcf-midatlantic wrote:

...

Thanks folks!  I have a spare step-up transformer.  I just wasn't sure if providing the UK Atari STE a 240V / 60 Hz power source would be a problem as it's probably designed for a 240V / 50 Hz source.

Its a switched mode PSU so the first thing that happens is that the mains is rectified and used to charge a high-voltage capacitor. So I think it will be fine with 240v/60v.

Really the usual cause of problems is running a US designed setup on 110v/50hz. Transformers efficiency depends on frequency, which is why a SMPSU which runs a kHz can have such a small transformer

I'm sorry Dave but this is wholly incorrect. Switching power supplies typically don't use transformers to step down voltages; they chop the rectified incoming voltage using a chopper transistor driven by an oscillator, and control the pulse width such that the integrated voltage (area under the curve) is the target voltage. An integrating network and filtering follows the chopper to produce a steady DC voltage, and a voltage divider pick-off is compared against a reference voltage and used to control the PWM width in a feedback network, resulting in regulation to the desired output voltage. It is not a matter of converting the frequency higher so a smaller transformer can be used in a linear power supply arrangement. -Dave -- Dave McGuire, AK4HZ New Kensington, PA

On 5/31/26 12:51, David Wade via vcf-midatlantic wrote:

...

  I'm sorry Dave but this is wholly incorrect. Switching power supplies typically don't use transformers to step down voltages;

I know, my bad. I meant to say...

Opps pressed send too quickly. I meant to say :-

Switched mode PSUs don't have issues with frequency. The usual cause of problems is running a US designed setup with a *linear *PSU on 110v/50hz in the UK because the transformer is less efficient and can overheat. This can be an issue with some Tandy Color Computers as the transformer runs warm anyway...

Yes, 100% correct.

...

they chop the rectified incoming voltage using a chopper transistor driven by an oscillator, and control the pulse width such that the integrated voltage (area under the curve) is the target voltage. An integrating network and filtering follows the chopper to produce a steady DC voltage, and a voltage divider pick-off is compared against a reference voltage and used to control the PWM width in a feedback network, resulting in regulation to the desired output voltage.  It is not a matter of converting the frequency higher so a smaller transformer can be used in a linear power supply arrangement. Then why does the Atari PSU have a transformer? It also puzzles me is that there is only one transformer, yet we have two output voltages, +5v and +12v. I can understand using PWM to control one voltage, but how do you keep one output at 12v and one at 5v when the load varies?

This is the schematic on my onedrive...

PSU_ASP34-1.pdf <https://1drv.ms/b/c/277a0739f125010e/IQBqZ48f- MTZTKUuEUKJyzYqAa4emkFH5F386XGY203tiug?e=g8v068>

Many (not all) switching power supplies have transformers, but they are not what are colloquially known as "power transformers" in the bulk step-down sense. The chopper system generates high frequency chopped DC which is fed to a transformer primary. There may be a step-down ratio involved, but that is not the primary function of the transformer; the PWM action and integration is what drops most or all of the voltage. In the case of the Atari power supply that you referenced, what's marked as T1 is a common-mode choke (and thus probably shouldn't use "T" as a reference designator), then the mains voltage is rectified by DB1, smoothed by C4 and C5, while C2 provides a path to ground for any high-frequency hash. Q1 and Q2 comprise the chopper drover and chopper to drive the primary of T2, and T2's two secondaries provide 12V and 5V, integrated by C14+L2+C15 and C12+L1+C13. The pickoff to close the feedback loop is derived from the +5V output. The real efficiency boost in a switching regulator comes from running the pass element either in saturation or in cutoff, with as little time as possible spent in the linear region. In saturation, as much current as possible flows through the transistor (low resistance -> low dissipation), and in cutoff, almost no current flows through the transistor (low current -> low dissipation). Contrast this to, say, a big honkin' 2N3055 pass transistor being servo'd around in its linear region, passing DC, acting as a current-controlled variable resistor, turning a chunk of the energy (Pd = (Vin - Vout) * Iout) that passes through that resistance straight into heat. Next...it is very reasonable to wonder how one point of regulation can produce two or more regulated voltages, but there is no magic here and it's just a big short-cut with minimal drawbacks. In the case of most TTL/MOS/etc systems, the heaviest load is on the +5V rail, so that's what will experience the most sag. The load on +12V is usually minimal; I don't know offhand what it's used for in the Atari system that this is the PSU for, but most of the time it's RS232 line drivers and that sort of thing, i.e. very light loads. So we'd naturally want to use the +5V rail for the regulation loop. That means, as you observed, that the +12V rail's actual voltage will vary based on the load on the +5V rail, but most +12V loads in typical computer systems are far less voltage-sensitive than +5V loads (typical TTL ICs are specified with a Vcc range of 4.75V to 5.25V) so in practical terms it just doesn't matter. Nearly all multi-rail switching power supplies use one voltage for the regulation loop's feedback input, and as you can imagine, choosing the right one is a critical part of the design process.

I know your love for all things Microsoft knoweth no bounds, so it came from here, but I couldn't figure out what the real URL for the PDF was

It is not a prejudice; please, I'm more professional than that. As an engineer, it is the essence of my job know a good and bad technologies and products when I see them. The day they start writing software competently, I will have less of an issue with their products. Excel isn't too bad, and the original Microsoft C is pretty good. But aside from that, the reality is that most of their products are toys, or worse, designed to create vendor lock-in situations. Sure, some people can muddle through like riding a bike with square wheels and occasionally get a little bit of work done...and construction workers could similarly haul construction materials in a Volkswagen. That doesn't mean it's a good idea. It is engineering professionality, knowing how things work, and seeing things for what they are that's in play here, not any sort of personal dislike, as people often dismissively say. And, frankly, when I see a good person beating themselves on the head with a hammer, I'm generally inclined to say "Hey, stop that, you'll hurt yourself!" -Dave -- Dave McGuire, AK4HZ New Kensington, PA