-
I've dabbled in low voltage DC electronics for a while, but I'm looking at learning and building a lot more. I intend to build Bob Heil's Pine Board AM Transmitter and part of it is building a power supply for it.
That got me thinking about PC and server PSU's in how they can accept both 240v and 120v inputs. How exactly do they handle the different voltages? Does it just detect if it's 240v and use a center tap or the entire winding if it detects 120v? Sorry if this sounds like a stupid question.
Thanks,
Dave -
Unlike linear power supplies, which use hefty laminated iron core transformers, computer power supplies use smaller, high frequency switching transformers. The latter support a much wider input voltage range. There is no need for any tap switching, the regulator simply adjusts the PWM duty cycle to get the correct output.
I would avoid a switchimg converter for your specific application, because your RF circuit might not interact well with the noisy power supply.
If you want to learn more about switching converters, I recommend you first try to wrap your head around simpler topologies, like the buck converter. -
I've dabbled in low voltage DC electronics for a while, but I'm looking at learning and building a lot more. I intend to build Bob Heil's Pine Board AM Transmitter and part of it is building a power supply for it.
That got me thinking about PC and server PSU's in how they can accept both 240v and 120v inputs. How exactly do they handle the different voltages? Does it just detect if it's 240v and use a center tap or the entire winding if it detects 120v? Sorry if this sounds like a stupid question.
For a simple linear power supply using a transformer, such as you might build yourself, this can be done using a transformer with two 120 V primary windings. You put them in parallel for 120 V mains and in series for 240 V mains. This is most simply done using a DPDT mains switch.
I have a power supply where you have to open the case and manually reconfigure the wires, on the assumption that you make such configuration changes very rarely. -
Think of a switch-mode power supply as a device that simply takes little sips of power from the primary side and sends them to the secondary, just often enough to average out to the necessary output voltage. With a higher input voltage, it simply takes shorter sips of power from the primary.
For example, imagine if the primary side is 120V, and you need 12V at 1A. Then (depending on the transformer winding) you might only draw power from the primary 5% of the time. If the load increases to 2A, you might need to draw power 10% of the time. At 240V, though, with the 2A load it'd only need 5% of the time.
Look into PWM (pulse width modulation), as it's a really good way to get a very rough understanding of the principles. -
Have a look at voltage doublers:
https://duckduckgo.com/?q=psu+"voltage+doubler"&iax=images&ia=images
The principle is quite simple.
For SMPS power supllies on 230Vac the input is a simple full bridge which rectifies to about 310Vdc.
Then there are 2 small modifications to have it run on 120Vac.
First the elco is split in 2 elco's in series.
Then the center tap between the elco's is shorted to one of the input terminals.
This means the top elco is rectified to around +150Vdc, and the bottom elco is rectified to around -150Vdc, and together they are 300V, about the same as from the full bridge.
Additonal advantage is that 200V elco's can be used instead of 400V elco's, Disadvantage is that somewhat bigger elco's must be used because of the (2x) half wave rectification for 120Vac.
--------------------------
With smaller power supplies it's more common that they simply accept a wide change of input volatage.
With a bit of forethought it's not too difficult to design an SMPS for which the input voltage can have a 1:4 ratio.
If you have a look at the very common small smps modules from Ali / Ebay / China with for example a LM2596 or MP1485, they have a wide input voltage, and they simply adjust the PWM duty cycle internally to regulate the output voltage.
-
Just a little FYI, native English speakers won't know what you mean by "elco"*, we just call them "electrolytics" if not calling them by their full name, or just "caps" when specifying the capacitor type isn't actually important in context.
*we stopped calling capacitors "condensers" a loooong time ago in English. -
Native English speaker here, years of repairing Grundig and Phillips electronics and reading Elektor taught me Elco
-
That got me thinking about PC and server PSU's in how they can accept both 240v and 120v inputs. How exactly do they handle the different voltages? Does it just detect if it's 240v and use a center tap or the entire winding if it detects 120v? Sorry if this sounds like a stupid question.
Most PC and server PSUs have a voltage selector switch to be set by the operator.
Small PSUs like for laptops have a wide input range, say 100 to 240 volts because they are designed to accept that range. They do not switch between voltages.
I suppose you could design a PSU that detected input voltage and switched but I do not think they are common at all.
-
That got me thinking about PC and server PSU's in how they can accept both 240v and 120v inputs. How exactly do they handle the different voltages? Does it just detect if it's 240v and use a center tap or the entire winding if it detects 120v? Sorry if this sounds like a stupid question.
Most PC and server PSUs have a voltage selector switch to be set by the operator.
Small PSUs like for laptops have a wide input range, say 100 to 240 volts because they are designed to accept that range. They do not switch between voltages.
I suppose you could design a PSU that detected input voltage and switched but I do not think they are common at all.
Are you joking?
The majority of power supplies in PCs for the past two decades have been wide input range or autoswitching, they are by far the most common type, there are even dedicated ICs for detecting line voltage and autoswitching. -
Are you joking?
The majority of power supplies in PCs for the past two decades have been wide input range or autoswitching, they are by far the most common type, there are even dedicated ICs for detecting line voltage and autoswitching.
You made me look. I have three desktops right here and I was sure I was going to find a switch that switches to voltage doubler configuration. You know what I found instead? Stickers that say "230 volts only". On all three.
I remember the 110/220 sliding switches on the older PSUs but I guess they got cheap and removed them.
I went to pccomponentes.com, an online PC components vendor, and most of the PSUs are 230 volt only. Only one says "full range". None have voltage selector switches.
-
The actual trick with modern PC power supplies (at least good ones) is that the first stage of a decent supply in the reasonably high power class is generally a BOOST converter that takes the input up to somewhere around 390V or so.....
This is independent of the input voltage and is done by some cleverness with a dual loop controller that makes the input current track the voltage waveform so that the load appears resistive. It is a requirement in larger supplies to control the power factor and harmonic distortion.
If you have to have such a stage to meet regulations it is cheap to accommodate a wide input range, hence the common 85-265V range seen on equipment.
Once you have your (more or less fixed) DC bus at 390V or thereabout you then do the main switching converter that produces the outputs and also provides the isolation.
The half wave voltage multiplier (with the slide switch) trick was common before PFC became a requirement.
Regards, Dan. -
Active PFC is expensive to implement and not really required on PC PSUs which can achieve harmonic reduction with passive low pass filters which are simpler and cheaper. I do not think active PFC is common in computer PSUs.
Yes, I know they exist, just like 500 euro bank notes, I have heard of a guy who once talked to someone who said he had seen one.
-
You made me look. I have three desktops right here and I was sure I was going to find a switch that switches to voltage doubler configuration. You know what I found instead? Stickers that say "230 volts only". On all three.
I remember the 110/220 sliding switches on the older PSUs but I guess they got cheap and removed them.
I went to pccomponentes.com, an online PC components vendor, and most of the PSUs are 230 volt only. Only one says "full range".
A bit of a mystery then as only one out of the first ten PSUs I looked at on a local supplier site was 230V only (it was one of the cheap ones, I might order one to satisfy my curiosity), I can't remember ever seeing a 230V only PSU in the past ten years in any PC, all the machines I have here (site with 800 users) and at home (with the possible exception of a HP Vectra 486) are wide input range.
I had thought it might be because your supplier offers high power gaming PSUs as some server PSUs derate output if operated on 110V (some HP supplies are derated to 960W from 1200W) but even that doesn't seem to be a reason for limiting to 230V as Scan offer similar and higher rating PSUs which are 100-230V input so I'm mystified as to why it would still be a thing when it's so simple to implement.
(BTW, the first two PSUs your supplier offers are wide input ones)
-
Maybe in the UK they have different requirements.
-
Dont look to the crap from Nox..
Look at decent PSU's, like Seasonic, and they all state this:
AC Input Voltage: 100 V - 240 V
And, they have active PFC, who knew..
https://www.pccomponentes.com/seasonic-s12ii-520w-80-plus-bronze -
Dont look to the crap from Nox..
Look at decent PSU's, like Seasonic, and they all state this:
AC Input Voltage: 100 V - 240 V
And, they have active PFC, who knew..
https://www.pccomponentes.com/seasonic-s12ii-520w-80-plus-bronze
I've a feeling the only way to get a efficiency rating is with active PFC? -
Harmonic suppression compliance can be generally achieved with passive filters but active PFC has other benefits, mainly that it is, in itself, a pre-regulator so the following switching part is working pretty much from a constant DC voltage.
-
99% of 80+ bronze/gold and higher ATX power supplies are wide input range
80+ "white" allows 230v only psus, which otherwise won't reach 80% effiiciency on 110v or they use small/cheap parts (ex a 4A bridge rectifier or 2-3A diodes on a 500w psu.. too low to get 500w on 110v)
A linear psu can be made using transformer with 2 primary windings and parallel or series them depending on voltage/
Or, you can do like atx psus AC->DC, optional boost to ~400-420v DC (actve pfc), then high frequency switching (typically at least 30kHz or higher to stay above hearing range, but 200kHz..500kHz is common) -
They don't do any autoswitching except very rare cases. Manually switchable input voltage is still more common than autoswitching. LOW power SMPS usually just work in wide input voltage range. Those of higher power have PFC (works in wide input voltage range as well) which converts input voltage into around 400VDC and then SMPS works from that.QuoteI suppose you could design a PSU that detected input voltage and switched but I do not think they are common at all.
Are you joking?
The majority of power supplies in PCs for the past two decades have been wide input range or autoswitching, they are by far the most common type, there are even dedicated ICs for detecting line voltage and autoswitching. -
You cannot legally sell them in EU without PFC. SMPS above certain power rating do have minimum power factor requirement. And simple rectifier + smoothing capacitor do not meet that.Dont look to the crap from Nox..
Look at decent PSU's, like Seasonic, and they all state this:
AC Input Voltage: 100 V - 240 V
And, they have active PFC, who knew..
https://www.pccomponentes.com/seasonic-s12ii-520w-80-plus-bronze
I've a feeling the only way to get a efficiency rating is with active PFC? -
The EU mandates limits on harmonics and does not care about how you achieve that.
https://en.wikipedia.org/wiki/IEC_61000-3-2
A passive low pass filter can meet the requirements and is cheaper than active PFC.
-
The EU mandates limits on harmonics and does not care about how you achieve that.
https://en.wikipedia.org/wiki/IEC_61000-3-2
A passive low pass filter can meet the requirements and is cheaper than active PFC.
I very much doubt that it is cheaper, third harmonic is only 150Hz, so that is a LARGE inductor to keep that out, while an active PFC is like one mosfet, a gapped inductor and a small controller chip, a boost stage running at maybe 50KHz is going to be cheaper then a LC filter cutting off below 150Hz, probably just on the shipping cost never mind anything else!
Certainly by the time you are looking at any kind of modern server power supply, I would be VERY surprised to see passive harmonic filtering, rather then active PFC.
Regards, Dan. -
I very much doubt that it is cheaper, third harmonic is only 150Hz, so that is a LARGE inductor to keep that out, while an active PFC is like one mosfet, a gapped inductor and a small controller chip, a boost stage running at maybe 50KHz is going to be cheaper then a LC filter cutting off below 150Hz, probably just on the shipping cost never mind anything else!
I have no opinion on what is cheaper but most cheap PC PSUs do not have active PFC and I assume they meet applicable regulations.
My TV, computer monitors and laptop PSUs do not have active PFC either.
Now that I think about it I can't think of any appliance or light in my home that has active PFC.
Residential customers do not pay for reactive power so they have no incentive to improve the power factor.
OTOH big businesses do pay for reactive power and have an incentive to improve PF. I remember reading about some company that had LED lighting with active PFC because they had thousands of such lights and PF was an important consideration.
-
I think there was an exception in EU rules for psus with less than 75w output, so your monitors may lack active pfc.
TV may use less than 75w, depends on backlight.
if your laptop's brick is over 65w, it should have active pfc.
-
I think there was an exception in EU rules for psus with less than 75w output, so your monitors may lack active pfc.
You think there is an exception? Why do you think that? And what exception?
TV may use less than 75w, depends on backlight.
if your laptop's brick is over 65w, it should have active pfc.
Because, you know, I just linked to the relevant regulation.
https://en.wikipedia.org/wiki/IEC_61000-3-2
Did you read it?
That regulation establishes allowable harmonics current and says absolutely nothing about active PFC. And I cannot imagine why they would care how you achieve the objective.
If you have better information I'm sure we'd all like to see it but "I think" doesn't cut it.
All my devices are over 80 W and none have active PFC.