Hello EEVBlog Forums,
Here's a schematic of a transformer with a dual primary:
I'm interested in how a lot of the power supplies, specifically computer ATX power supplies, switch between input voltages of 110V to 230V with the use of an SPST slide switch. See image below for visual of switch:
Any help is greatly appreciated!
-Bear
switchmode power supplies do not use a transformer like the one you are thinking of. The switch could for example alter a feedback network to change the output voltage. I'm sure myself what that switch does but don't confuse old 50Hz transformers with computer SMPS
Switchmode power supplies don't use a classic transformer like the one in the first picture.
They convert the AC voltage using a bridge rectifier BEFORE it reaches a transformer, smooth it out using capacitors and then there's a circuit which sends pulses of DC voltage on the primary side of the transformer therefore generating a square wave, lots of pulses of electricity, usually between around 20k and 100k pulses a second. Depending on what voltage comes out on the other side, the circuit adjusts the number of pulses sent on the primary side.
In the first picture of your post, the transformer is always running at 50-60 Hz, that's why it's so much heavier and larger compared to switch mode transformers.
The switch in the back of some old power supplies simply rearranges how two capacitors are seen in the circuit of the power supply.
It boils down to Figure 10 in this application note (see page 7) :
http://www.ti.com/lit/an/snva006b/snva006b.pdf The switch simply connects the middle of those two capacitors to the bridge rectifier, changing how those two capacitors are seen in circuit. Those capacitors are usually rated for 200v or something like that - if the voltage switch is set to 115v and the actual voltage is 230v then the rectified DC voltage is 230x1.414 = ~ 320v and therefore the 200v rated capacitors blow up.
Depending on what voltage comes out on the other side, the circuit adjusts the number of pulses sent on the primary side.
You mean the duty cycle or "on" versus "off" time of the pulse that is usually a fixed frequency.
Ironically, that type of switch pictured is usually a DPDT that they don't use all the contacts of.
You can do a series/parallel connection using exactly one DPDT switch. You need to arrange the switch so that either position achieves either jumper configuration in the diagram, of course.
Tim
Ironically, that type of switch pictured is usually a DPDT that they don't use all the contacts of.
You can do a series/parallel connection using exactly one DPDT switch. You need to arrange the switch so that either position achieves either jumper configuration in the diagram, of course.
Tim
Thanks, this answered my question perfectly! Without the DPDT possibility I might have never found out a way to do it with a conventional power transformer.
switchmode power supplies do not use a transformer like the one you are thinking of. The switch could for example alter a feedback network to change the output voltage. I'm sure myself what that switch does but don't confuse old 50Hz transformers with computer SMPS
While the original problem is solved already, let me add this note to the above: While Simon is of course correct in what he writes, the kind of switch we see in the original pic is not needed at all if you have a "proper" general purpose SMPS producing one unvarying set of output voltages . Such SMPSes usually accept inputs from say 90 VAC to 265 VAC or so, and have sufficient range in the duty cycle of the switcher to adjust the output voltage(s) into range regardless of input variations.
switchmode power supplies do not use a transformer like the one you are thinking of. The switch could for example alter a feedback network to change the output voltage. I'm sure myself what that switch does but don't confuse old 50Hz transformers with computer SMPS
While the original problem is solved already, let me add this note to the above: While Simon is of course correct in what he writes, the kind of switch we see in the original pic is not needed at all if you have a "proper" general purpose SMPS producing one unvarying set of output voltages . Such SMPSes usually accept inputs from say 90 VAC to 265 VAC or so, and have sufficient range in the duty cycle of the switcher to adjust the output voltage(s) into range regardless of input variations.
Although an SMPS might be the way to go, it's not something I feel comfortable designing myself as it is quite out of my realm of understanding. Many people suggest simply buying a power supply which does the job or re-purposing an existing power supply (i.e. a printer's power supply, etc.). However, I would very much like to do the power supply for my circuit myself as both a learning experience and a fun project which I can honestly say that I really did design and put all of it together myself.
Thanks for all the help so far!
If you plan to make a product with classic transformers, don't forget about the fact that in some places, the voltage can be much lower, or much higher than the default one... and sometimes it's not just temporary, for a few hours.
In some areas, there may be only 100v AC at the mains for long periods of time. When you determine the losses in bridge rectifier and how large the filter capacitor should be, keep that in mind.
Just the same, Europe standard voltage is 230v but there are some countries that are 240v and it's quite possible to have 210-245v AC
What I'm trying to say is... don't design something that will output (for example) a minimum 7v to a linear regulator, thinking 2v will be enough voltage drop for the linear regulator to work fine and output 5v.
Aim for minimums of 8-10v and don't ignore the fact that your maximum output voltage may be 10-30% higher than normal.
In some areas, there may be only 100v AC at the mains for long periods of time.
Like forever. In Japan, for example.
Which is why you see good SMPS units rated from 90V up to 250V or so.
If you plan to make a product with classic transformers, don't forget about the fact that in some places, the voltage can be much lower, or much higher than the default one... and sometimes it's not just temporary, for a few hours.
In some areas, there may be only 100v AC at the mains for long periods of time. When you determine the losses in bridge rectifier and how large the filter capacitor should be, keep that in mind.
Just the same, Europe standard voltage is 230v but there are some countries that are 240v and it's quite possible to have 210-245v AC
What I'm trying to say is... don't design something that will output (for example) a minimum 7v to a linear regulator, thinking 2v will be enough voltage drop for the linear regulator to work fine and output 5v.
Aim for minimums of 8-10v and don't ignore the fact that your maximum output voltage may be 10-30% higher than normal.
The circuit which is being powered consists of very flexible input voltage devices - roughly 18V to 40V is my optimal input range. If I design with a 28Vrms secondary, the peak voltage will be somewhere around 40V subtracting some voltage drops here and there I'd estimate to be around 32-37 Vdc after rectification and loading it down. The only reason I want higher voltages in the first place is due to the boost drivers being more efficient the closer the input voltage is to the output voltage. I can design around low voltages, but my output constant current won't drive the LEDs to high enough brightness (must be greater than or equal to 350mA at the least).
I may end up using a transformer with a 24Vrms secondary due to the maximum input voltage on mains. This should still give me enough wiggle-room along with being a tad more common than 28V secondaries.
EDIT: On second thought, I just received an email from DigiKey about some parts. I didn't realize that some of these really nice, high-output wattage supplies are so cheap:
here. I will probably go with this option as I really can't justify paying the same amount for simply the transformer alone. I could possibly analyze this circuit and learn a few things as well!
