# Transformer for a power supply

1. Jun 10, 2014

### Pharrahnox

I am going to make a power supply. I understand the smoothing capacitor and the full-wave bridge rectifier.

I am trying to make a 60V, maybe 10A power supply - maybe only 5A, but I would prefer 10A. I don't have a transformer for it, though. I have researched a fair bit on transformers recently, and am now looking to either buy or build one. It will need to be 300W or 600W, so a fair bit of power.

I am having trouble finding where to find one, and I don't know how to build one with the materials I have. Does anyone have any suggestions.

BTW, I should mention, my country's mains power is 240V, 50Hz, so I will need a 4:1 transformer.

On a side note, the 60V power supply that I had bought, Yihua 605d (it was cheap), has stopped working after I connected a slightly-charged large capacitor to it. I looked inside to find a fuse but couldn't see one. What else may have been damaged to cause it to stop working?
I do realise that I could pilfer the transformer out of the power supply, but I would prefer to try to fix it first.

2. Jun 10, 2014

### davenn

Building a transformer isn't impossible many people used to do it in years gone bye
but for a novice it would be a major undertaking

Transformers are readily available these days with all sorts of output voltages and current capabilities

one with an output of 300 to 600W at 60V would be readily available as there would be used in any decent hi power audio amplifier power supply
I'm sure you would probably fin one either through one of the major electronics suppliers like Digikey, Element 14, Mouser etc.

Have no idea where you are ??
But if you lived close to Sydney Australia you could have one of my old audio amp transformers :)

cheers
Dave

3. Jun 10, 2014

### Pharrahnox

Thanks for the offer, but I'm down towards the bottom of Victoria.

I had a look at DigiKey, Element14 and Mouser, but the 300VA 60V transformers all cost >= $100. I could almost buy another Yihua 605d for that much... I have looked up how to build a transformer a bit. I found this site : http://ludens.cl/Electron/trafos/trafos.html It is very detailed, there's a lot of information, but it makes it look quite complicated and difficult... I actually have found a transformer in my collection of junk circuit boards, it had 63V capacitors after it, so I would assume it to be somewhere near 60V. It has dimensions of 61*55*65mm and has 6 taps on one side. It is a SONY 1-433-729-12, but I couln't find any specs on it. I might just have to determine it experimentally. I have never really dealt with mains power directly, so if I were to plug this transformer in and have a very high resistance load 1MΩ on the secondary side, it would only use 3.6mW. That sounds safe, but are there any dangers, other than the obvious don't stick your tongue on it? Another question. How can I tell how much power is dissipated on the primary side of a transformer if the resistances of both sides are known? Or is only from power lost in the transformer? If so, then is it feasible to create a tiny, weak transformer with only a few turns of wire, with a very high resistance load on the secondary side? In this case I wouldn't care about the power, just transforming the voltage. 4. Jun 10, 2014 ### davenn a 300-600W transformer is going to be a respectable cost and they are bulky and heavy. With the work involved in building one of that rating, I personally wouldn't even consider it my time is too valuable to waste hours doing that instead of a couple of 100$ buying one
its just not worth the hassle

cheers
Dave

5. Jun 12, 2014

### Pharrahnox

I agree, it probably isn't worth the hassle.

I have possibly found the answer. I have found variable transformers online that can deliver even higher powers at lower costs. I am aware that they aren't isolated, and I will research safety precautions in the case that I do purchase one.

However, I would still like to know if a very small, very simple transformer could be constructed using a few dozen turns of wire in each winding, wrapped around a 10" iron nail, and then plugged into a power point to give the required voltage, at potentially very low powers, such as < 5W, with a high value series resistor in place, or something like that.
Is there any very simple and easy (and of course safe) way to get 60V out of a 230V powerpoint, that I can then rectify?

6. Jun 12, 2014

### davenn

No, you would have a bang, lots of sparks and a cloud of smoke.
PLEASE don't even consider trying it

Dave

7. Jun 12, 2014

### CWatters

8. Jun 12, 2014

### Pharrahnox

Thanks for the link. That looks quite promising. I may get that, depending on the shipping cost for an overseas 2kVA variac. I would prefer having variable voltages.

What I mean by the 5W transformer is just something I can easily build to give whatever voltage I want, and that is safe to use in a powerpoint.

9. Jun 12, 2014

### enorbet

I'd like to offer a few points of interest, caution, and suggestion.

Transformers do not have voltage beyond breakdown voltage of the insulator. They have ratios of impedance. Those 63v caps don't signify a particular transformer but rather a transformer rectifier combination. Example - Half Wave and Full Wave rectifiers, and Doublers, will deliver very different DC voltage to the caps using the identical transformer. It is an indication however that it will be easy to get the desired 60V or nominally close. BTW you're likely better off with a 3:1 or 3.5:1 since losses in ripple (filtered away to ground) will make it exceedingly difficult to render 60VDC from 60VAC.

Modern core technology and winding techniques have reduced the size of transformers tremendously. Winding your own by hand around a nail would be a waste of time and energy, both yours and the power company's.

Since the house power voltage is, I assume, consistent where you live almost any used transformer can be made to work provided it will handle the current you require. It is pretty trivial to convert voltage. Granted it is substantially easier to just buy "the right one" but since you balk at \$100+ for a 500 Watt transformer, it seems an option to scavenge one.

When working on power supplies the danger is not technically voltage but rather amperage and that applied across your heart. 0.1 Amp across your heart can kill you at any voltage. Always use the "Single Hand Technique". Put one hand behind your back when the other is close to any part of a powered PSU. The only time to break this rule is to take measurements through probes with a known voltage resistance, but it is wiser even then to make the ground probe an alligator clip or some variation, requiring only one hand on one probe. Then the VOM is at risk, not you.

10. Jun 13, 2014

### Pharrahnox

Thanks for your words of caution, and I'm aware of the dangers of electricity - not first hand experience, just research. I'll definitely use the "Single Hand Technique", especially when experimenting.

I just measured the actual voltage from the power point; it's 247V. What would be a good leeway to achieve 60V from this: what turns ratio should I have? It will be rectified by a full-wave diode bridge, so there would be ~1.4V drop across the diode bridge. A 4:1 ratio would give 60.35V after taking into account the voltage drop across the diodes. But that doesn't include losses in the wires or other things that might need to be considered.

There a few (important) things that I still don't understand about transformers:

If I have a transformer plugged into 240VAC, and it is a 4:1 ratio, and I have a secondary load of 0.5A, the current I would expect in the primary would be 0.5/4 = 0.125A.
Would I then use I2R to calculate the power lost in the primary winding? For example, if the primary winding had a resistance of 0.5Ω, then the power loss would be 0.125^2 * 0.5 = 7.8mW.

Would that then mean that the voltage drop across the primary winding is 0.125A*0.5Ω = 62.5mV?

If I were to make a shoddy transformer with a 4:1 turns ratio to use from mains power, what other considerations do I have to take into account? I don't overly care about the efficiency, I just want to be able to get 60V, at low powers, and for it not to melt or spark. What is the affect of having more turns of wire? Does it make the transformer more effective? I have loads of 18awg 26awg wire, but what amount of turns do I need to use to make it work?

Thanks for any input.

11. Jun 14, 2014

### enorbet

Greetings
In answer to your first part the calculation isn't based on voltage drop because this is conversion from AC to DC and you will get near RMS values and the rest will be ripple. In practice, the very best one can do is around 80% of AC input will result in DC output. Since it is easier to drop voltage than increase it, I would seek to have around 80-90VAC on the secondary. It might be valuable to remember that very few devices are voltage critical beyond 10-20% so it might be important to discover what this device really requires. Example - most 60VDC devices will operate with negligible losses well below 50VDC, provided sufficient current is present. This reflects the importance of OHM's Law since it is Wattage that is power, not merely Voltage. It is also why the power losses you mention are negligible especially in a "cheap and dirty" transformer.

The effect of more turns is better inductive coupling, resulting in greater transfer efficiency with less losses as heat.

Since we are talking real world here and not just hypothetical, let me kindly repeat that the time spent to wind one's own homebrew transformer is better spent "cobbling" an existing one unless your end goal is to discover just how far technology has come in core materials, geometry, and winding techniques, not to mention specialty wire of bifilar and trifilar design. Winding common wire around a nail does work to demonstrate the base simplicity of transformers but is far from practical to actually use to power a device.

People throw away electronic devices "by the ton" and a very large percentage use DC and therefore transformers - a scavengers gold mine. It would be vastly easier to modify your rectifier circuit for desired DC voltage (eg: choosing half wave rectification to drop a higher VAC, or simple RC networks after the fact, to usable DC levels) than to get or build an exact transformer for what appears to be casual usage.

12. Jun 17, 2014

### Pharrahnox

Even if it is only to demonstrate the principle practically, I still believe it to be worth a few minutes. I will use many of turns if it increases the efficiency. I assumed that it would, but wanted to be sure.

For most purposes, I wouldn't overly care about having fairly precise voltage limits, but because I would like to power circuits containing 63V capacitors, I believe it would be a good idea to not go much higher than 60V.

I have heard of RC networks before, but only to multiply voltage. How can I set one up to reduce the voltage by a known amount?

13. Jun 18, 2014

### enorbet

Regarding homebrew transformers - It is extremely doubtful that one layer of windings will be sufficient, nor is it wise to simply wind the primary and then wind the secondary on top of it. They should be wound together as much as is possible.

Let's say you calculate the number of Primary turns and subsequently, how many layers that will be. It will in all likelihood be substantially more than three (3) but lets use that number for simple demonstration and because it closely fits the turns ratio. Wind the first layer of Primary and then wind 1/3rd the number of turns of the Secondary over it and then resume with the Primary over the top of that and once again switch to Secondary when that layer is complete, and then repeat till complete. This is very primitive but will at least give some reasonable (for a homebrew) windings inductive coupling so it actually works as a transformer instead of as just a heater.

The "R" (resistance) part of an RC network is essentially just resistors in series, although some will use "bleeders" in parallel for further control but at the expense of a little loss. It is easy to calculate the value of resistance(s) required to achieve a given voltage drop if you know how much power the device draws.

The "C" (Capacitance) part is for DeCoupling. They isolate one section from another, further reduce ripple as well as ground loops. They also act something like a reservoir or flywheel in maintaining stable voltages under dynamic loading.

If you Google for "Power Supply Decoupling" you will find descriptions, formulas, and pictures that may help you decide the details of how you can and should proceed.

Additional components that are useful in dropping voltage, especially if precision is desired, are Zener Diodes or commercially available VoltageReg ICs.

You cannot multiply voltage with simple RC networks. This requires Diodes as well as RC and a thorough understanding of Ground/Earth techniques. Reducing voltage with R or RC is accomplished with the most basic application of Ohm's Law.

Regulation, however, is rather important so I suggest you consider the above mentioned VR IC method. They are simple and cheap. Here is a basic document describing power supply design and the use of such ICs.

http://www.egr.msu.edu/classes/ece480/capstone/spring09/group06/power supply app note.pdf