Info on custom power supplies, please

[swinginguitar]

Reviving the thread to ask:

when I fire up these transformers to test/measure, what sort of load do i need to put on them?

[R.G.]

Can you clarify "these transformers" a bit?

[swinginguitar]

Haha

Yeh - I scrounged some trafos - one from a printer PSU, one from a bookshelf stereo, and yet another from a timer/psu for a sprinkler system.

I want to fire them up and measure the actual voltages, then do some breadboarding with regulators and such.

Will they blow without a load (giggity)? Do I need to put a resistance across the 2ndaries?

[mattthegamer463]

For a more accurate measurement you should load with a high power resistor, maybe 100 ohms and 10-20 watts, should give a decent current draw from a low voltage transformer secondary.

Transformers will dispense their energy as heat when not loaded, it is possible they can burn out but if you're just measuring voltages for a few seconds they won't heat up and die.  The iron cores can absorb a massive amount of thermal energy.  Make sure you don't hook them up backwards or else a thousand volts might come out the other side.

Seriously, be careful.

[R.G.]

Actually, matt, that is a misconception. Transformers by themselves do not burn up or blow out when unloaded. Tube amplifiers with output transformers can do this, and that is related to there being a transformer on the output, but the transformer itself is not the problem, it's the interaction of the whole amp + transformer.

Transformers have voltage limits and also voltage divided by frequency limits.

You are correct that transformers will dispose of the internal energy as heat; however, the internal energy is roughly the same at all times when they're running as designed, whether loaded or not. The internal losses are disposed of as heat, but they are quite small; generally less than a couple of percent of the full load power.

In particular, for power transformers, it will not harm the transformer to run it unloaded (or loaded, for that matter) as long as the signal fed to the primary is (1) at no lower frequency than the transformer is rated (2) no higher voltage divided by frequency than the rating.

Since power line frequencies are fixed, number 2 translates to "no higher voltage than rated". Any voltage lower than rated is OK. It's not widely known, since you can't easily change the power line frequency, but a transformer can be run at twice the input voltage if you also double the input frequency; it's power rating goes up about 2:1 there as well. This is why switching power supply transformers can be so tiny.

Iron cores and copper windings can absorb a lot of heat. The time to reach thermal equilibrium for a substantial sized transformer can be hours.

No matter how you hook them up, the voltages on primary and secondary are related by the voltage ratio, Vprimary/Vsecondary. They're perfectly happy with a lower voltage on the primary. They're perfectly happy with the rated voltage on the secondary, or less. And they're perfectly happy (themselves) running without a load, although whatever is driving them may be unhappy. A transformer run backwards just makes the proportional voltage. Here's an example.

Let's say you have a 120Vac to 18Vac transformer. The peak voltage of an 18Vrms sine wave is 1.414*Vrms, or 18*1.414 = 25.45Vpk, and that's what DC this transformer will give (minus losses) if you rectify the secondary.  The voltage ratio is 120/18 = 6.67 from primary to secondary. So any signal less than 120Vac and faster than 60Hz will give 1/6.67 times the input signal at the secondary. If you put in 60Vac from another transformer secondary, the example will put out 60vac/6.667 = 9Vac on the secondary (ignoring losses).

It works the other way too. If you put in 9Vac on the secondary, you get 60V on the primary. If  you put in 10V on the secondary, you get 66.7V on the primary.

What if you put in 120V on the secondary? Will the primary go to 120*6.67 = 800Vac?

No. The reason it won't is that the volt-time product on the secondary is too big. This transformer is designed to withstand and work with a volt-time product proportional to 120Vac and 60Hz. It will be proportional to 120*0.0167 ~ 2. It will have a little bit of safety factor, so things don't fail at V*t = 2.0001, but it won't have infinite headroom. Notice that the volt-time product on the secondary is 18V and 60Hz, or 18*0.0167 = 0.3. If there is a 30% safety factor, things start going ugly when you get to 0.3 volt-seconds times 1.3 = 0.39, or a voltage of 23.4V.

What happens there is that the volt-time in the core saturates the core. The magnetic domains are all lined up, and the core can't absorb any more magnetic field. Whatever else happens, it can no longer transfer [approximation warning!] voltage or current to the opposite winding. The voltage on the opposite side is clipped when the M-field exceeds the saturation value. The inductance of the windings can no longer limit the current coming into the primary as it does before saturation, and smoke begins to curl out of the transformer.

But I digress... 

It's OK to run a transformer unloaded as long as it's at or below rated voltage and at or above rated minimum frequency.

[arawn]

Just to reiterate 317's should be out. From personal experience they are rated at 1.5 amps of current, and I had problems keeping them alive at 1.2amps of current. They also arent real fond of multiple load lines.

[R.G.]

Just to reiterate 317's should be out. From personal experience they are rated at 1.5 amps of current, and I had problems keeping them alive at 1.2amps of current. They also arent real fond of multiple load lines.

I'm perplexed. You mean LM317s? I use them all the time and find them very reliable. However, they do shut down when they get too hot, so unless they're on a heat sink that can get all the heat out of them, they shut down before they hit the separate current limit. The more excess voltage they have to wipe off, the more power they have to dissipate, so the hotter they get and the better the heat sink they need.

Can you expound on "multiple load lines"? Do you mean multiple loads?

[arawn]

yes multiple loads. and heatsinking did not seem to help me and i even went so far as to using thermal grease before bolting heat sinks in place. Course i could have got a bad lot of lm317's my last boss pulled them out of a drawer and dropped a handful in my  hands, said go play with these!

[swinginguitar]

As a side question on me testing the trafos, how much voltage/amperage can the typical radio shack type breadboard handle?

[arawn]

that's a good question, I just looked at so me of the breadboard i have and it does not specify any volt/amp handling characteristics.

[mattthegamer463]

As a side question on me testing the trafos, how much voltage/amperage can the typical radio shack type breadboard handle?

Not a lot/not a lot.

I wouldn't push it beyond a few amps at any voltage, and 50V.  The main issue is the separation between each conductor row, the voltage can jump through the inside of the board.  I will confess I once built a 250V tube PSU on a BB and it didn't short.  If you do such a thing make sure you don't use adjacent rows with high potential, spread it out a lot.  And be careful.