Author Topic: Is a triac on the primary side of a transformer a good idea?  (Read 8387 times)

0 Members and 1 Guest are viewing this topic.

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12807
Re: Is a triac on the primary side of a transformer a good idea?
« Reply #25 on: May 02, 2018, 02:18:01 am »
I'm revising my best-case estimate of what can be got out of it.  I've zoomed in the photo and done my best to take some measurements of the two secondaries' wire diameters.  It looks like about 1mm for the 20-0-20V  secondary and 0.5mm for the 15V secondary (both with approx 10% uncertainty).   The turns ratio between the secondaries must be 40:15, which reduces to 8:3, but as the  CSAs are in the ratio 4:1 the relative proportion of the two winding areas on the secondary bobbin are 32:3 so the 20-0-20V takes up 91%, and the 15V takes up 9% of the  winding area.   Assuming the VA rating divides with the winding area occupied, and going with my and Jwillis's guesstimate that its a 12VA transformer, that means the 20-0-20V should be good for 0.91*12/40 A or about 275mA RMS and the 15V winding should be good for 0.09*12/15 A or about 70mA RMS.   

Remembering that the conversion factor between RMS current into and DC current out of a bridge rectifier + reservoir capacitor (without a large DC rated choke in between) is 0.62,  that means, using  Oldways suggested LM317 & LM337 regulator pair, its good for about 160mA DC out on both adjustable outputs (because you loose 10MA to the voltage set divider current.

Failure to consider over-current protection is highly likely to result in it failing sooner rather than later.  Of course you can overload it for a short time, but you will cook it if you sustain that.   One answer would be to build it up, run it at 160mA load in its intended case for at least 6 hours, and measure the surface temperature of the top of the core,  get a mains rated, resettable (or maybe even a self-resetting) thermal trip with a trip point slightly above that temperature and put it in series with the primary, with an overload trip warning neon indicator in parallel with the thermal trip. 

However,  LM317/LM337s will typically pass a peak current of 2.2A until their thermal limiting cuts in, and, assuming   they are overloaded with the output voltage wound right up, due to the low voltage drop across them the thermal limiting may *NEVER* cut in if they are adequately heatsinked for 160mA right down to 1.2V out.   Converted back to RMS current and then VA, that's about 142VA or an absolutely ruinous overload of nearly 12 times its normal rating for the primary.   Also the I2 losses will be about 140 times higher than normal, so the primary is likely to burn up before the suggested external resettable thermal trip can react.   The key to avoiding this is IN ADDITION TO THE THERMAL TRIP,  to provide secondary side over-current protection that will act at under a factor of two overload.   If you fuse the secondary with a 250mA  slow blow antisurge fuse, it will be good for 150mA continuous, will provide 300mA for ten seconds, and at 500mA will blow in about a second well before the primary can burn up.

I'd prefer electronic protection but the circuits required get quite complex.

The 15V winding is difficult to say what to do with it.   You could just ignore it, or if you decide to build the 'lab' psu circuit with current limiting from the LM317 datasheet, you could use it with a rail splitter  to provide +/-10V rails for the current sinks needed for that circuit to work.   I've run the numbers and its not really worth using it with a buck module to provide 5V - you'd do better to use a separate low wattage regulated wallwart PSU to power any logic you want to mix with your analog circuits.
« Last Edit: May 02, 2018, 02:20:10 am by Ian.M »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf