Should I Worry?

[Jonaz]

I recently built a variable bench power supply that can deliver 9V-23V at 300mA. It works great and maintains good voltage regulation even under loads over a 100mA. However, the 20VCT Triad FP20-300 transformer I'm using is giving me 35.6V with no load after rectification @ 120V wall voltage. It's got me slightly concerned that it's hovering too close to the 40V input maximum of the LM317T. A small increase in wall voltage and...ZAP!!! Even under a load of close to 200mA, the voltage after rectification only sags to 33.9V.

My question is: what can I do to lower the voltage before it hits the input of the LM317T? I would feel safer if I could get it down to about 30V. My solution at the moment is to string a bunch of 1N4003 diodes in series with the PT secondaries before rectifying the AC but I'm sure there's a better way. Any help would be gratefully appreciated!

[Cliff Schecht]

You put a resistor in series with the input of the regulator to drop a certain amount of voltage. So you want to drop, say, 10 V across the resistor and still be able to deliver 300 mA of current? Some quick Ohm's law magic will tell us that to drop down to 30 V with a 40 V input at 300 mA, you do 40V minus 30V (giving 10V) and divide this by 300mA (or .3A) to end up with about a 33 Ohm resistor.

I think that you are being a bit of a weary of the wall power though which, from my experience, is usually around 117V-120V RMS. I doubt that your transformer/diode regulator will ever get enough voltage from a US wall socket to hurt anything. I think that a more reasonable value for your circuit can be found by using your real-life results: (35V-30V)/(.3A)= 16 Ohms. 15 Ohms is probably an easier value to find though, so we'll choose that.

Something to consider here is that this resistor is strictly in place to burn up power in the form of heat. A little more Ohm's law says that 15 Ohms*(.3A)^2 gives you about 1.35 watts of power dissipation at your maximum rated current. I'd put in a 15 Ohm, 2 Watt resistor and wouldn't think about it anymore .

[JKowalski]

I dont think you should worry. 40V on the LM317 isn't going to instantly destroy it, and the wall voltage stays pretty stable. If you are worried about transients, you could put some line filter components before the transformer, or buy one of those pre-filtered power entry modules. Or maybe put a ~40 volt zener before the regulator to dissipate transients on that side... 

[earthtonesaudio]

You might also check out National's appnotes for the LM317, they have at least one example of handling high input voltages while delivering modest output voltages.  Much more complex than a resistor though.

[darron]

i don't think that you should worry about it. there's a lot of headroom there now for an LM317T now!

if you did go with the 1n400x diodes in series i would think it would make more sense to put them after the rectifier, rather than before? but either the resistor or zener diode seem better ideas.

i had similar worries when i built my supply, but my voltage wasn't that high. i figure if something goes wrong then i'll replace the regulator and then put measures in to prevent it, nothing yet (:

[Jonaz]

Thanks guys, great suggestions! I think I'll just shoot for the series resistor then to drop some juice. Much obliged!

[Paul Marossy]

I think that you are being a bit of a weary of the wall power though which, from my experience, is usually around 117V-120V RMS.

Probably so. But in my neck of the woods, it can be as high as 125V in the summer when every air conditioning unit in the city is on and it's 115 deg F outside. I don't understand exactly why it's that high in the summer, but I have measured that before.

[GibsonGM]

Mine gets high like that, too.  I'd recommend the series resistor.  $.75 part, and then you don't have to dig into it again if it fries the regulator!  Just be conservative with the power rating and you're all set.