60hz hum

[PRR]

> the exact point where the regulator has the biggest power wasted

Half voltage?

i.e.: 27V in, 15V out. 12V total to drop. 10V of available drop. Half that is 5V.

Assume 50 ohms.

At zero mA, R drops nothing, Reg drops 12V, at zero mA, Reg throws zero Watts.

At 240mA, R drops 12V, Reg is under-supplied. Need to leave >2V across the regulator.

At 200mA, R drops 10V, Reg drops the last 2V, 2V*200mA is 400mW or 0.4W. R throws 10V 200mA or 2 Watts.

At 120mA, R drops 6V, Reg drops 6V. Both throw 0.72W.

You can try a few in-betweens but I think with the pure resistor the half-voltage point is max dissipation.

Therefore within (and just past) the "works good" range, the max Reg dissipation is 0.72W. This can be handled with a naked TO220.

The max resistor dissipation, in the "works" range, is 2W. BUT shorts happen. In dead-short nearly all the 27V appears in the resistor. Maybe minus 2V in the regulator. 25V/50r is 0.5A. The Reg will not shut-down on over-current. It dissipates just over 1W so will not shut-down on overheat. (Ah; yes, this is worst-case point, not the half-Voltage point.) So the resistor has 25V 0.5A or 12.5W!

If you are attentive, a 10W part will suffice. If you quickly smell the toast or note all-dead, and unplug ASAP, a 10W part will stand 12.5W for many-many seconds at a time. If you stick it in a closet powered-up, better use a 20W part.

[R.G.]

Half voltage?

That was my first guess, but I was too lazy to do the algebra. It will be near there. I had a thought that perhaps the regulator alone might have a bigger V*I product at a lower current, but - well, I need to do the symbolic math or cheat and sim it.

The max resistor dissipation, in the "works" range, is 2W. BUT shorts happen. In dead-short nearly all the 27V appears in the resistor. Maybe minus 2V in the regulator. 25V/50r is 0.5A. The Reg will not shut-down on over-current. It dissipates just over 1W so will not shut-down on overheat. (Ah; yes, this is worst-case point, not the half-Voltage point.) So the resistor has 25V 0.5A or 12.5W!

I might stick in a separate current limit clamp with a couple of transistors since there is so much overhead voltage available, just to catch that case and avoid paying for and mounting a 15-20W power resistor. One TO-220 device and a TO92, a couple of resistors, etc.

Or use a 78M15, which limits at 0.5A. Or multiple 78L15s at 100ma each.

[WhenBoredomPeaks]

TL;ĐR: rectified voltage is somehow 21VDC instead of 27VDC

I used the single resistor technique and it works.

What is interesting that i do not have 27VDC after rectification. I measured only 21VDC

I have 20VAC out of the transformer, which turns into 21VDC after rectification. The resistor drops about 3.5V so the voltage into the stabilizator is around 17.5VDC.

This is cool but the current draw is only ~60mA on both of the rails. (i used LEDs for "simulating" current draw instead of the synth circuit because i did not want to blow up the synth stuff if something goes bad)

After 30 minutes the stabilizators are run cool, their output is +15V and -15V.1 which looks good. The dissipating resistors (8W, 47ohm) are a bit hot. (you can keep your finger on them though)

If my calculations were right with this low input voltage (21VDC) the voltage into the regulators will be below 17V even at 100mA.

My rectifier is made of UF4001 diodes. Do they drop 6V? Or my rectifier is connected in the wrong way somehow?

[PRR]

Where is the filter cap?

Customarily we wire PT Rect CAP Resistor Regulator.

If you wired PT Rect Resistor CAP Regulator you would expect a DC closer to the AC rms. Also some real ugly wave-forms in the resistor which foil simple computations.

[WhenBoredomPeaks]

Where is the filter cap?

Customarily we wire PT Rect CAP Resistor Regulator.

If you wired PT Rect Resistor CAP Regulator you would expect a DC closer to the AC rms. Also some real ugly wave-forms in the resistor which foil simple computations.

Yeah my resistor is before the cap.

At this point it would be easier to add some caps before the resistor so make it like Pt-Rect-Cap-Resistor-Cap-Regulator than switching their order. Would that work or i have to switch their order?

(it is amazing what can go wrong in a supply, this forced me to do more calculations and stuff than something like an ada flanger, and sooner or later it will be more expensive too...)

edit: i added 1000uF 50V caps before the resistor while letting the 3300uFs after the resistor in their place and now it measures normal values, so it seems like it is working!

is this is an acceptable configuration? (rectifier-caps-resistor-caps-regulator?)

[amptramp]

You can let the free PSU design program on duncanamps show you exactly what to expect:

http://www.duncanamps.com/psud2/index.html

And it runs on some pretty primitive operating systems.