What is the simplest and most efficient way to reduce ±60V to ±15V?

[Michael Bean]

Google Translate seems to be unable to fully express my thoughts, haha. I plan to get a voltage of ±60V from the main power supply of the power amplifier because I don't want to add new windings to the transformer. While maintaining good performance, it needs to be reduced to ±15V in the simplest way. At least 200ma current is used to drive 5-6 operational amplifiers. It has a huge heat sink, so it can get good heat dissipation support, and I don't care about power consumption. Based on this requirement, I first thought of lm317/lm337 because it can provide good enough performance and is simple enough, but its maximum withstand voltage cannot meet the requirements. I thought of using two lm317s in series to solve the power consumption and withstand voltage problems, and considered using TVS to protect the lm317 from accidental damage. I don't know whether the whole circuit is strong enough after using TVS, or whether the operational amplifier will be damaged by accidental ultra-high voltage. If this idea is feasible, can lm337 do the same? This is my initial thought.

What opamps are you using that draw more than 30 mA each?

Mike

 

[PRR]

There is no elegant solution.

Simple voltage-drop of 40V at 200mA is a LOT of heat to dissipate, twice (+/-).

Switchers need careful selection against supersonic interference.

I too question the 200mA for a small number of opamps. 50mA might be generous, and significantly less waste heat.

Most opamps do not *NEED* regulated supplies. They only need to never see over +/-18V (and not under whatever the output needs to be). For most small-signal audio a little R-C filtering cleans the ripple and crosstalk well.

The plan below is fairly "safe". The transistor plus Zener part will keep everything after down below 20V, safe for LM7_15, safe for 36V opamps for momentary events. The 47r resistors limit the peak charging current and also take some heat off the transistors. Transistors should be 80W TO220 types on quite large heatsinks (like a 486 CPU but without the fan).

Actually I would change the Zeners to 15V and omit the regulators. Fender guitar amplifiers did that, simpler, a LOT. (With long-term trouble, which is why I prefer transistors and generous heatsinks.) If the load is really like 50mA, then 10r and 1000uFd filtering will clean the rails well.

 

[bansuri]

for 100mA
good hum and noise filtering (can be improved with darlington)
very low heat in semiconductors
very stable (no external feedback)

short proof

very low noise



You can put another 80V capacitor after the 330 ohm for better filtering

 

[bansuri]

some diagram





Output resistance 1.5 ohm