1
\$\begingroup\$

I have a regulated power source supplying +/-48 VDC that I would like drop to +/- 24 VDC. I think linear regulators are the way to go here since they are powering an audio circuit. I would like to use a pair of regulators, such as 7824 and 7924, but those particular ones have a max input voltage of +40 VDC and -40 VDC.

Are there similar regulators with higher maximum inputs? Nothing shows up on Mouser. If not, am I correct in thinking the best option would be to use transformers?

Alternatively, the device to be powered also has regulators at the input (can't quite see their model numbers, which are obscured by heatsinks), so maybe some sort of resistor-based solution would work?

edit: Since there were some questions about the power supply: it's not a standard power supply. The whole system is a guitar combo amp that was essentially built from a rackmount preamp installed into a powered speaker. The power amp had a +/-48V tap that was powering the preamp. I wanted the option to use the preamp separately, so I converted the preamp to use a separate +/-24V power supply, which had the added benefit of removing some resistors that were overheating and burning the circuit board. However, when using the whole system as it was originally designed, I'd like the option to power the preamp from the 48V power tap.

The designer recommended jumpering those resistors to convert the preamp to 24VDC. I can't tell how those resistors were originally dropping the voltage, but I'd rather not use a similar solution to convert the power amp's output to 24V. Some commenters already pointed out how inefficient that is.

\$\endgroup\$
12
  • \$\begingroup\$ I mentioned resistors because that's how it originally worked. The device originally took +/- 48 VDC, then using cement resistors to drop the voltage before three voltage regulators. However, I cannot tell exactly how it was wired. Also, at some point in the future I may want to power other devices from this tap, so I'd rather use regulators or transformers. \$\endgroup\$
    – meanerelk
    May 3 '20 at 1:16
  • 2
    \$\begingroup\$ linear regulators are probably not the way to go unless you only need 20mA or less. Too much waste heat from burning off an excess 24V. Doesnt matter what the linear regulator datasheet says about current; those are optimistic numbers. Transformers only work on AC. Resistor are never the way to go. Required current is missing. \$\endgroup\$
    – DKNguyen
    May 3 '20 at 1:32
  • 1
    \$\begingroup\$ I feel it would be a lot easier to make +24/-24V power supply. \$\endgroup\$
    – Leoman12
    May 3 '20 at 2:14
  • \$\begingroup\$ @Leoman12 agreed. \$\endgroup\$
    – DKNguyen
    May 3 '20 at 2:20
  • \$\begingroup\$ @DKNguyen Current is about 0.4A (at least that's what my benchtop power supply reported). That makes sense, thank you. There was definitely a lot of waste heat in the original design, which is why I jumpered the resistors from the PCB. At least if the new solution produces that much heat, it won't be burning the board. \$\endgroup\$
    – meanerelk
    May 3 '20 at 2:24
1
\$\begingroup\$

I suggest a separate power supply or two DC-DC converters as the easiest solution.

For audio you could use simple emitter-follower zener regulators, but the problem of burning off heat will still be there. Eg. (Q1 could also be a Darlington pair)

schematic

simulate this circuit – Schematic created using CircuitLab

The same circuit, flipped with a PNP power transistor would be used for the negative supply, and resistors could be added in the collector to burn off some of the heat. The calculations depend on the current requirements. It's easy to find TO-220 transistors or Darlington pairs good for 60V or 80V.

It's not inherently short-circuit proof and the efficiency is terrible. You could also use these as pre-regulators for 3-terminal regulators but that doesn't help much.

The 24V suggests a relatively high power amplifier, which means the horrific lack of efficiency is going to be an issue.

\$\endgroup\$
7
  • 3
    \$\begingroup\$ I hereby coin the word: horrificiency. \$\endgroup\$
    – DKNguyen
    May 3 '20 at 2:09
  • \$\begingroup\$ Thanks for the information! I am also building a separate power supply for the preamp, which will definitely be easier. But it would still be nice to be able to use the power amp's rails when using both units together, to avoid the inconvenience of carrying a separate power supply. Current is around 0.4A. I have been looking for DC-DC converters, but that seems to be a broad class of components. Do you mean something like a buck converter? This one looks like it might be close to what I need, at least for the positive rail. \$\endgroup\$
    – meanerelk
    May 3 '20 at 2:21
  • 1
    \$\begingroup\$ @meanerelk Yes, a buck. but those are not trivial to implement. Find a DC-DC converter module where it is already done for you. \$\endgroup\$
    – DKNguyen
    May 3 '20 at 2:30
  • \$\begingroup\$ For a one-part solution, look at a Meanwell NSD15-48D12. 48V->+/-12V at 600mA. About $16. \$\endgroup\$ May 3 '20 at 2:35
  • \$\begingroup\$ @DKNguyen Will do. I'd definitely rather buy than build. \$\endgroup\$
    – meanerelk
    May 3 '20 at 2:42
0
\$\begingroup\$

If you really want to get +48V/-48V to +24/-24V there is a way (possibly).

For the 48V to 24V you can use a buck converter such as TPS40200 which accepts input voltages up to 52V. There are varieties on Texas Instruments websites under buck converters.

For the -24V output you can use a buck-boost topology that would convert the 48V to a -24V. Here the -48v wouldn’t be needed.

Another possible way is to modify a normal buck converter to obtain negative voltage. It’s explained here

https://www.vishay.com/docs/76946/creatingnegativeoutputvoltage.pdf

And here

https://www.analog.com/media/en/reference-design-documentation/design-notes/dn1021fa.pdf

And also here

http://www.aosmd.com/res/application_notes/power-ics/pic-002.pdf

To obtain -24V from 48V, it might be possible to use TPS40200 using information from those documents.

Note: this method requires design of buck converters and PCB design.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.