Why use that old noisy regulators then? 78xx/79xx really have no place in audio anymore with their at least 40 µV noise and 3V dropout.
Using them is just keeping to what one knows.
There seems a tendency in general to go back to mediocre stuff and mediocre design choices of 40 years ago. This while such things have been solved/superseded for decades. Today I see electrolytic caps in the signal path while film caps are meanwhile the same size, I see lacking mute relays and people suddenly liking their woofers to fart, no DC protection of speakers as apparently no one needs that anymore and now a new design with 78xx/79xx. I am sure someone will pop up to suggest to use dinosaurs LM317/LM337
Both indeed are beter.If quality is a goal then there really are much better regs out there. TPS7A4700 and TPS7A3301 for instance.
edit: I see that a version is made with LM2941. Well it is just as noisy as 78xx. The negative version LM2991 is a famous one, it has a typical noise figure of 200 µv and a maximum of 450 µV!!!!!!! You should don't even want to use those.
I suggest people that need a good symmetric PSU look here:
VRDN: bipolar regulator PCB for line level ckts: ±11V to ±20V @ 1.5A with "De-Noiser"
Last edited:
Gaz2613, imho it is in your best interests to take on board the critique that can come from diyaudio. We all learn things from others, and that is often fed in to the next activity/project we do.
We are now in the days of digital measurement advances that allow a much better awareness of what ripple and noise is on the output of a power supply. A scope trace may look fine for high level ripple that is observable in a trace, but a spectrum plot will show so much more detail than a flat line on a scope trace. Your board can allow swap-out of diodes and regulators for lower noise outcomes, as well as bypassing redundant RC filter stages, if so desired.
There are very large threads on quasimodo snubbering of rectification induced noise that would be worth checking out on diyaudio, and I'd reckon many would like to use a C-RC snubber across the transformer secondary if that capability was available, or if you had done that measurement work.
We are now in the days of digital measurement advances that allow a much better awareness of what ripple and noise is on the output of a power supply. A scope trace may look fine for high level ripple that is observable in a trace, but a spectrum plot will show so much more detail than a flat line on a scope trace. Your board can allow swap-out of diodes and regulators for lower noise outcomes, as well as bypassing redundant RC filter stages, if so desired.
There are very large threads on quasimodo snubbering of rectification induced noise that would be worth checking out on diyaudio, and I'd reckon many would like to use a C-RC snubber across the transformer secondary if that capability was available, or if you had done that measurement work.
That comment shows a fundamental lack of awareness about how to route wiring in capacitor filtered power supplies - a standard technique described in any good power supply article/book from the 1940's onwards.
Last edited:
I have no intention of purchasing expensive equipment when I make these primarily for my own sole use. If someone wants something more professional then they can feel free to look at alternatives. I simply laid out the tracks in the CRC filter exactly as on the Whammy. So if it's wrong then it's an electrical engineer that's done it wrong. I try and do as much research as I can from people more knowledgeable than myself but my aim here is to just make a simple quiet power supply not reinvent the wheel. It serves its purpose and I am happy with it's performance. I have not had a single person yet come to me and tell me they are not happy with their board.
Will I make improvements and tweaks to designs in the future based on feedback? You bet I will as how else will I learn? DIY is fun right? Love it! And I especially love to try new things.
Hi please reread on LM2991. You can almost not choose a worse regulator for audio. Even 79xx is way better than that.
Even if you can not measure noise you have the freedom to choose low noise regulators. If you don't have the equipment to measure it the noise will be there regardless.
- If you lack the skills to solder SMD then LM317/LM337 is probably the best combo for a relatively low noise symmetric PSU. Make sure to decouple the ADJ pins.
- If you do have the skills to solder SMD ICs then there are enough nice regulator ICs to choose.
Every audio device will perform better when the PSU is excellent. Please keep the CRC arrangement, even better would be CLC.
Even if you can not measure noise you have the freedom to choose low noise regulators. If you don't have the equipment to measure it the noise will be there regardless.
- If you lack the skills to solder SMD then LM317/LM337 is probably the best combo for a relatively low noise symmetric PSU. Make sure to decouple the ADJ pins.
- If you do have the skills to solder SMD ICs then there are enough nice regulator ICs to choose.
Every audio device will perform better when the PSU is excellent. Please keep the CRC arrangement, even better would be CLC.
Last edited:
Hi, I might have a few left. If not I'm sure I still have some REV A boards though.