Moving Coil Prepre PSU: Basic Questions, What am i missing?

I won't lie, most of the time when i built DIY audio stuff, i'm blindly following ready made schematics. So i did 20 years ago when building my 2sk170 PrePre and the following RIAA stage. Now, over the years i have gathered crates full of electronic components stuff, often MIL spec parts, which i wanted to use some day to rebuild things that actually work, but have some flaws, maybe only optical, sometimes it was just me modifying into a wrong direction because i didn't have parts at hand, no money or whatsoever.

Now the time has come, the amount of parts in my garage is getting annoying. So i decided to "build it away" and get my final stereo setup done.
One of the DIY devices that has a flaw is the PSU of my phono amp. It humms a little bit. It is basically only LM7824/LM7812 with some filtering caps. So i thought to myself i had to go look into the crates what i have.
What i found is a nice little Collins choke with 2,5H/100R at 0,11A. I thought it could be a good idea to use it with a decent 30V toroid, some 4,7mF caps and two LM317 regulators for creating a dual power supply. The needed voltages are +12V@40mA and +24V@60mA.

I turned on PSU designer to simulate that and... wow... i'm getting 30V @ 1,4mV of ripple only? Can that be correct? I mean plus the additional rejection of the LM317s which should be around 60dB for each stage that should make a very quiet phono PSU in the nanoVolts region right?
Now what i would do is adding a LM317 stage to regulate for 24V for the MM stage (approx 60mA) and behind it another LM317 to regulate for 12V for the MC stage (approx 40mA). Maybe add another RC in front of the second LM 317. Should take off some heat from the reg's shoulders and should reduce ripple even more. Maybe something like 20R with 4,7mF or so.

simple choke PSU_2.JPG


Ok, so i'm not the smartest guy when it comes to designing circuits (and living with women). So i have some basic questions.
Everyone is building shunt regulated PSUs etc. No doubt they work good, but for me they are hard to understand and build. If i interpret PSU designer correctly, i wouldn't need more than these few parts plus etching a little PCB for regulators etc to get away with it? Does that mean all the effort put in shunts, low noise regs, huge filter caps etc is only for cost reasons because of rather expensive chokes? I mean actually you can get a choke with these values around USD 10 to 20 used or so. Can this be true?

Also another basic question: I would like to build a test setup. Do i have to worry inrush current could destroy the choke? Maybe should i add a current limiting resistor?

Thanks for your help!
 
...shunts...
Shunts are different from series regulators in that shunt output impedance tends to fall as frequency increases (at least up to some frequency), whereas output impedance of series regulators goes up as frequency increases.

Regarding regulators such as LM7512, and or LM317, in can be that just one of those is not enough to reduce low level AC line ripple to very low values. Using two regulators in series, or using something like RCRC or CRCRC before one such regulator may do the trick.

For old fashion LC power supply design info, you might look starting at page 1242 of the Radiotron Designer's Handbook at: http://www.tubebooks.org/books/rdh4.pdf

Beyond that, sources of hum don't necessarily have to be only as shown on a schematic (also some info on that starting at page 1246 of the above book). For one example, in some cases there can be stray magnetic coupling between a power transformer and a filter choke.

EDIT: From Martin Mallinson Patent:
...compared to a series regulator in which the output impedance increases with frequency, a shunt regulator has a lower output impedance as frequency increases...
https://patents.google.com/patent/US9383762B2/en
 
Last edited:
First off, i am not sure your current equipment hums because the 3-terminal regulators do not reduce sufficiently the 100Hz ripple. They do this very well. Adding a choke will change the perceived sound a bit but may not solve the hum issue, ripple is probably low enough already.

It certainly helps keeping some batteries around to solve issues of this nature. An 18v screwdriver battery is perhaps perfect for the purpose but even an old car battery is usable despite the low voltage.

Shunt regulators generally sound less intrusive than series. They seldom offer more ripple rejection or lower impedance but seem to react better to transients in the current draw. A lot of people simply prefer them for sonics without expecting any objective improvement. A resistor + zener is already a shunt, try it, you may be surprised.

For the better part of my hobby i have used chokes followed by regulators. In subjective terms chokes vary hugely in quality.
 
Thanks for answering to both of you.
I'm aware of other issues like grounding that can produce hum. That shouldn't be a problem with the current PSU. It's more or less only a LM7824 and LM7812 in series with 2200uF C filter in front. Not even a RC filter. I just didn't care enough.

Batteries:
That is actually the reason why i built a PSU. I was running it with three 12V Motorcycle LiPo batteries before. The preamp and the batteries sat in one of the crates in my garage for like 6 years, the batteries were dead (actually i found charging them annoying anyways). So i thought i had to get the amp going at first, using what i had at hand, now realizing i should put more effort in the PSU.

As a conclusion from your thoughts i'm leaning towards just using CRC filtering instead of CLC. Or using CRC with two LM317 in series to reduce ripple even more. I'm not sure how a choke can have an impact on sound/what impact a choke could have on sound. From reading i know that PSUs can introduce hiss. And my initial thought was that silicon could be the culprit, when i look at oscillioscope screenshots with slightly distorted sines because of switching. So in return i thought about building an unregulated supply, but then read that noise floor could be a problem because of resistance of CRC (i need to go down from 30V to 24V to 12V), what led me to chokes as the most desirable option (according to many writers on the web, but mostly referring to tube amps).