Riaa PSU ripple (single rail)

[brujo]

Hi fellas, I am Building this simple riaa preamp https://paulinthelab.blogspot.com/2014/05/riaa-phono-preamp-stripboard-layout.html?m=1
I am powering It with 12VDC, I know that It would benefit a dual Power line, but since I am trying very cheap solutions I would stick to single dc. Battery It Is quiter obviously.
A 7812 before the board with two cap would benefit It?
Any advices?
Cheers

[antonis]

A 7812 needs 14VDC (at least) to work properly..

I'm sure you know that you have to replace some of GNDs with a half-supply voltage point..

[m4268588]

C5 connect to Vref. Do not connect to GND.
(and coupling cap?)

Can PSU be occupied for EQ?

[Rob Strand]

For that circuit it's probably best to wire all the input ground, output ground, C1, C6, and especially the two C5's back to VREF.  Keep all the VREF connection antonis shows.    We are dealing  with small signal so it's best to have all ground references going to VREF.

You will be much better off with a regulator.

As I recall that RIAA circuit doesn't agree with the RIAA standard.  Quite crap off hand!

There was a thread on the forum discussing RIAA equalization around 2017, it might be in the lounge part of the forum.   This circuit might have even come up.

I could suggest some better values but these things can be a little finicky when using only E12 parts (1k0, 1k2, 1k5 etc...).  I suspect the old thread resolved the RIAA accuracy issue anyway.



Here's a quick comparison.   There are four examples:
- Reference response based on Wikipedia values
- Response from old National Semiconductors app note.  Very close to reference.
- The original response of the PITL circuit you posted.
- A hand modded version of the PITL ckt which pushes the response closer.
   (Don't consider this as optimal.)   There's a parallel cap 3n3 // 220pF.

What it shows is the PITL circuit is way off.

I only did a quick hack here but you should see the problem.




Here's V2 (MOD2).  It uses a different form of RC network which just happens to work out a little better.
(You need to consider controlling part tolerances at this level of matching.)

[PRR]

A phono preamp with midband gain of 150, working on 12V supply, will clip a lot.

There's no hard limit on what can come out of a cartridge.

I used to do gain of 50 and 30V supply. Even then a few hot platters had a few clipped transients, but it was rare.

The response errors you show are avoidable, but I have seen FAR worse. And didn't complain at the time.

[Rob Strand]

A phono preamp with midband gain of 150, working on 12V supply, will clip a lot.
I used to do gain of 50 and 30V supply. Even then a few hot platters had a few clipped transients, but it was rare.

Yes, the gain probably should be less than 60 (I started off just fixing the response as an example, and only knocked the gain back a bit to (107 in MOD2)).
200mV peak is a safe figure, that pushes the gain way down to 30 for the 12V single supply.

If the preamp is plugging into a PC then even a single supply 12V could clip the line input.  So we don't
need the gain anyway.

Probably don't want to increase the gain resistor any more than 1k so the network needs to be rescaled.   The
dual parallel RC form usually scales reasonably well if you allow for one parallel cap.

The response errors you show are avoidable, but I have seen FAR worse. And didn't complain at the time.

A less of than 1dB tilt is a good target, 0.5dB better.  IIRC the original has nearly 4dB tilt.


Here's a commercial unit which runs off 12V.
Gain is 100 and max in 50mV peak.  So it doesn't allow much for peaks.
It has a volume control on the unit.

https://pro2.com.au/26226/allproducts/avaccessories/turntable-accessories/pre-amps-turntable-accessories/pro1387-riaa-phono-preamp-with-aux-input/

No knowing if this thing has an internal voltage doubler or similar.


FWIW, the previous MOD2 circuit has a gain of x107 at 1kHz

Here's some reduced gain alternatives.   I'll leave it to people's discretion what gain they want.  The 1kHz gain is the value in the last E section, so x34.45 for MOD3 and x42.98 for MOD4. (Don't forget the RC network is different to the original circuit.)

[PRR]

The response errors you show are avoidable, but I have seen FAR worse. And didn't complain at the time.

A less of than 1dB tilt is a good target, 0.5dB better.  IIRC the original has nearly 4dB tilt.

I know what is good, now. Back in the 1960s even high-class gear was full of 2dB errors sometimes spanning 4dB overall. As the OP is looking for low investment, that may be OK.

Another remarkably popular phono preamp, with awful low supply voltage, was the Dynaco PAT:
Dynaco PAT-4 Phono Preamp Schematic with DC voltages

Transistors can be any NPN Silicon with high hFE, the more the better. For 12V you can try bumping R4 to 470 ohms. Q2 collector should then be about 6v. For less splatt in loud passages R2 can be 220r.

[merlinb]

Nothing wrong with a single-rail phono preamp, people have been building them for donkey's years.

[m4268588]

It's my experience using a pickup output volts 0.4mV(1kHz 5cm/sec.) and 40db at 1kHz phono amp.

• 1kHz 50mm/sec. reference signal = 112.6mV PtoP
• MAX peak in general vinyls = 0.6V P-P

Set to t4=7.96u(20kHz)
(t1=3180u, t2=318u, t3=75u)
R3 & C3 may not be necessary. I don't know.
R0 you need to set. You can change gain without change to frequency response using divider.


Passive CR