RIAA pre-amp for Moving Magnet cartridges

It's an ESP Project 6 followed by a Project 99, with some changes.

Firstly, C2 and C3 of Project 6 looked wrong to me, with + connected to the common rail. If you google the basic schematic, you will find versions with the caps the other way 'round, and versions without the cap. These capacitors are there to compensate for any DC offset in the input signal or from imbalances in the op-amp -- ideal components don't exist. One can use either a series capacitor (in series with the signal, and therefore potentially degrading said signal) or this system, where the gain at DC becomes one. But the signal from the pickup has no DC component, and in my case the op-amp errors give me around 40mV DC at the output of the RIAA stage -- so in my opinion these caps are not needed.

And of course if you're connecting the two stages together you can get rid of some of the intermediate components, like C5 and R9.

Also, I ended up using 100nF capacitors in the rumble filter (more below) for a 27Hz -3dB frequency.

Assembly

This is a perfect project for perfboard. Then I saw the price of perfboard. So I made a plan. I had some PCB material lying around, so...

I laid out a PCB in KiCad, printed it as seen from the top, stuck it to the copper side of the PCB, and drilled all the holes.

Then countersunk the holes not connected to 0V.

Then I printed the copper traces as seen from the solder side (reversed) and stuck that to the non-copper side.

OK, in hindsight this was stupid. I could have just stuck the copper layout to the non-copper side, drilled all the holes except the ones connected to 0V, and then countersunk them from the copper side. Next time.

Because as you can see, all the 0V connections are to the copper on top.

And on the bottom I just persuaded the component leads to follow the path the copper would have, had this been a real PCB. It's a bit like this but more robust.

No, I am not an artist like AA7EE but it'll do.

So here's the thing. About half-way through I decided I was going to use only components I already had. Buy NOTHING. Maybe I should have sprung for metal film resistors in the places where I only had carbon in stock, or for modern op-amps instead of the LM358/MC1458/TS272/LF412 options I had, but hey. Also I cheated, using a bunch of 100nF 250V capacitors I had (150nF is a funny value, not many people stock them), this raises the rumble cut-off frequency, and they are also fairly large (I had already drilled the PCB by then) but they will work, so there it is.

The box

This is a Henrian Electronics MME-2. I have no idea what its purpose is. It generates 64, 128, 192, 256, 320, 384, 448, 512, 576, 640, 704, 768, 832, 896, 960, 1024, 1088, 1152, 1216, 1280, 1344, 1408, 1472, 1536, 1600, 1664, 1728, 1792, 1856, 1920, 1984 or 2048 kHz from a 4.096 MHz crystal and a couple of CMOS dividers. Then there's an 8031 with a little bit of code to configure the counters and to drive the display. And a whole lot more power supply than I would'a thunk is needed, with two 3VA transformers, 8V and 2x12V, generating 5V twice and +- 9V, which would make sense had this been RS-232 but the line drivers are 26LS30. There's a TL7705 supervisor, an XLE93C46 EEPROM, and some driver stuff for the 7-segment LED.

Anywayz, the box is about the right size for this preamp, and the 2x12V transformer is perfect, so this thing is getting recycled.

I ended up going with the one LF412 (25nV/root Hz noise) I had for the first opamp followed by three MC1458s (45nV/root Hz*). I also had a TS272 (30nV/root Hz noise) but it was bad. The NE5532 that ESP uses is low noise at 6nV/root Hz but (1) I don't have any in my junkbox and (2) I used sockets, if I think it's a problem, I can fix it. For now, I don't hear any noise, so I'm good to go.

* They don't measure the noise in exactly the same way so I don't know how valid a comparison this is.