Classic Audio '2068' phonostage development!

A very off-the-cuff 9 minute self-satisfied ramble about the events of this afternoon! Once in a blue moon a wacky idea pays off and gives a much better result than expected after a bit of tuning. This needs to be a Classic Audio product!

[@danniielle]

Love your work my friend! 😎

[@terrydemol5354]

Hello Michael, 2068 is a decent OPA - OPA1612 is worth a try, lower noise and distortion. Might be a little more finicky with PS bypassing.

[@michaelfidler8218]

Not suitable for MM inputs, as it has about 1.7pA/sqrt hz input current noise density. 2068 has 0.3, so much lower noise when used with a high-impedance source such as an MM cartridge.

[@terrydemol5354]

@@michaelfidler8218 Yes, you are on the money, 2068 has a sweet spot combination of fairly low voltage noise and current noise. TI do have a couple of BJT IP similar and also the Cmos OPA1656 but the 1656 has a lot of LF noise - otherwise really good OPA.

[@hoobsgroove]

But you've got a type of regulator there what is that a half bridge using the diodes, how's that better than using a dedicated regulator less forward current? Is it a real issue tho using say N10005? Can't you just have greater capacitance to make up the difference?

[@michaelfidler8218]

Lower quiescent current, no dropout margin + excess required.

[@hoobsgroove]

​​@@michaelfidler8218right does that lower the noise floor as well make it more linear than low switching? I can't understand the capacitor set up you've got looks like some of the negatives are going on the positive rail how is that working? You use them as bipolar set up? Actually you might be able to give me some advice if you don't mind!! I saw a valve buffer stage by musical fidelity x10d to work with CD, and I had to query three of the electrolytic capacitors was connected to ground to do with the power supply from AC 12 volts, they all shared the same ground plane, but three of the capacitors where reversed meaning plus was to negative ground, to me that doesn't make sense, they weren't set up as bipolar but they were working for a B- power rail power supply when I looked at a circuit diagram. But still can't seem to work out why an electrolytic would have its positive connected to ground ? even if it's the minus of the B- I had a look on forums and nobody's mention anything about this assume it must be correct?? If you're having a b- power rail do you have the electrolytics in a reversed priority like that? If you could advise me on that be great thanks!

[@michaelfidler8218]

@@hoobsgroove the noise floor on an amplifier output connected to a linear supply tends be lower as any noise will probably be close to mains hum (low frequency) as opposed to switching hash (high frequency) which most amplifiers are much more effective at rejecting. A single transformer winding is half-wave rectified into a positive and negative voltage to create a split supply (michaelfidler.com/articles/practical-vu-meter-circuits/#8) like so... If you have a negative voltage then electrolytic capacitor positive terminals will reference to ground (otherwise they'll be the wrong way around). It's pretty standard practice to do this for reservoire/filter capacitors.

[@hoobsgroove]

@@michaelfidler8218 ok so the capacitors are correct when it comes to negative voltage B- the positive becomes ground in that case, so they're not the wrong way around on the x10d ok thanks clears that up. Is that what you're doing with your power supply with the capacitors? the adjacent capacitor is flipped. Oh I see capacitors are flipped because it's a split rail supply on yours, isn't that more difficult too reduce means ripple? But if you use a one ferret polyurethane capacitor after the diode in conjunction with the electrolytic would that give you enough smoothing? One thing I'm not sure about is how you get good smoothing from the transformer? I have to read the article you posted in more depth Cheers for that.