Any ways to improve this LM4562 preamplifier circuit?

The cap on the treble pot is frequency determining and should not be changed, for DC blocking you want it to have no/little effect down to 20Hz, so it would have an impedance 10 times less than other components at 20Hz typically. 4.1k seems the relevant impedance, so 400 ohms at 20Hz for the DC blocker, which suggest 22µF or larger.
 
.... I definitely did not consider the phase inversion that would result from the tone control bypass ... to correct for phase inversion there, would a unity gain inverting op-amp in the bypass circuit be a good way to go?
Yes! I have seen this approach in commercial pre-amp circuits with tone control defeat.

Looking at the volume control, it looks like being inverting as well, so a unity gain inverting OpAmp seems to be the correct solution and keeps the whole pre-amp non-inverting overall.

Considering polarity, though, some/many (especially CD) Recordings are inverted, i.e. may sound different/better depending on pre-amp output polarity, I have seen people liking the ability to switch over all polarity between inverting and non-inverting (a few DACs provide this feature). Well, this may lead to overfeaturing your design, but maybe it's worth a thought...?

Regards,
Winfried
 
  • Like
Reactions: Palmtrees
Hi all,

Thanks for all of your ideas on how to make this preamplifier sound better. Here is version 4 of the schematic with the latest suggestions included...

1706081994713.png


I added DC blocking caps to the volume, bass, and balance potentiometer wipers as recommended by @Mark Tillotson and @epicyclic. @epicyclic recommended 22 uF, but it seems like bigger is better with electrolytic capacitors so I used 47 uF for these caps (C30, C31, and C32). Is this a good capacitance to use here? Capacitors C25, and C26 on the other side of the volume and balance pots are 470 uF, so would 470 uF be a better capacitance for C30, C31, and C32?

The cap on the treble pot is frequency determining and should not be changed, for DC blocking you want it to have no/little effect down to 20Hz, so it would have an impedance 10 times less than other components at 20Hz typically. 4.1k seems the relevant impedance, so 400 ohms at 20Hz for the DC blocker, which suggest 22µF or larger.
Is the position of the DC blocking cap (C32) for the bass pot correct? Does the treble pot need a DC blocking cap on the wiper as well? It already has a 8 nF cap (C13) on the wiper as part of the tone control circuit but is that enough to block DC? Should I add another 47 uF cap closer to the wiper?

One more question, the unity gain buffer in the volume control section has a 100K resistor to ground (R26) between C30 and the op amp. Is there any benefit to adding a similar resistor to the balance control between C31 and U10?

I also added an inverting unity gain op-amp (U15) to the tone control bypass to maintain phase as recommended by @wgh52.
 
Yes the bass pot DC blocker is correct. The treble already has a capacitor on the wiper so its not an issue. All capacitors block DC, otherwise they wouldn't be capacitors.

C31 is in the wrong place, it must be on the wiper - where it is it prevents any bias current to the opamp so the opamp will latchup and not work at all. The rules are every opamp input requires a DC path to bias it, and that path shouldn't go through a pot wiper - that's all. R20 and the pot's track provide the DC path for U10.

Your tone control bypass opamp U15 lacks the two resistors required for an inverting opamp stage, you've in effect shorted the output of U15 to the output of U4, which again will latchup.
 
  • Like
Reactions: Palmtrees
Personally I would keep C31 where it is and add a 100K resistor to ground between C31 and pin 3 of U10 to avoid a varying bias resistance .

As Mark says you have shorted out U15 you just cant copy the non inverting buffer and put the input on the negative , it should have eg 1K in series with the negative pin and 1K around the feedback path and should end up looking like U4 .

You need a 100K resistor to ground on the left hand side of C1 to stop a charge building up over it .
 
Last edited:
PS if you meant to parallel U10 and U11 then the positive pin of U11 should go to the positive pin of U10 . I would put a 470uF cap in series with the final output with a 22K bleeder resistor to earth unless you don't mind connecting DC to the power amp .
 
Ok here's the latest version:

1706159119782.png

C31 is in the wrong place, it must be on the wiper - where it is it prevents any bias current to the opamp so the opamp will latchup and not work at all. The rules are every opamp input requires a DC path to bias it, and that path shouldn't go through a pot wiper - that's all. R20 and the pot's track provide the DC path for U10.
I added a 100K resistor to ground (R29) between C31 and U10 to provide a DC path for U10 as suggested by @epicyclic. Will this be sufficient?

As Mark says you have shorted out U15 you just cant copy the non inverting buffer and put the input on the negative , it should have eg 1K in series with the negative pin and 1K around the feedback path and should end up looking like U4 .
Ok thanks for pointing that out. I added 1K input and feedback resistors to U15... Would there be any issues using lower value resistors here, say 600R?

You need a 100K resistor to ground on the left hand side of C1 to stop a charge building up over it .
That makes sense. The 100K resistor will also change the pre-amp's input impedance right? I calculated that the parallel resistance of 100K and 47K makes the input impedance 32K. Am I calculating that correctly?

Is anything else misplaced or missing? Any component values that could be improved? I feel like it's coming together pretty well and I'm looking forward to building this thing. But I'm still open to any suggestions if the circuit can be improved further.
 
PS if you meant to parallel U10 and U11 then the positive pin of U11 should go to the positive pin of U10 . I would put a 470uF cap in series with the final output with a 22K bleeder resistor to earth unless you don't mind connecting DC to the power amp .
The output buffer is a design shown in Small Signal Audio Design by Doug Self. The op-amps are in series, which is supposed to have better drive capability than a single op-amp. Since I'm driving two sources at the same time (power amp input and active sub input) I thought the extra drive capability would be useful. But please let me know if there is a better way.
 
.... The op-amps are in series, which is supposed to have better drive capability than a single op-amp. Since I'm driving two sources at the same time (power amp input and active sub input) I thought the extra drive capability would be useful....
Well, all that may be true, but the, maybe better alternative could be to have dedicated OPs drive each cable individually! In this alternative you would have two "parallel" OPs connected to C31 "output", each driving one cable. If the chosen OP is the best for cable drive is another issue to potentially be discussed: There are designs out there using much more drive capable BUF634s (which is designed for "driving work") instead and with better sound results. Maybe worth considering...

Kind greetings,
Winfried
 
But I'm still open to any suggestions if the circuit can be improved further.
The warmest recommendation is not to do it. You will have nothing left of the sound with a bunch of OPAs and coupled polarized electrolytic capacitors and multiple unnecessary signal processing. Make a good tube preamp for that money. It will literally be a quantum leap for the sound of your audio system.

The LM3886 amplifier (and many others) with digital sound sources doesn't even need a preamp, just a buffer (unity gain). Classic cathode follower or say Aikido cathode follower (Broskie) get the job done. No gain - no pain.

https://www.tubecad.com/2011/01/blog0198.htm
https://www.tubecad.com/2010/06/blog0188.htm
 
Last edited:
The warmest recommendation is not to do it. You will have nothing left of the sound with a bunch of OPAs and coupled polarized electrolytic capacitors and multiple unnecessary signal processing. Make a good tube preamp for that money. It will literally be a quantum leap for the sound of your audio system.
Quantum leap I'd concur with, but not for the better... This prejudice against high performance opamps is baseless and untenable. All the recorded material you listen too is passed through opamps in the mastering process, and older recordings may have been through many dozen opamps. Far from nothing left of the sound these include all times classics. So take that erroneous statement back please.

Valve circuitry I'd usually avoid for all the usual reasons, expense, weight, unreliability, microphony, high voltages, high distortion, and hum pickup - as a curiosity its a thing, and for high power RF transmitters its the best technology, but not for audio, and certainly not for line-level audio. I've nothing against people building valve circuitry, just don't try to claim its better quality sound reproduction without giving hard data.

The BOM for this opamp circuit is perhaps tens of dollars, compared to 100's or more for valves - no chance of making a tube amp for the money!
 
  • Like
Reactions: wgh52
This prejudice against high performance opamps is baseless and untenable.
This is not prejudice, I speak from experience. For 40 years I have used various tube preamplifiers and buffers, I would never change them for any OPA.
The above schematic (#68) is evil in the house and waste of time. And it doesn't cost that little. 🤣
To make a single buffer/preamp with OPA is fine and cheap, but multiple copying of the signal through so many OPA and other parts is doomed in advance.
A single OPA stereo preamp/buffer, the simplest possible, with a nice metal case, PCBs, a quality (preferably discrete) OPA, switches, connectors, a quality power supply and wiring cannot cost less than $150.
 
Last edited:
I'm with Mark. This is not the place for evangelising tubes amps! This is just confusing and off topic!

And please consider: We contribute here by helping and educating a (seemingly) electronics beginner building his (first) pre-amp project. Thankfully, he's willing to listen and do things right! With this in mind I'd really feel uncomfortable to have him work with high voltages for tubes and all the other hassle tubes provide.

Respecting your tube experience never the less,
Greetings,
Winfried
 
Respecting your tube experience never the less,
Greetings,
Winfried
I came across this topic by chance, and felt the need to share my opinion. Those unnecessarily complicated schematic caught my eye, the minimalistic approach is proven to be better, every additional stage will only worsen the sound quality. Music is a very subjective thing, and so are audio devices. Measurement is a tool, not an end goal.

There is no need to have tubes in the preamp/buffer, of course, although there is nothing so dangerous or too complicated. There is mains voltage in any device, so it is always dangerous.

If tone controls are required, it is sufficient to have an input selector, a potentiometer, an (inverting) buffer and a tone control stage. Without tone control, the buffer is without inverting the signal. And that's it, for digital signal sources and LM3886 no voltage gain is needed, even undesirable.

No coupling capacitors are needed, the DC offset is small. With fewer components, you can go for higher quality parts. Let's say discrete OPAs of the Sparkos, Burson and similar types, voltage regulators better than 78XX, LM317/337 and similar. I had good results with OPA1611/1612, OPA1655/56, OPA1641/42 and Muses 02. I recommend the LT1963 and LT3015 voltage regulators.

As for what will be made, the choice is free.

Cheers to everyone in this thread, I have nothing more to add. :wave2:
 
Last edited: