We should accept there are different "schools" of designing audio circuits and both have right to existence. One is called KISS and appreciates simplicity as guidance to quality. The other school doesn't hesitate to add complexity to address even minor or just theoretical issues: "better one ton too much than one gram too little".
It's up to @Palmtrees to decide which style suits him the best for his project. It's always possible to change approach in each consecutive project.
It's up to @Palmtrees to decide which style suits him the best for his project. It's always possible to change approach in each consecutive project.
I'm sure I can design a preamp with 100 op amps in the signal path and it will work perfectly. Likewise, I am sure that in a blind test a preamplifier with only 2-3 operational that perform exactly the same functions will be preferred.
Ok that's definitely something I would like to avoid, as it would be really nice to switch silently between preamp vs preamp bypass. Would similar caps/resistors also be needed before the tone control bypass switch (S5) to prevent transients when switching on/off the tone bypass?
For the output C/R, do the resistors go after the caps like this:
This will create a high pass filter, which means I should consider the input impedances of the power amp and active sub to make sure the Fc is not too high right? I know that my power amp has an input impedance of ~21K ohms, but I'm not sure about the sub. I am planning to get a 12" Klipsch Reference Series active sub, but I can't find any information about the input impedance online. What would be good values for these?
Could you explain more about what this means? Does this effect sound quality, or just the characteristics of the control law?
My main sources will be an iPhone, TV, and CD player. These are all line level out, so the source impedance should be pretty low right?
Exactly. The R33 & R34 bleeders should be something like 100kOhm (that is, considerably larger then the highest input impedance). Caps, I'd settle to one type of larger electrolytic for most of the coupling caps. Like 220uF/25V Nichicon Muse (UES series).
BTW, you have the same issue with the input cap for the balance control, again, you'd need two, one for each path. And you should keep the Baxandall tone control connected, otherwise its input coupling cap has no way to charge when tone control is bypassed.
Yep, those will be all very low output impedance, about 100Ohms.
I wouldn't worry. A volume pot circuit always goes from -infinity (gain=0) to some fixed gain value (gain=x), doesn't matter what that value is, the ratio is always infinity ;-)
The Baxandall is quite good resolution in the main region of operation of the pot and that's where it counts. Near the end stops action accelerates. For the max end stop, the circuit gain determines how much. 12dB, a common value, is completely fine. 30dB would not be.
Do we have any measurements on DIYaudio of where the source signal level causes the input distortion to exceed the output’s distortion?
Often, the distortion graphs within data sheet are showing the op amp with some gain and the output distorting only. If both channels of an LM4562 were arranged for analysis in unity gain with one CH inverted and the other CH non-inverted, it might be interesting to see how the distortion-vs-input level results compare to the more common tests.
I'm not sure what you mean by this, do you mean that I should keep one side of the Baxandall tone control connected at all times, i.e by removing one of the tone control bypass switches (S1 or S6)?
Or should I add another 470 uF cap the tone control bypass circuit before R30?
Like this, with C26 removed and C36 and C37 added?
Are the 100K bleeder resistors needed here too?
Ok great, will add bleeder resistors there in the next version.
This has got me thinking about the input select switch… This will be a dual gang rotary switch. If I want to avoid pops when switching between inputs, do I also need caps and bleeder resistors for each input line into the select switch?
This has got me thinking about the input select switch… This will be a dual gang rotary switch. If I want to avoid pops when switching between inputs, do I also need caps and bleeder resistors for each input line into the select switch?
Ok next version... I'll keep going as long as there are still improvements to be made! 🤓 😈 👽
I added 470 uF capacitors (C36, C37, C34, and C35) and 100K bleeder resistors (R35, R36, R33, and R34) to the preamp bypass and tone control bypass circuits to avoid pops when switching the bypasses on and off as suggested by @KSTR.
I also started wondering if I have too many electrolytic capacitors in the signal path. So I decided to change the value of the capacitors (C1, C38, and C31) forming high pass filters at the inputs of (U1, U16, and U10) from 47 uF to 4.7 uF, which will allow me to use high quality metal film capacitors there (i.e. Mundorf aluminum foil in oil or similar), especially at the input with C1. I calculated that the Fc would be 0.34 Hz, which is more than acceptable right? Are there any other negative repercussions to doing this?
I added 470 uF capacitors (C36, C37, C34, and C35) and 100K bleeder resistors (R35, R36, R33, and R34) to the preamp bypass and tone control bypass circuits to avoid pops when switching the bypasses on and off as suggested by @KSTR.
I also started wondering if I have too many electrolytic capacitors in the signal path. So I decided to change the value of the capacitors (C1, C38, and C31) forming high pass filters at the inputs of (U1, U16, and U10) from 47 uF to 4.7 uF, which will allow me to use high quality metal film capacitors there (i.e. Mundorf aluminum foil in oil or similar), especially at the input with C1. I calculated that the Fc would be 0.34 Hz, which is more than acceptable right? Are there any other negative repercussions to doing this?
No, everything is fine. The more things you put in the signal path the better. 🤣
Yes, as it's the same scenario. Ideally, you have all the inputs "preconditioned" before the source selector and that includes removal of any DC offset.
But of course one might have to ask if all this is really required, especially if the use case is very clear and well-defined.
As mentioned in a previous post, it'll be good to have enough circuit option in the layout so you can freely experiment without butchering up the PCB to much.
That is correct. The distance input to opamp should be as small as possible.
I’m planning to do a breadboard prototype before designing the PCB so that will give me a chance to try different configurations and components
Preamp looks nice .
Given the amount of components in your design, it would make more sense IMHO to fab a PCB. For 3-10 run is very affordable (10-20 usd/euros+ shipping). I have found that the time to route the PCB, specially with thru holes, is less that the time it takes to to a bread board as once the bom increases it can get very hard to have it stable with all the myriad problems of bad and loose connection and changes on capacitance and inductace. Also, you can test multichannel by selectively populatiimg the boards and daisy chaining them.
It can be a basic route with test point and using the auto route. A prototype to test ideas and component placement.As I said, at least for me, it is a save time and sanity.
Given the amount of components in your design, it would make more sense IMHO to fab a PCB. For 3-10 run is very affordable (10-20 usd/euros+ shipping). I have found that the time to route the PCB, specially with thru holes, is less that the time it takes to to a bread board as once the bom increases it can get very hard to have it stable with all the myriad problems of bad and loose connection and changes on capacitance and inductace. Also, you can test multichannel by selectively populatiimg the boards and daisy chaining them.
It can be a basic route with test point and using the auto route. A prototype to test ideas and component placement.As I said, at least for me, it is a save time and sanity.