I think probably its fine, you'll want a low resistance pot there - I'd go for 5k (Alps RK27 is available in that value).
Btw the pot's still in the 'signal path' even if its only the shunt element in an attenuator. 😵
Btw the pot's still in the 'signal path' even if its only the shunt element in an attenuator. 😵
Ok.
I still don't see the point to add another active stage in the signal path:
Instead of your 1k/100k resistor on buffer input ( first opamp) implement your attenuator there ( both opamps ( input buffer and the one added after your filter cell) are non inverting buffer so they are exchangeable).
And i would not implement a pot as you'll never be able to dial up same level between both chanel ( in other words your intended xover could only be expected not achieved in practice). You would face an ergonomy issue too: how are you going to lower volume for your whole loudspeaker?
But please let me go back to your root problem: you talk about noise issue. Ok from dsp? If it is an issue with your dac nothing upstream in the signal path is going to solve it.
If it is related to your gain staging then you need a fixed attenuation not something variable.
Could you make a global schemo of your system (Including source) it'll be easier to spot issue and how to solve them.
I still don't see the point to add another active stage in the signal path:
Instead of your 1k/100k resistor on buffer input ( first opamp) implement your attenuator there ( both opamps ( input buffer and the one added after your filter cell) are non inverting buffer so they are exchangeable).
And i would not implement a pot as you'll never be able to dial up same level between both chanel ( in other words your intended xover could only be expected not achieved in practice). You would face an ergonomy issue too: how are you going to lower volume for your whole loudspeaker?
But please let me go back to your root problem: you talk about noise issue. Ok from dsp? If it is an issue with your dac nothing upstream in the signal path is going to solve it.
If it is related to your gain staging then you need a fixed attenuation not something variable.
Could you make a global schemo of your system (Including source) it'll be easier to spot issue and how to solve them.
The miniDSP produces a fair amount of background noise/hiss when using the analog input, which i have to. I have gain to spare, so the solution is to attenuate after the miniDSP. I need to be able to fine-tune the level of the low section relative to the mid, this has to happen after the input stage, as they are sharing this stage. This fine-tuning will be a set and forget thing. I just want to choose the best way to do this, last time i got it wrong.
My drawing is pretty much complete. The voltage source is the minidsp, and the amplifier comes right after the output stage.
My drawing is pretty much complete. The voltage source is the minidsp, and the amplifier comes right after the output stage.
Because i think the issue is not dac related but ADC related. This is a gain staging issue to me. Iow you got it wrong because you attack the wrong place.
https://www.diyaudio.com/community/threads/what-is-gain-structure.186018/
I know it'll ask a bit of time but please describe your sytem within a schemo.
https://www.diyaudio.com/community/threads/what-is-gain-structure.186018/
I know it'll ask a bit of time but please describe your sytem within a schemo.
Denon 3700x -> analog input miniDSP 2x4 HD ->custom circuit as shown -> Icepower 200AS2
The denon is controlling volume. I agree its a gain staging issue, thats what im trying to fix.
The denon is controlling volume. I agree its a gain staging issue, thats what im trying to fix.
Bingo. Read Pano's article and i'll back latter ( have the kids to go back from school).
Hint: your ADC is under modulated and you experience quantification noise. Or the other way around... 😉
We are getting off track here. As i said, i have gain to spare, so i can afford to attenuate right before the power amp sufficiently to remove any unwanted noise and still achieve the max spl i need.
All i need is a good solution to attenuate while introducing the least amount of noise and distortion. I can't see any way around using a dedicated opamp for this purpose. But i don't have confidence in my current suggestion on how to arrange resistors and pots in this arrangement.
All i need is a good solution to attenuate while introducing the least amount of noise and distortion. I can't see any way around using a dedicated opamp for this purpose. But i don't have confidence in my current suggestion on how to arrange resistors and pots in this arrangement.
Sorry but i disagree, if the noise are introduced at ADC stage ( because your out level of Denon is too low) you could do whatever you want upstream, once noise are introduced you can't make it disapear.
You have spare gain at will it is a symptom of under modulated adc. Digital system differs in the way the gain staging is implemented to analog ones.
You have spare gain at will it is a symptom of under modulated adc. Digital system differs in the way the gain staging is implemented to analog ones.
Without any attenuation upstream i could theoretically reach 20db higher volume then i would ever need. I also have about 10db too high background noise. WITH 10db attenuation upstream, i can still reach higher volumes than i actually need, AND background noise is gone. I don't understand what you think is the problem.
Converters need to be 'attacked' with the highest level possible. If you attenuate before the ADC you'll induce quantification noise (the hiss and noise you talk about).
Once this is converted into digital stream here they are.
Iow you can't control your overall level before the adc. This have to be done analog post dac or into digital domain post adc.
I repeat i am almost sure your issue come from there. Try to drive your minidsp by denon with volume fully open and listen if there is still noise and hiss.
Of course not your amp, if you have a mixing desk ( or anything which can accept the minidsp out max voltage) listen through headphone.
Once this is converted into digital stream here they are.
Iow you can't control your overall level before the adc. This have to be done analog post dac or into digital domain post adc.
I repeat i am almost sure your issue come from there. Try to drive your minidsp by denon with volume fully open and listen if there is still noise and hiss.
Of course not your amp, if you have a mixing desk ( or anything which can accept the minidsp out max voltage) listen through headphone.
Yes there are reasons, don't put it there, its detuning the Sallen Key filter (which expects zero source impedance to
achieve its designed response curve).
Basically add another opamp stage specifically to attenuate, and use lowish value resistors, like 1k:1k to keep the noise
down (most opamps can happily drive 2k loads) - there is absolutely no benefit to scrimping on opamp stages, adding
them usually increases performance as here as you need each part of the circuit to do one thing well, not two things
badly. Use as many opamps as it takes - no more, no fewer. An illogical desire to reduce the number of opamps is
going to cause problems, not solve them.
Here you want an input stage with suitable input impedance, you want some attenuation, and you want filtering - that's 3 stages.
(However there might be a way to combine filtering and attenuation, but not with standard Sallen-Key design - you'd be
starting from scratch).
I note the input stage lacks RF filtering - I'd suggest fixing that if the input is via a cable from another box - 330pF to ground on
the input connector at least.
