I'm not sure the capacitor would isolate anything, as the noise currents (AC) could still flow through the capacitor. But here's another line of thought:
Use an air-core current transformer (CT) with a suitable turns ratio after a current output DAC. The D/A gives a low current (mA) but the turns ratio could compensate for that. It is also possible to print this transformer onto a PCB or integrate it into a chip (like the Icoupler from ADI).
There're some challenges regarding the above but it's a benign isolator, as there is no iron / ferrite core or a non-linear BH curve to cause distortion.
A hall effect current sensor could be used instead (with multiple turns as below) but it has noise and distortion, much like an opto-coupler.
Use an air-core current transformer (CT) with a suitable turns ratio after a current output DAC. The D/A gives a low current (mA) but the turns ratio could compensate for that. It is also possible to print this transformer onto a PCB or integrate it into a chip (like the Icoupler from ADI).
There're some challenges regarding the above but it's a benign isolator, as there is no iron / ferrite core or a non-linear BH curve to cause distortion.
A hall effect current sensor could be used instead (with multiple turns as below) but it has noise and distortion, much like an opto-coupler.
Last edited:
My take is that it is DC like fluctuations that is the nasty - that would be removed by a capacitor I would think. Higher frequency noice traveling backward towards a DAC I suppose is hard to stop, if not impossible, in general if it is present in the audio passband.
But yes, quite a few attest to a DAC with a transformer coupled output. Maybe it is this aspect that makes them a the real benefit. This is system thinking... how does it all interact to create the best net result....
//
But yes, quite a few attest to a DAC with a transformer coupled output. Maybe it is this aspect that makes them a the real benefit. This is system thinking... how does it all interact to create the best net result....
//
Having output transformers can make a difference mitigating leakage current because of power supplies (either linear or SMPS) I found in my DACs. Besides that they can sound great. But they have one very bad parameter.
Last edited:
Just joking. Try them out one time, you will like them. I used to sell them to fellow hobbyists and there is a trade off between quality and size. The Haufe studio equipment ones suitable for DAC output voltages are relatively small and have pretty good specs. Very hard to find though.
Neutrik NTL-1 is OK but design a PCB and only solder them once.
Neutrik NTL-1 is OK but design a PCB and only solder them once.
Last edited:
I did and used Lundahl LL1684 - I think I mentioned in the thread. I couldn't make it silent if I didn't connect the grounds of each side. These where Nc400 <-> Soekris DAC in balanced mode (one board / channel) - if left floating i.e. only hot and cold connected, there was slight hum - and they where close so no long cabling.
Still own them.
So with these the DAC is isolated... but perhaps there is still a little internal problem inside a class-d amp... 🙂
//
Still own them.
So with these the DAC is isolated... but perhaps there is still a little internal problem inside a class-d amp... 🙂
//
Another issue maybe and for sure not all transformers perform the same. They should be in a shielded metal box with correct grounding scheme definitely. I don't know those exact Lundahls but the ones I tried were expensive, ugly (not entirely unimportant) and vulnerable with pressed paper PCB material. As said transformers can improve some issues that may or may not be of your likings. Some sound OK, some are not an improvement etc. They seem simple devices but are not so simple in reality.
You could try another amplifier with normal output power and have volume control at its inputs instead of connecting DACs directly (a system choice). Open input amplifiers just give too many unnecessary side effects. It probably is a soup with a few wrong ingredients and adding ingredients won't make the soup any better.
You could try another amplifier with normal output power and have volume control at its inputs instead of connecting DACs directly (a system choice). Open input amplifiers just give too many unnecessary side effects. It probably is a soup with a few wrong ingredients and adding ingredients won't make the soup any better.
Last edited:
The units mentioned is non of the ones I use today. Current DACs are AKM based and galvanically connected to ordinary class-d amps. This sound great - but perhaps there is untapped potential....
//
//
I can't see any advantage with that - and it would not be a desired solution. I'll stick with the 64 bit volume control in CDSP thank you.
if its just a matter of gain stage I might do a passive attenuator as this could increase SNR.
//
if its just a matter of gain stage I might do a passive attenuator as this could increase SNR.
//
Not checking = not knowing. No need to thank me, it could be in your benefit. Passive attenuator is the same of course. Maximum gain available at all times amplifies all what happens with maximum gain (at all times) and maybe you do not want to amplify all what happens with maximum gain (at all times). Combine this static high gain with high possible output power of the amplifier and see this is a far from ideal way of doing things. Maybe it looks good on theoretical paper but analog reality speaks some words too.
The advantage would be that noise, hum, results from not clean ground and what not that appear full force at the amplifiers inputs/GND are not amplified with maximum gain (besides the 64 bit volume controlled signals that should be amplified). Possibly a solution the simplest possible way. Well in fact it is, maybe undesired for whatever reasons but for sure simple & effective.
The advantage would be that noise, hum, results from not clean ground and what not that appear full force at the amplifiers inputs/GND are not amplified with maximum gain (besides the 64 bit volume controlled signals that should be amplified). Possibly a solution the simplest possible way. Well in fact it is, maybe undesired for whatever reasons but for sure simple & effective.
Last edited:
That's more a figure of speech... 😎
Clearly a well balanced gain structure is desired - no doubt.
//
Digital systems are sensitive to timing errors, analog to amplitude, both to noise. For a digital source there have to be a point in the system where it is converted to analog as speakers are analog. So where is the optimal point to have a gain control instead of a volume control? Or both? Just wondering....
Yes. Volume control does not change gain. It reduces input signal level.
So it depends on where the "volume control" is situated if its a gain control or a volume control. Because you could have volume control in the feedback loop....
Potato / tomato... 🙂
//
Potato / tomato... 🙂
//
Don't know what you are trying to explain but this is basic stuff. With added volume control (like fully normal in analog days) all signals will be reduced in level but gain is constant. What you do with DACs connected directly is amplify all that the amplifier sees with maximum gain. Exactly what should not be done.
One of the best examples what can happen is to power the DAC off and on again after a few seconds. With volume control this would also be way less. Again it is a system choice with too many drawbacks to ignore.
One of the best examples what can happen is to power the DAC off and on again after a few seconds. With volume control this would also be way less. Again it is a system choice with too many drawbacks to ignore.
Last edited: