Is it possible to get lower volages than 2 VRMS/lower powers from DACs? ... And use external standalone preamps or preamps from integrated amps?
AFAIK some old vintage preamps have attenuators and others don't. Some people even modify their pres to take up to 7 VRMS (for ex.: Sansui's AU-999).
How do you handle high outputs from DAC/CD players when connecting them to vintage preamps/power amps?
AFAIK some old vintage preamps have attenuators and others don't. Some people even modify their pres to take up to 7 VRMS (for ex.: Sansui's AU-999).
How do you handle high outputs from DAC/CD players when connecting them to vintage preamps/power amps?
Many older preamps have the inputs go directly to a volume control, so there's no overload problem.
And many haven't. For ex.: Sansuis. Some models have built-in attenuators. And you can always add attenuators to the line (I saw them small as RCA connectors).
DACs have opamps pre preamps. I'm new into HiFi/Audio, so correct me if I'm wrong. Is it possible to use external preamp instead and get the signal at it's source (before opamp in line)?
DACs have opamps pre preamps. I'm new into HiFi/Audio, so correct me if I'm wrong. Is it possible to use external preamp instead and get the signal at it's source (before opamp in line)?
Depends on the DAC design - many DAC chips have internal opamps and its impossible to get the signal prior to these. Most designs which aren't using chips with on-board opamps use them externally. A minority use a current out DAC chip followed by passive I/V and non opamp-based amplification.
abraxalito, thank you!
So in designs with by the book IC implementations, precise clocks and good PSUs the difference should be minimal or virtually none. Is it correct?
What is the difference between IC opamping and use of "external" (onboard) opamp? In dynamics? Less distortion?
As far as I Googled both popular ICs (ES9018 and PCM1794) permits passive I/V implementations. Examples are audiodesign's and DD's.
So in designs with by the book IC implementations, precise clocks and good PSUs the difference should be minimal or virtually none. Is it correct?
What is the difference between IC opamping and use of "external" (onboard) opamp? In dynamics? Less distortion?
As far as I Googled both popular ICs (ES9018 and PCM1794) permits passive I/V implementations. Examples are audiodesign's and DD's.
In my experience the problems with opamp sound quality are down to PSU quality. That is - opamps are much more demanding of power supplies than most designers realize. So you can undo a lot of the damage to SQ which is laid at the feet of opamps by improving their supplies.
There is an exception to this principle though, that's when an opamp is being fed with fast edges direct from a DAC chip - I don't believe the SQ problems introduced by such an opamp can be fixed up with better supplies. The opamp shouldn't be fed such out of band (OOB) signals in the first place, so a passive filter is in order between the DAC and the opamp.
IC opamps tend to be built on CMOS processes because the digital parts of those chips need CMOS. CMOS is my last choice of process technology when looking at an opamp for audio use. Using external opamps allows them to be BiFET or bipolar which are more appropriate where SQ is the main issue.
Yes dynamics is the issue - a poorly decoupled opamp squashes the perceived dynamics.
There is an exception to this principle though, that's when an opamp is being fed with fast edges direct from a DAC chip - I don't believe the SQ problems introduced by such an opamp can be fixed up with better supplies. The opamp shouldn't be fed such out of band (OOB) signals in the first place, so a passive filter is in order between the DAC and the opamp.
IC opamps tend to be built on CMOS processes because the digital parts of those chips need CMOS. CMOS is my last choice of process technology when looking at an opamp for audio use. Using external opamps allows them to be BiFET or bipolar which are more appropriate where SQ is the main issue.
Yes dynamics is the issue - a poorly decoupled opamp squashes the perceived dynamics.
How much power would you recommand for ESS9023 based board (with IC's opamp)? 5, 10 W? (1, 2 A @ 5 V)?
Agree with all the above posts from abraxalito.
It is not about quantity (a supply of a few tens of mA are enough for your average opamp) but about quality.
Since opamps have no HF PSRR, the best way to screw up the dynamic performance of a good opamp is to apply resonant decoupling. That is very easy to do : just add a 100nF ceramic in parallel with any low-ESR capacitor, ceramic, film, oscon, whatever. Then add one ceramic per opamp and connect it all together with highly inductive traces. Then connect to DAC. This pretty much guarantees all the kinds of fail than can happen, will. For extra fail, use a regulator that is unstable with the chosen capacitors.
Bulletproof recipe is 1uF or larger X7R SMD, plus high-ESR cap around 50-100 mOhm, aluminium tantalum whatever, with reasonably low inductance, ie not more than 3.5mm pin spacing. I like Panasonic FR/FM electrolytics, cheap, very high quality, no microphony, very reliable. Add suitable regulator and you're good to go.
It is not about quantity (a supply of a few tens of mA are enough for your average opamp) but about quality.
Since opamps have no HF PSRR, the best way to screw up the dynamic performance of a good opamp is to apply resonant decoupling. That is very easy to do : just add a 100nF ceramic in parallel with any low-ESR capacitor, ceramic, film, oscon, whatever. Then add one ceramic per opamp and connect it all together with highly inductive traces. Then connect to DAC. This pretty much guarantees all the kinds of fail than can happen, will. For extra fail, use a regulator that is unstable with the chosen capacitors.
Bulletproof recipe is 1uF or larger X7R SMD, plus high-ESR cap around 50-100 mOhm, aluminium tantalum whatever, with reasonably low inductance, ie not more than 3.5mm pin spacing. I like Panasonic FR/FM electrolytics, cheap, very high quality, no microphony, very reliable. Add suitable regulator and you're good to go.
The most simple answer to the original question would be to build a simple attenuator into an RCA cable and be able to use any modern DAC...
Btw, while I agree with everything peufeu said, I can't make sense of abraxalito's comments, unless we're told what he means in technical terms by "dynamics", how they are "squashed" and how to measure these things in practice.
I have low-pass LC filter on infput from PSU and ULN regulator, peufe. I'm asking for predesigned DIY kit. Maybe I can bypass or add this circuit.
00940, attenuators are an option. Some preamps have pots first on signal path to regulate signal power, others have built in switch attenuators. Some people even mod their pres by simply adding attenuators on board/signal path (what I consider a non-sense as you modify an vintage piece of electronic when you can simply add external attenuator on RCA - as you proposed). But attenuators have an impact on sound, you lose some dynamics and sound is a little bit wormer (as I noticed).
Well, for attenuators, there isn't a value that will work in every case. You'll need to adjust it at least for the DAC output impedance and the preamp input impedance. It can't be too low in value (or you ask too much of the DAC) nor too high (or you risk to pick up noise).
Anyway, decreasing the DAC output, either by lowering the gain of the I/V stage or by attenuating the output, is technically a bad solution wrt gain structure. If you're able to build or modify a DAC, you should be able to reduce the gain of the preamp.
Anyway, decreasing the DAC output, either by lowering the gain of the I/V stage or by attenuating the output, is technically a bad solution wrt gain structure. If you're able to build or modify a DAC, you should be able to reduce the gain of the preamp.
Let's say that I have a pair of poweramp with 1 V sensitiviti input and preamp with 150 mV. What would you do? Which DAC will you chose? What kind of modification/design regarding analog signal output would you go for?
Depends...
In theory: I wouldn't want to be limited by one DAC. Changing the power amps gain might be difficult since their stability will be affected. So this leaves the preamp to be modified.
So I'd keep the box of the preamp, the input switching and the volume pot. The electronics would be modified to have one (optional) unity gain buffer before the pot, another after the pot. The actual required mods would depend on the preamp's internals.
In theory: I wouldn't want to be limited by one DAC. Changing the power amps gain might be difficult since their stability will be affected. So this leaves the preamp to be modified.
So I'd keep the box of the preamp, the input switching and the volume pot. The electronics would be modified to have one (optional) unity gain buffer before the pot, another after the pot. The actual required mods would depend on the preamp's internals.
If you use a current output DAC with passive I/V, you can (within reason) lower the I/V resistor value to reduce the voltage out. I'm going to do that on my new design to get exactly the voltage I need for my specific headphones or power amp.
I don't have any technical measurements for dynamics, that's entirely a subjective evaluation. But I am sure that eventually measurements will be developed for this aspect of perceived sound quality - probably they'll involve measurement of noise modulation. A 'dynamic' sound is one with lower noise modulation than a 'squashed' one.
Are you talking about what size of wallwart? If so, then the absolute minimum required to do the job. As peufeu says, opamps take a few mA each.
Why I say the minimum size is because transformers all couple noise from the mains supply. The larger the physical dimensions of the trafo, the more capacitance between your DAC and the mains. We want this capacitance to be as small as possible, hence choosing the smallest trafo is what I recommend. Split bobbin trafos are best, if you can find one.
On my blog I have a ferrite cored trafo which has almost immeasurably low capacitance between windings, but its very inefficient due to needing a lot of copper. But it'll most likely power my DAC seeing as I only need 1W max.
Another important preference is that I don't want to modify my preamp (or at least not in irreversible way). As I have 2 AUX inputs I could modify one of them in that way.Are we discussing about modifications that are basically on board attenuators?
Do I compromise anything lowering the voltage?
BTW what is best for dynamics, playing with software volume high, pre/opamp high or both high?
- Status
- Not open for further replies.