as do I Eric, more than that, its compulsory!, but it is not a substitute for measurement and should not be framed as trumping measurements already taken in a way that is meaningful to others, with their ears and their gear.eganz1 said:
its insightful into our own response to the current internal emotional/chemical/electrical mind/body state and that is all. thats a big thing, but its not terribly reliable.
I appreciate that Coris is at least trying to make measurements (even though every single one has been different in some way), but i'm afraid anything that shows up in a meaningful way wityh the current setup is more an indication of how bad things are, rather than how good. operating correctly there simply isnt a lot this scope and its input stage can tell us about something more than -110dB down (THD+N) of -135 or lower for other types. so you are effectively measuring the scope noise floor, with nothing meaningful making it through that.
I dont think you want to intentionally hobble the measurement rig, to show 'worst case' whatever that means. when its already struggling.
Last edited:
yes, but the resistor does not convert the current from the DAC in an inverting opamp IV convertor (your schematic example), it presents no load to the dac at all, it looks like a short to a virtual ground. a dac IV resistor is technically a resistor that sets a load point for a current output DAC and the voltage is created across the resistor to ground (or to whatever the load is, basically across the resistor), this loads the dac and drops some voltage (so there are losses)
is this correction constructive enough for you
You have to accept that through that resistor (schematic example) it have to flow a current, to generate a voltage.
From where it come that current if not from the DAC (in this case)? The meaning here is to convert the current which it came out from the DAC pin to a voltage, which is to be amplified further. There is used this schematic to do this job and not something else. Right?
Where is is generated that voltage in this case if not at the ends of that resistor? And the op amp amplify and compensate the loss of energy...
There is that resistor actually, which it set the load (in current) to the DAC output. It have a feedback function strictly for the compensation device (op amp), but is in the same time the I/V resistor.
You forget in your assertion the AC world of this circuit. There that resistor it lead the (AC) current to the very solid ground. There is here first of all about AC current. Isn't it? The DAC output is in AC current which is to be converted in AC voltage... That AC current is limited internally (DAC) by a resistor (195 ohm), or so...
This way it works. Is not the same in your world?
Last edited:
I will try again in HRes mode. But I think the scope can only up to (or down to) -100dB. I will see what it may be done...
I may recognise that is a little bit frustrating the impossibility to let the critics hear/listen what it come out from this my modded device. There is definitely obvious for me how much difference is between a standard and this my modded player. I listened it quite long so as it were, and I know very well how it sounded. And now I can hear it too at another (higher) quality level...
This a real fact! There is a fact too when some say that all is bull ****... But so it is at least, and this way it works the world...
I understand now about the caps... Too much caps in the signal path is not good (for me...). Less is better here. In opposite when about decoupling. It may be those too in the signal path at least, but not in that way the filters one are... This is another controversial theme...
It still quite much work yet. The only mod which is now done is the power system, and some mechanical things. This DAC post processing and the clock system are still yet two big parts of this work, before it will be finish. Then it will come for sure out on the marked the 205 model...
Last edited:
and you have to accept that you dont know how opamps, and or inverting opamp IV convertors work
I question EVERYTHING I do - and I don't add noise and can't understand why anybody would suggest that I do. But can we also have a reality check then, and ditch ALL those low output Moving Coils out there? Both MC and MM have RIAA curves that at least helps towards noise, but MC is potentially 20dB worse off - so maybe I have made my last MC phono stage? Unless you know how to make them very low noise, then what is the problem? As long as noise is satisfactorily low enough.
I don't add noise - and no player (or phono preamp) I have ever done has been described as such.
As for adding noise, while not my thing, natural tape hiss on classic master tapes was a natural from of dither when converting into bits and may well have contributed to why many said back in the 80's that analog recordings converted into digits sounded better than actual new digital recordings. Maybe they had a point - in which case 'accidental' noise improved a situation?
I do wonder if there is such a thing as a simplistic A-B comparison that works reliably - nothing replaces actually spending time getting to know what a piece of equipment does in a familiar surrounding. The only way to come to a valid judgement is to USE it the way you would USE it if you decided to keep it.
Often I have been taken at my word - and chanced it. I rarely if ever disappoint. Just amazed that they put so much trust in me and I would not want to betray that. I believe I am a very moral person.
I have never cared for shoot-outs, often the best equipment becomes the victim. The best is often ejected first - because that dirty word "neutrality" is actually very important long-term and a potential for failure in shoot-outs. An important positive becomes a negative. So subtlety has near no value.
Anyway, those are my tuppence worth.
All the best to you likewise, I am sure you would be welcome if you were to come back to this thread, at least welcome by me.
Cheers, Joe
There's a point: if you say someone is wrong, you should also explain why.
OK.
A current output DAC is designed to be loaded into a (virtual) ground, meaning that its output should be at zero voltage. This is generally done with a circuit as shown by Coris, with an opamp (IC or discrete) with 100% feedback. Due to the feedback action, the DAC output is held at zero volts, and the current flows through the feedback resistor to develop Vout.
There are people who don't like this setup and instead opt for a resistive I/V converter, by loading the DAC output to ground with a resistor. They then amplify the voltage across this resistor with active means or a transformer to develop Vout.
The disadvantage is that this method puts the DAC output not at zero volts but at some (m)volts above it. For instance, with a 100 ohms resistor and 2mA DAC output current, the DAC output rises to 200mV. As the DAC is not designed for this, the distortion is worse than with the output at zero volts.
Therefore the resistor should be kept as low as possible, and I have seen values of just 10 ohms, but this means it only gets very little signal voltage. This puts a very high requirement on the noise performance of the following amplification stage.
Some here have picked up the take-home message that the I/V resistance should be as small as possible, but not understood that it is ONLY an issue with resistive loading. With the opamp I/V conversion circuit as shown by Coris this is NOT an issue because in that circuit the DAC output is nicely held at zero volts.
Therefore, the feedback resistor (that he also calls the I/V resistor) can be as large as possible with not effect on the DAC linearity. By mistakenly using a very low feedback resistor value you paint yourself into a corner because it only develops a very low signal at the opamp output. You then need another gain stage and you end up with a lot of noise and mush that really kills the signal to noise ration and drops the dynamic range by 40dB.
jan
PS Joe I was not accusing you of adding noise btw....
OK.
A current output DAC is designed to be loaded into a (virtual) ground, meaning that its output should be at zero voltage. This is generally done with a circuit as shown by Coris, with an opamp (IC or discrete) with 100% feedback. Due to the feedback action, the DAC output is held at zero volts, and the current flows through the feedback resistor to develop Vout.
There are people who don't like this setup and instead opt for a resistive I/V converter, by loading the DAC output to ground with a resistor. They then amplify the voltage across this resistor with active means or a transformer to develop Vout.
The disadvantage is that this method puts the DAC output not at zero volts but at some (m)volts above it. For instance, with a 100 ohms resistor and 2mA DAC output current, the DAC output rises to 200mV. As the DAC is not designed for this, the distortion is worse than with the output at zero volts.
Therefore the resistor should be kept as low as possible, and I have seen values of just 10 ohms, but this means it only gets very little signal voltage. This puts a very high requirement on the noise performance of the following amplification stage.
Some here have picked up the take-home message that the I/V resistance should be as small as possible, but not understood that it is ONLY an issue with resistive loading. With the opamp I/V conversion circuit as shown by Coris this is NOT an issue because in that circuit the DAC output is nicely held at zero volts.
Therefore, the feedback resistor (that he also calls the I/V resistor) can be as large as possible with not effect on the DAC linearity. By mistakenly using a very low feedback resistor value you paint yourself into a corner because it only develops a very low signal at the opamp output. You then need another gain stage and you end up with a lot of noise and mush that really kills the signal to noise ration and drops the dynamic range by 40dB.
jan
PS Joe I was not accusing you of adding noise btw....
I Did Jan, not as completely in the one post as you, but add my explanation of the resistor not loading the dac in a trans-impedance amp, in comparison to a passive IV (buffered or not), to my comment on the low FB resistance lessening SNR before an additional and significant gain stage. In Coris's case this is then compounded by an analogue attenuator (more noise), before a poweramp which also has a huge amount of gain. he has to attenuate the DAC output at the poweramp input significantly even for high volume (in his words).
i'm not sure how much more detailed i'm expected to get for an expert modder that has been playing with this exact IV stage, with this exact dac chip, for years and I too have been harping on the gain structure elements of that here for years. I have spent quite a lot of time trying to explain this point to him before this started to get heated, because IMO it renders the rest of the mods/changes pretty much meaningless.
I gave numerous links for further reading, spent significant amounts of my own time, but every time its greeted with the sort of crap and backslapping you see here, all sorts of personal jokes and characterization. Sometimes you get a thanks, before the online equivalent of backstabbing occurs. then its ignored, denied and the hand-waving continues. if all this isnt meant to be instructive and nothing is taken on, I have no idea what its all for? with all the pics, explanations, impressions, stories, but no feedback is welcomed, or accepted really.
it tends to put one off putting in too much effort.
combine that with DC coupling, wide bandwidth, no filtering and the thing is an EMC nightmare, you could use a TV in the room as a sort of scope, but those silly engineers ...
i'm not sure how much more detailed i'm expected to get for an expert modder that has been playing with this exact IV stage, with this exact dac chip, for years and I too have been harping on the gain structure elements of that here for years. I have spent quite a lot of time trying to explain this point to him before this started to get heated, because IMO it renders the rest of the mods/changes pretty much meaningless.
I gave numerous links for further reading, spent significant amounts of my own time, but every time its greeted with the sort of crap and backslapping you see here, all sorts of personal jokes and characterization. Sometimes you get a thanks, before the online equivalent of backstabbing occurs. then its ignored, denied and the hand-waving continues. if all this isnt meant to be instructive and nothing is taken on, I have no idea what its all for? with all the pics, explanations, impressions, stories, but no feedback is welcomed, or accepted really.
it tends to put one off putting in too much effort.
combine that with DC coupling, wide bandwidth, no filtering and the thing is an EMC nightmare, you could use a TV in the room as a sort of scope, but those silly engineers ...
Last edited:
If you are on forums like this for more than 10 years like I have, you Get It
I operate from the assumption that there are always lurkers that don't come into the open but do understand and appreciate your efforts at explanation.
The horse and water, remember?
Non carborundum illigitum.
jan
I operate from the assumption that there are always lurkers that don't come into the open but do understand and appreciate your efforts at explanation.
The horse and water, remember?
Non carborundum illigitum.
jan
If you are on forums like this for more than 10 years like I have, you Get It
I operate from the assumption that there are always lurkers that don't come into the open but do understand and appreciate your efforts at explanation.
The horse and water, remember?
Non carborundum illigitum.
jan
I may say that I appreciate too, your attitude, and efforts for coming with explanations.
This is a much more civilised way to be member in a forum, and really contribute with something.
Does doing 'mods' starting with the earliest Philips 14bit x4 players count?
I am near my thirtieth anniversary - no backslapping I fear.
Hey, lighten up. Do you know what the definition of a balanced person is? Somebody with a chip on BOTH shoulders.
Cheers, Joe
Getting late here. So will be as brief as possible.
The truth is that a current output DAC is in its sweet spot when it is shortened. But the output impedance of a device relative to the load it sees is what determines "current mode" - and hence passive I/V at very low impedance still qualifies.
When we speak of "Voltage mode" we do not need zero Ohm source impedance to qualify. Indeed the acceptable ratio is 10:1 as a rule of thumb (actually it is an insertion loss of a maximum of 1dB - which is pretty close to the same thing).
Hence "Current mode" can also have an acceptable ratio - and 10:1 is maybe just OK, but I certainly like to see something rather higher, like 50+:1 or better (note the output of the DAC sees the "1" at the end of the "50+" - it hardly sees it at all).
I use 3R3 on plus and minus phase and get that ratio. The output is low, about 33mV RMS, but still more than ample to deal with. Low noise circuit and extremely wide bandwidth, there is no noise problem whatsoever. If you can design a phono stage, then you have a solution.
Just listen to it... all that is needed - the rest is just words.
Cheers, Joe
PS: For many years have been working of current drive of loudspeakers, where it is not the voltage across the Voice Coil that makes the cone move. It is the current through the Voice Coil that makes the cone move and hence sound. The ideal amplifier is a transconductance amp, where input V is converted to A output. I built such an amp and with 100mV signal input gives 250mA output, 1V signal input gives 2.5A output and etc. How is that achieved? The output impedance of the amp is 270 Ohm. Studying on the bench, you learn a LOT about how "Current mode" behaves - the rules are very different. There is the ideal theory, infinite Z and there are real world practicalities - it still works.
The truth is that a current output DAC is in its sweet spot when it is shortened. But the output impedance of a device relative to the load it sees is what determines "current mode" - and hence passive I/V at very low impedance still qualifies.
When we speak of "Voltage mode" we do not need zero Ohm source impedance to qualify. Indeed the acceptable ratio is 10:1 as a rule of thumb (actually it is an insertion loss of a maximum of 1dB - which is pretty close to the same thing).
Hence "Current mode" can also have an acceptable ratio - and 10:1 is maybe just OK, but I certainly like to see something rather higher, like 50+:1 or better (note the output of the DAC sees the "1" at the end of the "50+" - it hardly sees it at all).
I use 3R3 on plus and minus phase and get that ratio. The output is low, about 33mV RMS, but still more than ample to deal with. Low noise circuit and extremely wide bandwidth, there is no noise problem whatsoever. If you can design a phono stage, then you have a solution.
Just listen to it... all that is needed - the rest is just words.
Cheers, Joe
PS: For many years have been working of current drive of loudspeakers, where it is not the voltage across the Voice Coil that makes the cone move. It is the current through the Voice Coil that makes the cone move and hence sound. The ideal amplifier is a transconductance amp, where input V is converted to A output. I built such an amp and with 100mV signal input gives 250mA output, 1V signal input gives 2.5A output and etc. How is that achieved? The output impedance of the amp is 270 Ohm. Studying on the bench, you learn a LOT about how "Current mode" behaves - the rules are very different. There is the ideal theory, infinite Z and there are real world practicalities - it still works.
Last edited:
important note, that 781Ω number is per dac in 8 channel mode, so for stereo balanced out, ~195Ω and 15.6mA. ive seen that number posted around a few places without mentioning this important detail.
the rest holds, not sure on the in/out Cap ie. whether that is the spec for 8 channel or 2
ahh ... but... FWIR not here with the oppo, because they havent paralleled the dacs for the stereo out as recommended, an odd choice ...probably to avoid relays or some other switching and logic I suppose, but its unfortunate if you only want stereo
the rest holds, not sure on the in/out Cap ie. whether that is the spec for 8 channel or 2
ahh ... but... FWIR not here with the oppo, because they havent paralleled the dacs for the stereo out as recommended, an odd choice ...probably to avoid relays or some other switching and logic I suppose, but its unfortunate if you only want stereo
Last edited:
Actually the ESS DAC is a bit of the odd man out, sort of schitzophrenic, not able to decide whether it wants to be a voltage out DAC (Zout = 0) or a current out DAC (Zout = infinite).
The cleanest option, if you insist on buffering the output, would be the following.
Use an I/V opamp (discrete or chip as you fancy) like Coris showed. Use a feedback resistor that has the same value as the DAC Zout (around 780 ohms). That will give you current loading of the DAC and an output level the same as the open output level of the unloaded DAC. Also, (if you use a good opamp), no noise and distortion penalty!
It may not conform to current fashion, and you may subjectively prefer an other configuration, but for the cleanest sound there's nothing better. The best way to reap the benefits of Joe's clock and low jitter mods!
jan
The cleanest option, if you insist on buffering the output, would be the following.
Use an I/V opamp (discrete or chip as you fancy) like Coris showed. Use a feedback resistor that has the same value as the DAC Zout (around 780 ohms). That will give you current loading of the DAC and an output level the same as the open output level of the unloaded DAC. Also, (if you use a good opamp), no noise and distortion penalty!
It may not conform to current fashion, and you may subjectively prefer an other configuration, but for the cleanest sound there's nothing better. The best way to reap the benefits of Joe's clock and low jitter mods!
jan
Well, it is not so much secret that datasheet now. ESS have revoked the confidentiality of the document for approx two years ago. If you need it, just PM me...