There is 8mA output per channel if using the ES9018 in stereo mode.. so 0,008x300 = 2.4v, but that is a Vpp value - so ~0.85Vrms. and 680R means 5.44Vpp or 1.923Vrms.
As I already mentioned before, you can use a LU1014 triode cell as in the DAO headphone amp to replace the 2SK369 in the SEN IV.
200mA bias should be sufficient for any DACs that I know of.
And it can also swing a lot of voltage.
Patrick
Another discontinued Jfet
Any hope that those epay ones are the real deal ?
It's paired with a preamp with gain but it used below 0db level so it certainly doesn't feel like 0.85V (very big room to fill with small cone speaker)
I should have used my scope to have a better picture but it was so obvious that I just did it (hard task the way it's build) and happy with the results
Furthermore, according to dac's specs it's 16mA, to be specific 92,4% of AVCC, specified at 3.3V (in my case 3.4V) across the 195R (that's 3.05Vpp or 1Vrms, tried that, just a capacitor at output, not bad but kind of lifeless). That's how it feels like (1.7Vrms) just little bit less than normal.
Anyway, I think Patrick's favorite DAC using Sen doesn't come from ESS. I'm curious about AD1865 and PCM1704.
Anyway, I think Patrick's favorite DAC using Sen doesn't come from ESS. I'm curious about AD1865 and PCM1704.
Last edited:
As i have understand, each phase output half avcc voltage, and effective series/output resistance with four parallelled "channels" is 195R. So if you take 1.7/195 = 8.7mA
Check this picture - http://www.audiodesignguide.com/DAC32/ES9018out1.gif
Check this picture - http://www.audiodesignguide.com/DAC32/ES9018out1.gif
According to this, yep, you are right. Still can't recommended easily to ichiban the 680R resistor for direct output to amp (I surely recommend building it though). Sadly it's so PITA to change them I won't bother anymore.
I'm building another in the future but I think I'll use another DAC for it. For the present, I'll do an A/B test in another system, still I think that Sen will be the most neutral one
savvas: but i forgot that we are talking about a balanced design, so you are right. The output voltages i wrote above is per phase, so total output voltage is twice that. I used 160R at first for my ES9018, and later changed to 240R if i remember correctly, due to needing alot bigger caps otherwise. Also it is built into my pre-amp case, and i do have 6dB gain direct after the SEN.
Since you are writing „most“ of your hum problem, here is the solution that gave me a completely silent SEN with regulated mains supply: The Vref ground reference should not be connected to the AVCC ground at the 9018, but to the power supply ground that feeds the AVCC. This is probably a Placid or something similar upstream. Go as far upstream as possible. This completely eliminated my hum problem.
Thanks for advice kumori, I remember that's the advice also from TP team, for I/V connect the GND from power supply and that mod is more easily implemented.
With an Riv of 150Ω I tried some Civ cap rolling. Most of the ones I tried had varying amounts of noise and/or oscillation. No rhyme or reason that I could detect between value or quality. All were precisely matched.
A copper foil 470pF PPS was the worst. Expensive 6800pF RTX polystyrene a small amount. wima 4700pF FKS bettered the RTX. Amtrans 470pF almost silent. The only one with no issue was a 290pF silver mica, but it made almost no difference really. Without Civ, dead silence. I suspect some interaction with the F5X amp downstream where I had installed 39pF silver micas.
A better result was changing the 470uF Muse ES bipolars with Kansei Bipolars. Yes, at about 10usd each (x8) it will cost you, but its completely worth it.
Sorry, was Kaisei, not kansei
HiFiCollective
Audio Note Kaisei Electrolytic Capacitors | Hifi Collective
Partsconnexion
More Info Page
Guys,
I have an issue that is driving me crazy: I have the ESS9018 SEN DAC. The right channel is fine. Yesterday I noticed that the left had a lower level and I was trying to understand why. What I noticed is that if I connect only one phase for the channel that works, the other phase is flat to zero. On the channel that gives me trouble, if I connect one phase, I observe on the unconnected phase a curve shifted 45 degrees.
Picture attached. Having this strange cross-talk, when it sums up it makes a mess. Where does it pick it up ?
D.
I have an issue that is driving me crazy: I have the ESS9018 SEN DAC. The right channel is fine. Yesterday I noticed that the left had a lower level and I was trying to understand why. What I noticed is that if I connect only one phase for the channel that works, the other phase is flat to zero. On the channel that gives me trouble, if I connect one phase, I observe on the unconnected phase a curve shifted 45 degrees.
Picture attached. Having this strange cross-talk, when it sums up it makes a mess. Where does it pick it up ?
D.
Attachments
Guys,
This was one of the weirdest thing that happened to me in long time:
As I said before, I found that my DAC had one channel at different level than the other. It also sounded strange.
I first thought it was the DAC that died, but I tested it and it was OK. So I checked the IV, compare readings with the working channel and it was OK. What I was observing, as posted before, that there was a strange cross talk between the phases. I had switched from batteries to the lab power supply, to have a control over the currents.
At the end I decided to pull the plug and unsoldered the active devices. I did not want to sacrifice eight 2SK369, so I managed to take them in one piece, as they are glued to the heatsink. I replaced that with other devices, using only one couple per channel to simplify things. No change. What I observed is that with no power, there was no crosstalk. Also I observed that the crosstalk was different with the power supply and with the batteries. With batteries, it was in phase, with the power supply it was 45 degrees shifted. This pointed me to the power source. Analyzing what I had done recently, I remembered that I was experimenting with the chargers for the lithium batteries and that in the process I had killed and replaced one of the relay board that I am using.
End of the story, the new relay board must be defective and there is a cross talk between the channels that prevents the batteries to float independently. I eliminated the shitty board and everything came back to normality.
I spent 15 hours on this issue almost bought a new DAC module, almost scrapped a completed project for something that was really unthinkable.
For this circuit, floating is the key word.
Good Night,
Davide
This was one of the weirdest thing that happened to me in long time:
As I said before, I found that my DAC had one channel at different level than the other. It also sounded strange.
I first thought it was the DAC that died, but I tested it and it was OK. So I checked the IV, compare readings with the working channel and it was OK. What I was observing, as posted before, that there was a strange cross talk between the phases. I had switched from batteries to the lab power supply, to have a control over the currents.
At the end I decided to pull the plug and unsoldered the active devices. I did not want to sacrifice eight 2SK369, so I managed to take them in one piece, as they are glued to the heatsink. I replaced that with other devices, using only one couple per channel to simplify things. No change. What I observed is that with no power, there was no crosstalk. Also I observed that the crosstalk was different with the power supply and with the batteries. With batteries, it was in phase, with the power supply it was 45 degrees shifted. This pointed me to the power source. Analyzing what I had done recently, I remembered that I was experimenting with the chargers for the lithium batteries and that in the process I had killed and replaced one of the relay board that I am using.
End of the story, the new relay board must be defective and there is a cross talk between the channels that prevents the batteries to float independently. I eliminated the shitty board and everything came back to normality.
I spent 15 hours on this issue almost bought a new DAC module, almost scrapped a completed project for something that was really unthinkable.
For this circuit, floating is the key word.
Good Night,
Davide