It's well matched to my needs, Henry and I have been friends for a long time. I've been following his project and interpretation of your design since he first started. Henry is very particular, methodical, and scientific in his approach, the fact that the DCG3 was what he chose for his project speaks volumes I think about its design quality. And I heard that immediately when I hooked it up.
It is the only hifi amplifier in the house that I did not design. (and that includes some commercial kit lurking around the place)
I'm very pleased with it, and Henry did a great job implementing your design.
It is the only hifi amplifier in the house that I did not design. (and that includes some commercial kit lurking around the place)
I'm very pleased with it, and Henry did a great job implementing your design.
Didn't know you are friends, cool. Thanks for the good words. You are a professional engineer and I am an audio amateur so I double appreciate.
You are as good an audio design engineer as anyone I know, and your designs are well appreciated anywhere I hang out. 😀
Well, thanks for the kind words, Kevin. A good scientific theory not only answers a question but prompts further questions and learning. I hope Salas will take pride that his circuit inspires others to use it as a basis for their own exploration.
That said, my build has turned up some unexpected sensitivities. My advice to others is to stick with the official build, which has been thoroughly wrung-out. If you decide to go your own way and run into troubles, you're on your own to sort them out.
At this point, I don't know for sure why I've had oscillations with my boards. They are very close to the official layout. The substitution of discrete 2SK170s for LSK389s doesn't seem off the bat to be a problem. Possibly there is an interaction with the Super Regulator, though I didn't see any RF on the supply rails when I was debugging the prototype.
I've added dominant-pole compensation to the unit in Kevin's hands. My other unit doesn't require it. SPICE simulation probably isn't accurate enough to explain the behavior. The mystery will remain unsolved for now because I don't have the instrumentation to do open-loop gain/phase measurements, and I don't feel like tearing into my finished unit which looks great and is working very well.
-Henry
Please what gain for my headphones of 80 dB/V, planar magnetic transducer and 34ohm.
Now I must turn volume dial up to 9 out of 10 on headphone. Line output volume is much better at 5 out of 10 (on a 5x gain power amp)
Current settings on DCG are 3x gain and 150mA bias.
Now I must turn volume dial up to 9 out of 10 on headphone. Line output volume is much better at 5 out of 10 (on a 5x gain power amp)
Current settings on DCG are 3x gain and 150mA bias.
Last edited:
80dBSPL @ 1V sensitivity headphones need a 30W at 34 Ohm amplifier with +/- 45V rails and 24dB gain for 110dBSPL
Are you sure they are such a heavy a load?
Are you sure they are such a heavy a load?
If on the other hand they are 80dB/mW then 1W is enough for 110dB and that's possible if driven balanced from double DCG3 with 1.2W capability. Gain setting 6X should be compatible.
morning Salas
from specifications: 34ohm / sens 93dB / mW
I filled as in attach below.
Just wanted to balance amplifier and headphone loads more even in the middle of dial.
Anyway DCG3 can drive both loads, despite turning up and down the knob each time, no complains, sound is amazing no hum at all, even at 10 to 10 pot dial, .
from specifications: 34ohm / sens 93dB / mW
I filled as in attach below.
Just wanted to balance amplifier and headphone loads more even in the middle of dial.
Anyway DCG3 can drive both loads, despite turning up and down the knob each time, no complains, sound is amazing no hum at all, even at 10 to 10 pot dial, .
Attachments
In audio dBm is ref to 600 Ohm. When finally 93dB/mW sens (much more normal no need for double dcg3 bridged/balanced) try 5X gain.
headphones are intended for chinesse market, brand is somewhat translated as "old week 2017" and model "SK2000".
老周2017旗艦，王者系列SK2000 平板耳機關聯LCD3 HE1000 HD800
And amplifier is a heavily modified ACA without feedback and simplified input circuitery that turned to an gain increase from 3x as stock, to an 5x.
老周2017旗艦，王者系列SK2000 平板耳機關聯LCD3 HE1000 HD800
And amplifier is a heavily modified ACA without feedback and simplified input circuitery that turned to an gain increase from 3x as stock, to an 5x.
Forget to mention also added to the NPass desing an input to ground 47k resistor for better impedance matching.
Salas please, it will degrade the sound on DCG3 adding an switch on the front panel to change gain from 3x -line to 5x -headphone?
Salas please, it will degrade the sound on DCG3 adding an switch on the front panel to change gain from 3x -line to 5x -headphone?
If you can do it with front panel switch control for mini relay located very near to the dcg3 board's gain resistors it could possibly work without quality degradation but beware of loud pops if switching them live.
Pops are less likely when having one of two gain resistors permanently in circuit and applying another in parallel to reach desired value.
Pops are less likely when having one of two gain resistors permanently in circuit and applying another in parallel to reach desired value.
Hello, I received the ordered parts. Tomorrow I will test the necessary parameters. BC327 transistors from DIOTEC 50 pieces. selectively measured 10 pieces of hFE 450-470. Close enough you can choose. Who has such transistors, you can check whether the transistor pins are magnetic with an ordinary magnet ( I read about this method of checking for originality) My BC327S are magnetic. I ordered earlier MPSA42/92 also magnetic, there were no complaints about the work. Transistors BF256 and PF5102 are not magnetic. I'm making my UBiB1.3 board to power the preamp. there is a great desire to place the radiators on the board, but there were doubts about the size. Tell me the acceptable dimensions for radiators with a bias current of 100-110 mA. Thanks.
Attachments
Such sinks will go very hot in shunt regs. This reg type works like a heavily biased Class A amp. Consuming as much as it gives out or more. If you can accept high sinks temperature it also takes very good ventilation not to impact a box's internal ambient temperature too much. And to use 105C or 125C capacitors in the vicinity of the sinks of course. M1 & M2 themselves can work at near boiling temperature but with a shorter life expectancy. Use three to four times better thermal capacity sinks.
