OK, I hope you do one day.
Simple as that.
Except I know very litle about Circlotrons, so that will now be another Saturday afternoon in front of this PC !!!
🙂
1Audio, my headamp differs a bit from the National application. I use Shunt Feedback for less distortion and i force the Opamp into Class A.
> Except I know very litle about Circlotrons, so that will now be another Saturday afternoon in front of this PC !!!
I started here :
The Tube CAD Journal, Circlotron Relativity
Tubecad has published many versions of circlotron, including ones with MOSFETs.
Another info source is the Pass forum, where I have also published a few simplified circuits.
And many others also at the same forum from Circlomanen.
Happy reading,
Patrick
PS the reason why I use circlotron is because I only need N devices and they are by design balanced. You could also do the same (unity gain output stage) using the DAO with Taylor Current Source. Only thing you need to do is to run the LU1014s as in Zen V9 with higher bias and lower source degenration, and you need to change a few things around the Taylor current source to balance current modulation. See also Tubecad for the Taylor current source.
I started here :
The Tube CAD Journal, Circlotron Relativity
Tubecad has published many versions of circlotron, including ones with MOSFETs.
Another info source is the Pass forum, where I have also published a few simplified circuits.
And many others also at the same forum from Circlomanen.
Happy reading,
Patrick
PS the reason why I use circlotron is because I only need N devices and they are by design balanced. You could also do the same (unity gain output stage) using the DAO with Taylor Current Source. Only thing you need to do is to run the LU1014s as in Zen V9 with higher bias and lower source degenration, and you need to change a few things around the Taylor current source to balance current modulation. See also Tubecad for the Taylor current source.
EUVL, you have the right spirit for the Blowtorch thread. You have made an example of a most elegant approach to making a headphone amp. Simple, yet not too simple, sophisticated where it gives improvement, and your own take on the optimum output stages. Well done!
I'm not sure how the output buffer could add offset that the input couldn't correct. Is it possible it was oscillating? With a 100 MHz power bandwidth it could oscillate at a very high frequency.
In many cases discrete circuits offer the ability to use currents not possible in an IC. Something like these buffers seem to be special cases where the internal currents are high. The biasing and matching are probably very good. That was why I was asking about the line between discrete and integrated design.
I have LME49600 working in my DA converter, i use a servo there. No problem with oscillation there.
Sofar i had no success using the LME49600 without servo but that does not mean it can not
be made to work without. Maybe it needs a stopper resistor, maybe something else.
John mentioned in another thread that a preamp should never leave class a.
Look at my headphone amp. It has an bias current ( new version ) of 70mA. The Buf634 has a maximum bias current of 15mA so it leaves class at loud volume on a low impedance, low sensitivity headphone. Is that one reason that my discrete circuit does work well with the demanding AKG 701 ?
Sofar i had no success using the LME49600 without servo but that does not mean it can not
be made to work without. Maybe it needs a stopper resistor, maybe something else.
John mentioned in another thread that a preamp should never leave class a.
Look at my headphone amp. It has an bias current ( new version ) of 70mA. The Buf634 has a maximum bias current of 15mA so it leaves class at loud volume on a low impedance, low sensitivity headphone. Is that one reason that my discrete circuit does work well with the demanding AKG 701 ?
Probably, 15 ma is to little for a headphone amp. Of course, it will work and measure 'OK' but it still can be bettered. I found the difference not in THD, in the amount of higher odd order distortion that gets generated, in my power amp designs.
When i raised the amount of class a in my headphone amp the sound got "softer ", more listanable, less edgy. I also had the feeling that the bass quality was better, deeper and more clear the same time.
One needs 40mA at least if one wants to drive terminated cables.
I have 70mA driving 200R.
The DAO follower (headphone) has 200mA, so chosen because it is the most linear region of the cascoded LU1014.
But it also drives 25 ohm phones with comfort.
I know it is not a bias contest. 😉
Patrick
I have 70mA driving 200R.
The DAO follower (headphone) has 200mA, so chosen because it is the most linear region of the cascoded LU1014.
But it also drives 25 ohm phones with comfort.
I know it is not a bias contest. 😉
Patrick
I agree, EUVL. I used about 40-50ma for the Parasound JC-2 and almost the same in the Levinson JC-2, 35 years before. It is best to operate the output devices where they are the fastest and most linear, if possible.
All Erno line amps use about 70 ma bias ( Mosfets at the output) - they can comfortably drive 50 Ohm headphones. They are also suitable for terminated cabling
kannan
kannan
10 times the diameter. Try chain nose pliers, hold then next to the component body and then bend the lead to avoid stressing the lead to inside joint.
I tend to use a lead bender to make the radius. I have personally never had any problems, BUT with very tight boards, compromises are a problem.
IPC-610-D "Acceptability of Electronic Assemblies" gives suitable recommendations in Table 4-1 for wires and as usual, the answer is "it depends".
Fixed coaxial cable - 5 x OD (outside diameter)
Flexible coaxial cable - 10x OD
Unshielded wires - 3xOD < AWG10, 5XOD >AWG10
Shielded wires/cables - 5x OD
Harness assembly - Bend radius equal to or greater than the minimum bend radius of any individual cable within the harness
And later on in Section 7 for through hole components, for those people still using this technology...
For components (where D = lead diameter or max thickness), minimum inside bend radius is:
D < 0.8mm - 1D
0.8mm < D < 1.2mm - 1.5D
D > 1.2mm - 2D
- Status
- Not open for further replies.