Reply
 
Thread Tools Display Modes
  #1   Report Post  
Posted to rec.audio.tubes
Patrick Turner Patrick Turner is offline
external usenet poster
 
Posts: 3,964
Default Long tail differential pair?? Who has circuit like that?

I have no idea if this reply post will appear at rec.audio.tubes because of recent history with not getting access via Google from where I am in Australia.

On Monday, 3 February 2014 00:12:21 UTC+11, wrote:
Yes, I'm building a linestage with my technician, and based on this circuit:
https://web.archive.org/web/19990223...train1_prj.htm


I just wanna to have a balanced version from the link of above.
The circuit should have two dual-triode 6SN7 tubes per channel. The lower tube is set up as a long-tailed differential pair. one of the grids is grounded, the other is the input grid from the volume control. Input signals are amplified by this stage and converted to differential (push-pull, or balanced ) outputs by this circuit. One half of the second dual triode is the follower for the one half of the lower tube, and the second half of the upper tube is the follower for the other half of the lower tube.


Imagine that the lower tube is a differential pair and there is a second follower above the other half of the differential pair. Two tubes, single-ended input, balanced outs."

Anyone build something like that before. My friend has sent me a message, it might need a negative PS from what he learnt....??


You could take a look at my website which has a lot about amplifiers and load matching. http://www.turneraudio.com.au

Last lime I made a headphone amp for a customer it had 6BQ5 in SE triode with SE OPT about 7k0 : 16r and 4r taken from old AM/FM receiver from 1960s. I had 1/2 a 12AU7 driving the 6BQ5 and with 12dB GNFB. It was extremely quiet with resistance divider to drive headphones and the sound was glorious. The unit could drive an 8 ohm speaker OK and headphones of 32r were easily driven off the R divider from OPT sec.

But if you don't want an OPT, and you can live with an unbalanced input, but want a balanced output able to power 32 ohms, I'd be thinking of a differential input pair with 6CG7
MJE340 transistor cathode current sink. RLa for each 1/2 6CG7 would be say 50k, with Ia at least 4mAdc. Then you can direct couple each anode to grids of a pair of 6BQ5 in triode mode and set up as direct coupled cathode followers. The dc carrying load of 6BQ5 cathodes could each be 10k taken to -150Vdc rail.
Probably much better would be 4 x EL86 arranged as a pair of White cathode followers, and the bottom tube acts as a live current source. You'd be surprised by what can be done with a White CF!
The two 6BQ5 cathodes will be at say +150Vdc potential. The output load of 32r may be driven from cathode to cathode. Each 6BQ5 sees a load of 16r, so its a very poor load match for the tube and THD will be high even and low Vo levels.
However the THD of each tube is 2H so you get cancelling with the PP operation which leaves the 3H. maybe 0.1Vrms is enough for loudest listening, and that means tube current into 16r = 6.25mA rms, or 8.8mA pk. Idle current in each 6BQ5 should be 30mA.
But to connect the 32r phones cathode to cathode is dangerous because the phones are at +150Vdc, and maybe some unbalanced Idc flows - not good.
So then you need 1,000uF from each cathode to each output terminal, with say 22k to 0V to keep the output terminals at 0Vdc.
Of course another method would have the pair of 6BQ5 followers driving a small 5VA rated PP OPT of say at least 2k2 : 8r0, but don'r ask me where you'd get one. The OPT allows lower Iadc of say 15mA. So OPT primary may have say 5k0 between CT and 0V for 30mA flow. The 8r Sec powers phones of 32r, and load tubes get is 8k8 at extremely low THD. Hammond might have something entirely suitable and yet small and cheap.

Headphone amps need to be much quieter than a normal power amp, I like no more than 50uV. This is 1/20 of 1mV which is often normal from a power amp. This is the reason you see a resistance divider driving the headphone outlet on many power amps. The divider can say 33r + 8r2. This means the power amp works in class A even if its an SS amp with low bias current in bjts. The divider reduces the Vo level which is much less than that used to power a speaker. It also reduces the noise.
A typical good quiet sample of 1/2 6CG7 may produce 16uV noise at its anode even with grounded grid. EL84 or EL86 in triode and in CF mode will be quiet enough.

Just my 2C worth, many ways to make a headphone amp.......

Patrick Turner.


Thanks in advance!


Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Long Tail Phase Splitter John L Stewart Vacuum Tubes 5 March 16th 13 11:03 PM
long tail pair grounded grid question? MarkX Vacuum Tubes 14 August 27th 09 02:28 PM
What a Long Tail ! Thousands of goods, which is your favourite ? [email protected] Vacuum Tubes 1 April 5th 08 09:08 PM
Tail Resistor in long tail pair with CCS. Vacuum Tubes 13 October 15th 05 05:13 PM
Long Tail Pair with SS CCS in the Cathodes Vacuum Tubes 11 September 24th 05 03:55 PM


All times are GMT +1. The time now is 11:20 PM.

Powered by: vBulletin
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 AudioBanter.com.
The comments are property of their posters.
 

About Us

"It's about Audio and hi-fi"