Class A, DHT driven, headphone amp - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Headphone Systems

Headphone Systems Everything to do with Headphones

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 1st July 2009, 06:59 PM   #11
diyAudio Member
 
dsavitsk's Avatar
 
Join Date: Jan 2005
Location: Hartford
Quote:
Originally posted by astouffer
You could always use a DHT for gain and a buffer chip like a BUF634 or HA-5002.

These chips really sound awful, and both need to be inside feedback loops of opamps or else they need output caps.
  Reply With Quote
Old 1st July 2009, 07:41 PM   #12
bigwill is offline bigwill  United Kingdom
diyAudio Member
 
Join Date: Dec 2004
Location: UK
Quote:
Originally posted by dsavitsk
The coupling cap is going to undo any goodness done by the DHT. It is the bottleneck of the amp.
Pfff. A properly biased electrolytic sounds fine
  Reply With Quote
Old 1st July 2009, 08:24 PM   #13
Gordy is offline Gordy  United Kingdom
diyAudio Member
 
Join Date: Nov 2006
Quote:
Originally posted by ikoflexer

As for headphones, I have the run of the mill Sony MDR-V6, claiming 106 dB/mW, 63 ohm at 1kHz.
OK, then 8 mW will very likely be more than loud enough! Allow a bit above that for 'headroom'. Say 32 mW.

Quote:
Originally posted by ikoflexer
[B]
The all of a sudden a power triode + transformer output doesn't look so unappealing, considering that it does away with the extra psu needed for the mosfet output (just thinking aloud here).
Volume control > 6H30 (run at about 25 mA from a CCS load, bias about -3V, about 80V on the anode) > small (about 3 microfarad) parallel-feed capacitor from anode > output transformer primary (with the other end of primary returned to ground).

For ac coupled o/p transformer see Sowter or Electra-Print (or possibly Hammond 119DA).

Quote:
Originally posted by ikoflexer

rman, I was thinking perhaps a DC servo might work, as wavebourne has done (somewhere else on this forum).
...but if the servo fails you get hundreds of volts very close to your head?


: )
  Reply With Quote
Old 1st July 2009, 09:50 PM   #14
Sheldon is offline Sheldon  United States
diyAudio Member
 
Join Date: Dec 2003
Location: San Diego
No need to be afraid of OPT's. Your power requirements are very low, and your phones are pretty high impedance. Edcor has SE transformers that start at around $20. At 10W they will only go to about 70Hz. But if you really only need 50mW or so, the LF extension will be much lower. You can even regap the transformer for more inductance, since your current will be low. You should easily be able to get 50mW output from a paralleled 3A5.

Sheldon
  Reply With Quote
Old 2nd July 2009, 02:34 AM   #15
iko is offline iko  Canada
diyAudio Moderator
 
iko's Avatar
 
Join Date: May 2008
Location: Toronto
Alright, I'm back.

OK, I would not want to use an opamp. I think I should add that to the list of requirements.

Coupling cap; the general consensus seems to be that direct coupling is better, but yes, in my case there is an extra mosfet. I'll try again for a design with a coupling cap to have something to fall back on in case direct coupling would not work well.

dsavitsk, I've had a look at your site for some time now, nice! I have all the parts to build the mosfet version, and if it sucks I'll swallow the bitter pill and buy a couple of OTs. Yes, I had the Edcors in sight. Sowter and such, more money than I want to spend.

I should say that I built a 21LU8 based headphone amp (pentode part triode strapped), RC interstage coupling, power transformers re-purposed to OTs. Total cost not more than $50 and it sounds great. Now if I go to a cost of $100 for the DHT amp, it better be twice as good; or so I hope. You may laugh, but that's my personal challenge.

Gordy, the 6H30 has good rep, but I really want a DHT.

Sheldon, good tip about paralleling the 3a5. Didn't think about that. I have some other DHTs coming my way that I will consider for this project. Some 1j6g, 30, and even 71a.

Weird that from some comments I find on the web it seems so easy to build an excellent amp. Makes one wonder then why all those people over at headwize expend so much effort to get that special sound.

bigwill, what do you mean by a properly biased coupling cap? Sorry, I just don't understand what you meant by biasing in this context. Are you referring to the RC filter value?

Edit: typo
  Reply With Quote
Old 2nd July 2009, 04:52 AM   #16
Sheldon is offline Sheldon  United States
diyAudio Member
 
Join Date: Dec 2003
Location: San Diego
If you want to stay with the FET output, you might consider a simpler design. The 3a5 is a very linear tube. I think you'd do fine with a resistor load. Here's a simple output stage: http://www.tubecad.com/april_may2001/page13.html

The filament as a load wouldn't work with the DHT, but you can substitute a resistor there too. Without a transformer, the coupling cap that will influence the sound most is on the output. With that in place, I wouldn't add all the extra stuff to direct couple to the FET stage. You can use a good quality cap there at little expense.

You are going to have to have a good filament supply. I'd spend effort there. You will need a very quiet supply, constant current is a good choice. Search Rod Coleman's design. It looks complicated, but it's not if you break it down. Batteries are an option for a 3A5.

Sheldon
  Reply With Quote
Old 2nd July 2009, 12:06 PM   #17
bigwill is offline bigwill  United Kingdom
diyAudio Member
 
Join Date: Dec 2004
Location: UK
Quote:
Originally posted by ikoflexer
Alright, I'm back.

bigwill, what do you mean by a properly biased coupling cap? Sorry, I just don't understand what you meant by biasing in this context. Are you referring to the RC filter value?

Simply having a big enough DC voltage across the capacitor to accommodate the voltage swing
  Reply With Quote
Old 2nd July 2009, 07:44 PM   #18
iko is offline iko  Canada
diyAudio Moderator
 
iko's Avatar
 
Join Date: May 2008
Location: Toronto
sheldon, interesting output stages over at tubecad. Ok, I've simplified the abomination somewhat. Have a look.

Click the image to open in full size.

You will see the simple ccs load on the 3a5. I use double the jfets (if different Idss the source resistor needs to be adjusted offline until the voltage drop across the two jfets is roughly equal; the reason for this is that the 2sk170 has a 40V Vds max; I know it's a hack, but will look for more suitable jfet, with higher voltage).

Funny, that simple ccs has this output impedance. Over 100kOhm at 1Mhz. Of course, in the real world it's not like that

Click the image to open in full size.

So, the simpler circuit simulates a bit better for distortion. Nice figures for the voltage on the 3a5 plate (V(dplate)).

Code:
Harmonic	Frequency	 Fourier 	Normalized	 Phase  	Normalized
 Number 	  [Hz]   	Component	 Component	[degree]	Phase [deg]
    1   	1.000e+03	1.335e+00	1.000e+00	    1.02	    0.00
    2   	2.000e+03	1.854e-04	1.389e-04	 -130.62	 -131.64
    3   	3.000e+03	2.844e-05	2.130e-05	   55.56	   54.54
    4   	4.000e+03	2.726e-06	2.042e-06	   11.59	   10.57
    5   	5.000e+03	4.017e-07	3.009e-07	  -94.13	  -95.14
    6   	6.000e+03	1.669e-07	1.250e-07	  105.29	  104.28
    7   	7.000e+03	1.916e-07	1.435e-07	   78.39	   77.37
    8   	8.000e+03	1.755e-07	1.315e-07	   73.31	   72.30
    9   	9.000e+03	1.813e-07	1.358e-07	   72.40	   71.38
Total Harmonic Distortion: 0.014049%


Fourier components of V(dplate)
DC component:85.6847

Harmonic	Frequency	 Fourier 	Normalized	 Phase  	Normalized
 Number 	  [Hz]   	Component	 Component	[degree]	Phase [deg]
    1   	1.000e+03	8.646e+00	1.000e+00	  179.64	    0.00
    2   	2.000e+03	8.206e-04	9.492e-05	   74.33	 -105.31
    3   	3.000e+03	6.109e-05	7.066e-06	 -140.98	 -320.62
    4   	4.000e+03	1.102e-06	1.275e-07	   22.94	 -156.70
    5   	5.000e+03	2.250e-06	2.602e-07	   93.33	  -86.32
    6   	6.000e+03	2.151e-06	2.488e-07	  111.49	  -68.15
    7   	7.000e+03	2.090e-06	2.417e-07	  143.39	  -36.25
    8   	8.000e+03	2.807e-06	3.246e-07	  123.89	  -55.75
    9   	9.000e+03	1.581e-06	1.828e-07	  113.01	  -66.63
Total Harmonic Distortion: 0.009518%


Fourier components of V(couple)
DC component:8.42537

Harmonic	Frequency	 Fourier 	Normalized	 Phase  	Normalized
 Number 	  [Hz]   	Component	 Component	[degree]	Phase [deg]
    1   	1.000e+03	1.360e+00	1.000e+00	   -0.41	    0.00
    2   	2.000e+03	1.504e-04	1.106e-04	 -123.51	 -123.10
    3   	3.000e+03	1.764e-05	1.297e-05	   52.67	   53.08
    4   	4.000e+03	1.318e-06	9.691e-07	    0.02	    0.43
    5   	5.000e+03	6.421e-07	4.721e-07	  -81.55	  -81.14
    6   	6.000e+03	3.784e-07	2.782e-07	  -74.86	  -74.45
    7   	7.000e+03	3.488e-07	2.565e-07	  -71.25	  -70.84
    8   	8.000e+03	3.301e-07	2.427e-07	  -75.17	  -74.76
    9   	9.000e+03	2.918e-07	2.145e-07	  -52.70	  -52.28
Total Harmonic Distortion: 0.011134%
sheldon, using a suitable resistor to get the 3a5 in the same operating point simulates as seen below. I'm inclined to stay with the simple ccs. What do you think?

Code:
Fourier components of V(out)
DC component:-8.26397e-008

Harmonic	Frequency	 Fourier 	Normalized	 Phase  	Normalized
 Number 	  [Hz]   	Component	 Component	[degree]	Phase [deg]
    1   	1.000e+03	6.720e-01	1.000e+00	    1.20	    0.00
    2   	2.000e+03	4.781e-03	7.115e-03	  -89.77	  -90.97
    3   	3.000e+03	1.080e-04	1.607e-04	    1.38	    0.18
    4   	4.000e+03	3.678e-06	5.473e-06	   85.54	   84.35
    5   	5.000e+03	2.154e-07	3.206e-07	 -123.01	 -124.20
    6   	6.000e+03	1.772e-07	2.637e-07	  -71.79	  -72.99
    7   	7.000e+03	1.748e-07	2.602e-07	  -68.22	  -69.42
    8   	8.000e+03	1.764e-07	2.625e-07	  -65.25	  -66.45
    9   	9.000e+03	1.775e-07	2.641e-07	  -61.29	  -62.49
Total Harmonic Distortion: 0.711664%


Fourier components of V(dplate)
DC component:85.7718

Harmonic	Frequency	 Fourier 	Normalized	 Phase  	Normalized
 Number 	  [Hz]   	Component	 Component	[degree]	Phase [deg]
    1   	1.000e+03	4.352e+00	1.000e+00	  179.83	    0.00
    2   	2.000e+03	3.092e-02	7.106e-03	   89.91	  -89.91
    3   	3.000e+03	6.925e-04	1.591e-04	  179.69	   -0.14
    4   	4.000e+03	2.173e-05	4.994e-06	  -93.44	 -273.27
    5   	5.000e+03	2.485e-06	5.711e-07	   63.45	 -116.37
    6   	6.000e+03	2.447e-06	5.624e-07	  105.75	  -74.08
    7   	7.000e+03	2.930e-06	6.734e-07	   97.13	  -82.69
    8   	8.000e+03	2.847e-06	6.543e-07	  105.15	  -74.68
    9   	9.000e+03	2.739e-06	6.295e-07	  113.49	  -66.33
Total Harmonic Distortion: 0.710828%


Fourier components of V(couple)
DC component:8.41166

Harmonic	Frequency	 Fourier 	Normalized	 Phase  	Normalized
 Number 	  [Hz]   	Component	 Component	[degree]	Phase [deg]
    1   	1.000e+03	6.845e-01	1.000e+00	   -0.23	    0.00
    2   	2.000e+03	4.865e-03	7.107e-03	  -90.34	  -90.11
    3   	3.000e+03	1.093e-04	1.597e-04	    0.16	    0.39
    4   	4.000e+03	3.562e-06	5.204e-06	   85.10	   85.33
    5   	5.000e+03	4.304e-07	6.288e-07	  -95.38	  -95.15
    6   	6.000e+03	4.220e-07	6.165e-07	  -76.21	  -75.98
    7   	7.000e+03	3.722e-07	5.438e-07	  -72.80	  -72.57
    8   	8.000e+03	4.392e-07	6.416e-07	  -62.21	  -61.98
    9   	9.000e+03	5.011e-07	7.320e-07	  -68.29	  -68.06
Total Harmonic Distortion: 0.710885%
  Reply With Quote
Old 2nd July 2009, 08:07 PM   #19
diyAudio Member
 
dsavitsk's Avatar
 
Join Date: Jan 2005
Location: Hartford
Quote:
Originally posted by ikoflexer
sheldon, interesting output stages over at tubecad. Ok, I've simplified the abomination somewhat. Have a look.

First, why not use a part for the CCS that is actually intended for this purpose and that will work better: DN2540 or IXYS 10M45. Both are depletion mode mosfets and both will work better here. Moreover, they are 400 and 450V parts.

Click the image to open in full size.

Second, why are you using two mosfet followers? Why not just take the headphone out from the first one? So you'll need a bigger heatsink. But, you already have about 30mA through the first fet, so just bump it up a tad and save yourself a second mosfet, ps, ccs, etc.
  Reply With Quote
Old 2nd July 2009, 09:19 PM   #20
iko is offline iko  Canada
diyAudio Moderator
 
iko's Avatar
 
Join Date: May 2008
Location: Toronto
Quote:
Originally posted by dsavitsk



First, why not use a part for the CCS that is actually intended for this purpose and that will work better: DN2540 or IXYS 10M45. Both are depletion mode mosfets and both will work better here. Moreover, they are 400 and 450V parts.

Click the image to open in full size.

Second, why are you using two mosfet followers? Why not just take the headphone out from the first one? So you'll need a bigger heatsink. But, you already have about 30mA through the first fet, so just bump it up a tad and save yourself a second mosfet, ps, ccs, etc.
Second suggestion I've tried initially, when I first looked at direct coupling, but couldn't get it working well. Didn't spend much time trying to fix it though, I'll revisit the issue. It'd be nice, of course, to get rid of using a whole lot of parts.

Regarding the DN2540 ccs; I've also tried that, but for some reason the simulated output impedance was not very good at all. I can give more details if interested.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
simplest LED circuit driven by mp3 player headphone amp groovebot Headphone Systems 23 12th March 2010 02:45 PM
Very nice sounding chipamp-driven working class A amp ionomolo Chip Amps 66 29th March 2009 12:03 AM
Class D driven active Transducer Happyface Class D 2 22nd March 2009 03:09 PM


New To Site? Need Help?

All times are GMT. The time now is 02:50 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2