So I threw this together for lack of a better thing to do....

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

CurtZHP

Well-known member
Joined
Mar 21, 2005
Messages
634
Location
Allentown, PA
Lockdown induced boredom has compelled me to issue myself a challenge.  I tried to build something using ONLY what parts I had on hand.  I came up with an amusing attempt at a  vacuum tube headphone amp.

Hat tip to our esteemed colleague, Mr. Ian Thompson-Bell.  In my search for information on such a thing, I came across his design and borrowed from it.  (I may as well paint a mustache on the Mona Lisa while I'm at it!)

Much to my horrified glee (is that even a thing??), it actually works!  A tiny bit on the noisy side (very, very slight noise that only shows up on the scope), depending on where I stick my fingers; but it's basically just a pile of guts on the workbench, so that's not at all surprising.

For those of you in need of internet target practice, fire at will!  I'll be taking cover under the workbench, in the same place I was when I first plugged it in.

 

Attachments

  • Tube Headphone Amp.jpg
    Tube Headphone Amp.jpg
    386 KB · Views: 163
I'd put C1/2 on the input side of the gain pots, and they should be much larger, like 1mfd. 

R11/12 will be lower distortion and slightly higher gain if more typical 100K-220K. 

C5/6 could be much smaller (0.022 - 0.01), but you're done there so....

C9/10 into low-Z headphones should be very large, they're scaled now for 10KΩ.  330mfd or so for Sony's I think.    I see 470mfd in some.   

Not sure what level that 12BH7 stage is even capable of driving at that low an impedance, if at all.  Haven't thought about that. 

All that off the top of my head......





 
EmRR said:
I'd put C1/2 on the input side of the gain pots, and they should be much larger, like 1mfd. 

Dumb question:  are C1/2 even necessary?  I only put them in because I've seen that done elsewhere (different circuits), and I happen to notice that Ian didn't even include such a thing.


R11/12 will be lower distortion and slightly higher gain if more typical 100K-220K.

Might be able to scare something up.



C5/6 could be much smaller (0.022 - 0.01), but you're done there so....

Again, might be able to find something.  Starting to run low on poly caps, though.


C9/10 into low-Z headphones should be very large, they're scaled now for 10KΩ.  330mfd or so for Sony's I think.    I see 470mfd in some.   

That would mean using large electrolytics, which I assume is OK for this purpose.  Any downside to that?


Not sure what level that 12BH7 stage is even capable of driving at that low an impedance, if at all.  Haven't thought about that. 

I started out with a 12AX7 in there too.  (Don't laugh too hard...)  As expected, performance was....less than ideal.  Further reading found that others had done similar using the BH tubes with much better results, Ian included.  I happened to find a couple in the junk box, old and used, and likely not even close in terms of matching.

Thanks!
 
CurtZHP said:
Dumb question:  are C1/2 even necessary?  I only put them in because I've seen that done elsewhere (different circuits), and I happen to notice that Ian didn't even include such a thing.

Keeps DC off the pot if it's before it, minimize scratchiness.  Ian argues for one after because of grid bias DC I think?  But I've never seen that be a problem.  You might get away with none. 
CurtZHP said:
That would mean using large electrolytics, which I assume is OK for this purpose.  Any downside to that?

Well, you have to use electrolytic.  Unless you find some dynamite sized film caps, I do have some 180mfd films that look like a stick of dynamite somewhere.  Something like 8-10 inches long. 
 
EmRR said:
Keeps DC off the pot if it's before it, minimize scratchiness.  Ian argues for one after because of grid bias DC I think?  But I've never seen that be a problem.  You might get away with none.

I'll just move 'em.


Well, you have to use electrolytic.  Unless you find some dynamite sized film caps, I do have some 180mfd films that look like a stick of dynamite somewhere.  Something like 8-10 inches long.

I think I saw a picture of one once!  :D

Managed to find a pair of 200K resistors for R11/12.  They're 5W monsters, but what else am I gonna use them for?

 
EmRR said:
Keeps DC off the pot if it's before it, minimize scratchiness.
There should be no DC. Doesn't matter if it's before or after. But I would put it after since a smaller cap is going to perform better.
 
C1/C2 are fine where they are. Value probably OK too; 3dB down point is about 3Hz.

The big problem is the 12BH7 SRPP stage. Those two 330K resistors are far to big. Should be more like 330 ohms.

Size of C9 and C10 depend on what load you want to drive.

The most likely reason it is noisy is because you probably have about 60dB of gain. With the level pots at the input you have always got that gain and hence that noise. If you place the pots between stages then when you turn them down you also turn down the stage 1 noise.

Cheers

Iab
 
Ha, I didn't even look at that 330K. 

Is the 10K pot value NOT part of the response determination for C1/2? 
 
EmRR said:
Ha, I didn't even look at that 330K. 

Is the 10K pot value NOT part of the response determination for C1/2?

With the cap after the pot the low end response is determined principally by the 1Meg grid resistor and the series cap (as long as the pot value is significantly lower than 1Meg which it is). If you put the cap before the pot then the cap and the pot determine the LF response.

Cheers

Ian
 
ruffrecords said:
C1/C2 are fine where they are. Value probably OK too; 3dB down point is about 3Hz.

The big problem is the 12BH7 SRPP stage. Those two 330K resistors are far to big. Should be more like 330 ohms.

Size of C9 and C10 depend on what load you want to drive.

The most likely reason it is noisy is because you probably have about 60dB of gain. With the level pots at the input you have always got that gain and hence that noise. If you place the pots between stages then when you turn them down you also turn down the stage 1 noise.

Cheers

Iab

Sorry, Ian.  I must have mislabeled them.  They actually ARE 330Ohm!
 
ruffrecords said:
The most likely reason it is noisy is because you probably have about 60dB of gain. With the level pots at the input you have always got that gain and hence that noise. If you place the pots between stages then when you turn them down you also turn down the stage 1 noise.

Cheers

Iab

I actually kicked that idea around a bit.  As I mentioned, though, wiggling things a bit seems to make it come and go.  Probably just sloppy wiring.
 
CurtZHP said:
I actually kicked that idea around a bit.  As I mentioned, though, wiggling things a bit seems to make it come and go.  Probably just sloppy wiring.

Well the big question is do you really need 60dB of gain??

Cheers

Ian
 
CurtZHP said:
Probably not for a headphone amp.

For a headphone amp you will almost certainly need a step down output transformer because the circuit as it stands will not drive loads below about 2K. You could add some negative feedback to set the gain to a more reasonable level and to reduce the distortion in the SRPP output stage. Here is my solution from a few years ago:

phonesampcctsht1of2scaled.jpg


Cheers

Ian
 
ruffrecords said:
For a headphone amp you will almost certainly need a step down output transformer because the circuit as it stands will not drive loads below about 2K. You could add some negative feedback to set the gain to a more reasonable level and to reduce the distortion in the SRPP output stage. Here is my solution from a few years ago:

phonesampcctsht1of2scaled.jpg


Cheers

Ian

I was trying to get away with building it using just what I had on hand.  Haven't got any Sowters laying around.
It seems to be driving a pair of 600ohm cans right now, but that's probably just by sheer brute force (60dB?).
Not quite ready to dump $200 worth of iron into it.
 
ruffrecords said:
The most likely reason it is noisy is because you probably have about 60dB of gain. With the level pots at the input you have always got that gain and hence that noise. If you place the pots between stages then when you turn them down you also turn down the stage 1 noise.


I think the noise was me.  I happened to notice that it came and went whenever I was adjusting the volume.  With no knob on the pot shaft, I was basically touching the bare metal shaft.  As long as I'm not touching anything, it's clean.  Good thing I got the power supply sorted out beforehand!
 
CurtZHP said:
I was trying to get away with building it using just what I had on hand.  Haven't got any Sowters laying around.
It seems to be driving a pair of 600ohm cans right now, but that's probably just by sheer brute force (60dB?).
Not quite ready to dump $200 worth of iron into it.
[/quote

You don't need to spend $200 on iron. The first prototype I built used $15 Edcor transformers. For 600 ohm cans you could use an XSM10K/600 at just under $19 each.

Cheers

Ian
 
ruffrecords said:
You don't need to spend $200 on iron. The first prototype I built used $15 Edcor transformers. For 600 ohm cans you could use an XSM10K/600 at just under $19 each.

Cheers

Ian

That's good to know.  I figured there had to be an alternative.  Those Sowters gave me sticker shock.
I wonder if OEP has something viable.  I've been using their input transformers in mic preamps and really like them.

I figure I'll go ahead and finish this thing as is and leave room for the transformers to be added once I find them.
 
..your TL783 regulator will probably very much like a layer or two of protection: 3x 39V Zeners protecting it from higher than 120V across (e.g. at power-up), and a 1n4007 protecting it from negative voltage across (e.g. at power-down).

And depending on the winding resistance of your secondary transformer, a series resistor (470 R?) will help a lot to protect against ramping-up voltage differentials..

Look at the G9 supply, that was the minimum configuration I could get long-term-stable :)

/Jakob E.
 
gyraf said:
..your TL783 regulator will probably very much like a layer or two of protection: 3x 39V Zeners protecting it from higher than 120V across (e.g. at power-up), and a 1n4007 protecting it from negative voltage across (e.g. at power-down).

And depending on the winding resistance of your secondary transformer, a series resistor (470 R?) will help a lot to protect against ramping-up voltage differentials..

Look at the G9 supply, that was the minimum configuration I could get long-term-stable :)

/Jakob E.

I have.  The last two builds I did included those features.  If this thing behaves, I may just spend a few bucks and add that.  Now that I think of it, I might have those parts already.  Just need a pair of output transformers.  Speaking of which, would the same OEP output transformers used in the G9 also work here?
 

Latest posts

Back
Top