[tomchr]

Having the ability to drive both speakers and headphones would certainly be nice. I think I have found a way to do it. I just need to verify that it'll work (and what the performance is).
 
I've seen designs where the secondary of the OPT is loaded by, say, 50~100 ohm. The speaker or headphone is connected in parallel with this load. I think that should work pretty well and would allow someone to connect a headphone to their speaker amp (or vice versa). It would, however, require that the OPTs are wound for use with speakers (4/8 ohm tap). That said, I'd like to verify the performance before recommending the circuit. Expect results by mid September...
 
~Tom

 

[tomchr]

Quote:

Jack at Electro-Print wound these for Audeze LCD2's.  50 ohm I think.
Jack can do Anything and although I have no experience with other OPT's I am glad I had him make them.
Very Old School.

 
Jack's good people. He can do anything the laws of physics will allow. I was impressed by his honesty and interest in customer success. He's not just a product pusher. He wants you to succeed using his products. Big difference. That's my mode of operation as well.
 
Electra-Print. (No affiliation, just happy customer).
 
~Tom

 

[muskyhuntr]

Hi All,
 
Looking forward to what Tom comes up with.  I was originally planning on driving a set of speakers with mine, but would add head phone functionality in a second!  As of now, I have all of my power supply boards done (only the filament boards tested) and the 300B board populated with the resistors. If I am understanding Tom correctly, we are talking only about changes to the design from the output transformers on?  Any changes on the boards?  I don't plan on putting any music through this until next year, so I have plenty of time to incorporate design changes!
 
Jim

 

[sceleratus]

Quote:

Hi All,
 
Looking forward to what Tom comes up with.  I was originally planning on driving a set of speakers with mine, but would add head phone functionality in a second!  As of now, I have all of my power supply boards done (only the filament boards tested) and the 300B board populated with the resistors. If I am understanding Tom correctly, we are talking only about changes to the design from the output transformers on?  Any changes on the boards?  I don't plan on putting any music through this until next year, so I have plenty of time to incorporate design changes!
 
Jim

NEXT YEAR !!!

 

[holland]

Quote:

Having the ability to drive both speakers and headphones would certainly be nice. I think I have found a way to do it. I just need to verify that it'll work (and what the performance is).
 
I've seen designs where the secondary of the OPT is loaded by, say, 50~100 ohm. The speaker or headphone is connected in parallel with this load. I think that should work pretty well and would allow someone to connect a headphone to their speaker amp (or vice versa). It would, however, require that the OPTs are wound for use with speakers (4/8 ohm tap). That said, I'd like to verify the performance before recommending the circuit. Expect results by mid September...
 
~Tom

 
Yeah, I've seen that around lots.  A parallel and possibly a series network to do 2 things, bring voltage down and to reflect a load closer to 8 ohm.  Usually it's a 9-10 ohm power resistor loading the secondaries.  I'm not really a fan of that.
 
I was, however, thinking about an interstage transformer....possibly, to bring down gain.  Possibly of a switchable auto transformer type.  I haven't worked the math to see if it's viable.  These are just musings as this is so far down on my list of things to do, after a number of other headphone amps and a power amp or two.  I'm going back to SS for a bit after I explore a balanced parafeed topology for a few months.

 

[sceleratus]

Quote:

The bakelite volume knob wasn't doing it for me so I went looking for someone that could turn one out of maple for me.
Today I found a great guy.  He might even try a couple inlay options I drew up.
Here are a couple thoughts.
 
 

 

Knob Update:
 
I found a gentleman in Vista, CA thay does wood turning.
He wasn't too keen about a metal inlay.  I asked about wood and he thought ebony would work well.
That sound good and it make sense to have a wood guy do wood.
 
[size=small]Looking for opinions:[/size]
[size=small]I have 3 choices.[/size]
[size=small]1) An inlayed "dot"[/size]
[size=small]2) An inlayed 1/8" sliver, "flush"[/size]
[size=small]3) An inlayed 1/8" sliver  "protruding"  (See diagram)[/size]
 
[size=small]#3 Would require the most work but the added dimension could be cool.[/size]

 

[tomchr]

I'd go with either a flush dot or a flush sliver. If you want to go with dark wood for the sliver, why not continue the walnut theme?
 
~Tom

 

[sceleratus]

I think ebony is harder.  It's also darker so a small piece will stand out more.

 

[tomchr]

Quote:

Yeah, I've seen that around lots.  A parallel and possibly a series network to do 2 things, bring voltage down and to reflect a load closer to 8 ohm.  Usually it's a 9-10 ohm power resistor loading the secondaries.  I'm not really a fan of that.

 
Triodes tend to like high-impedance loads. Higher load impedance --> higher gain, lower distortion. You just want enough load on the secondary that the inductive flyback is tamed. There is some uncertainty around how the OPT will operate under these conditions that make me want to try it out before recommending it. Yeah... I'm not a fan of burning gobs of power in a resistor to allow for use with headphones either. 
 
~Tom

 

[tomchr]

Quote:

I think ebony is harder.  It's also darker so a small piece will stand out more.

 
But it'll be a different color than the walnut. Arh... Grain structures. Splitting. Issues I don't have working in metals... Oh, well. It looks like it'll be a really nice box once you're done. I can see how that bakelite knob doesn't quite fit the high-end image. A nice solid maple knob would be much nicer.
 
Is the indicator needed? Don't you usually turn it up until it's loud enough anyway?
 
~Tom

 

[sceleratus]

Quote:

 
But it'll be a different color than the walnut. Arh... Grain structures. Splitting. Issues I don't have working in metals... Oh, well. It looks like it'll be a really nice box once you're done. I can see how that bakelite knob doesn't quite fit the high-end image. A nice solid maple knob would be much nicer.
 
Is the indicator needed? Don't you usually turn it up until it's loud enough anyway?
 
~Tom

The indicator is mostly a design detail.

 

[tomchr]

Quote:

Originally Posted by muskyhuntr /img/forum/go_quote.gif
If I am understanding Tom correctly, we are talking only about changes to the design from the output transformers on?

 
Correct. The headphone mod would attach to the secondary of the output transformer. It would allow for seamless switching between headphone and speaker. I need to try it out first, though.
 
I also need to update my design collateral with the component values for the E88CC. To change to the E88CC, one of the LEDs is shorted out and two 1/4 W resistors need to be changed to different values. It's a pretty flexible driver board. I must say that I really like the E88CC as the driver tube. The 6N6P comes in as a close second followed by ECC99. But that's my ears, my speakers. I use JJ E88CC tubes ($12/ea). Nuttin' fancy.
 
~Tom

 

[holland]

It would be interesting to see how the 300B reacts to a reflected load of over 30Kohms.  If it does measure the same (minus some power perhaps), then scelartus could have used a 31250:50 transformer.
 
Also some OPTs react to it OK, others don't and would prefer to see the nominal load of it's secondaries.  Though I think actual impedances might fluctuate more than that, and the parallel resistor would normalize them (affecting sound).  Though I do still wonder about an interstage transformer between the driver and cathode follower, to knock down some voltage gain, coupled with secondaries on the OPT for impedance matching.
 
Another note, you mentioned doing something similar before, you can also apply a series resistor after the parallel to bring the output impedance up some more.  Some people like it, some people don't.  I have to admit to being one of those that prefer output impedances of 120ohms on some headphones (Beyerdyanmic), not so much on others (HD650).

 

[tomchr]

As long as the OPT bandwidth remains the same under light load as under nominal load, I don't really foresee any issues with the light load. It should give lower THD. The higher gain would probably not be noticeable (though, probably measurable). The mu of a 300B is about 4 if my memory serves. So with the nominal load, you might get a gain of 2 out of it. With the lighter load, perhaps you get 3.5. That's about 5 dB. Not enough to write home about.
 
~Tom

 

[sceleratus]

The Misumi L-bracket for the volume attenuator arrived and thankfully it fits.
 
One unexpected downside is the knob is spongy when connected to a rotary switch.  The reason is the extension shaft is a very tight spring, if there is any resistance when you turn it, it needs to build up torque before it can "snap" and move the switch.  This give the knob a very strange feel.  A volume pot would not have this problem.
 
Nonetheless I am now going to order the Goldpoint 23 step attenuator.  Hopefully the detents are softer than the cheap import rotary switch I used for testing.  I don't have another option so it is....going to be this way.  I'm very happy because the clearances are on the order of 1/8".  I was lucky.
 
Note the use of a $20 filament PCB as a base for the bracket.