Cavalli Compact Tube Hybrid (CTH) Tube & Tweak thread.
Jun 1, 2010 at 4:36 AM Post #526 of 691
Would it be possible to use a 6SN7 tube in this amp?  Besides the different pinout, the CTH could manage the heater requirements and all the 6SN7 specs look reasonably close to other working tubes.
 
Jun 5, 2010 at 5:44 AM Post #528 of 691
Just replaced my Alps RK27 pot with a TKD 2CP-2508 and the improvement is astounding.  While it was expensive, I'd definitely recommend it for someone's upgraded the caps and has a nice tube in their build already.  The bass is much tighter and deeper though that might be because I went from a 50k to 100k pot and my DAC has a cap on its output.  The sound is also much smoother.  Whereas there used to some sounds that seemed too loud and harsh especially during female vocals, the TKD pot doesn't seem to bite on these frequencies like the Alps did.  Since the higher frequencies aren't as rough, I can make out a lot more detail.  One interesting effect is that the soundstage seems a little farther back which feels different especially with Grados.  And the knob turns so smoothly!
 
While it wouldn't be the first thing I'd upgrade if I were to build a second CTH, I wouldn't be done upgrading without a better pot.
 
Jun 6, 2010 at 5:16 PM Post #529 of 691
That's great to hear:)  I've been considering getting a couple of the RK40s that have been up on the bay lately @ $25 or so a pop + shipping.
Even tho they are new old stock, perhaps quite old stock, they probably would still represent an improvement.
 
Jun 11, 2010 at 4:22 PM Post #530 of 691
Is there a way to reduce the gain of this amp?  I'm already using a 100ohm output resistor and find myself using the bottom 15 degrees of my pot at most.  Even with a nice pot, the channel matching is poor at the very bottom end.  I'd also like to not use the output resistor since its probably not ideal with Grados.  Maybe some way to add some negative feedback?  I'm also using a  6JD8 tube.  Maybe there are nice tubes with significantly lower gain?
 
EDIT This amp has way too much gain for every pair of headphones I've used with it, not just Grados.
 
Jun 11, 2010 at 8:03 PM Post #532 of 691
This is one of the downsides of listening to Grado's: they are very easy to drive.  I'm in the same situation.
 
Sticking to 12AU7 and variant tubes will give you lower gain.  You could also add a resistor (50-100K) on the left and right channel between the input jack and the volume pot to attenuate the input signal.  This will probably yield another 30 degrees of pot rotation.
 
Jun 12, 2010 at 10:02 PM Post #533 of 691
You could also put resistance in front of the pot.  In my opinion this is the best option, assuming you choose your pot and resistor values such that the input resistance is not too high.
However, the output resistors have their own function which is to reduce "hiss" from the amp.
 
Jun 13, 2010 at 9:24 AM Post #534 of 691


Quote:
You could also put resistance in front of the pot.  In my opinion this is the best option, assuming you choose your pot and resistor values such that the input resistance is not too high.
However, the output resistors have their own function which is to reduce "hiss" from the amp.


Any rules of thumb for determining if input resistance is too high?
 
Jun 13, 2010 at 2:43 PM Post #535 of 691


Quote:
Any rules of thumb for determining if input resistance is too high?


Depending on who you ask 10k - 100k ohms is a good input input impedance.  So as an example you could put 10k resistors in front of a 10k pot for 20k input impedance, or 50k in front of a 50k pot for 100k input impedance.  I think either one would be perfectly fine.
 
Jul 16, 2010 at 2:22 PM Post #536 of 691
Jul 17, 2010 at 12:09 AM Post #538 of 691
So I'm determined to add some negative feedback to my CTH.  I just hate having to turn up the volume just to center the music and the idea of further attenuating the signal before amplifying it so much seems silly.  I spent forever working through the schematics with my very basic understanding of tubes and transistors and ended up deciding that the input stage is inverting which the output buffer is not.  And then a minute of google searching confirmed what took me what seemed like an hour:
 
Quote:
For CTH, gain is directly related to the tube. If I recall the circuit properly, and you bump the input resistor up to around 10K, you should be able to feed back the output to the input via a 100K resistor (or larger). The output from the tube should be inverting and the buffer is non-inverting.

Give it a try or yell at me for even broaching this.
smily_headphones1.gif


The problem is that the output is biased up 12V so there would need to be a cap in the feedback loop.  The cap would be pretty small though to as long as you don't add much feedback.  I'd like to add variable feedback a la the bijou.  Instead of this global feedback, local feedback just around the tube would be possible in basically the same way but this might present too large of a load to the tube.  At the same time, I feel like local feedback might be less influential on the overall sound signature than global feedback though I have no logic to back that up.
 
Any suggestions?
 
 
Jul 17, 2010 at 12:49 PM Post #539 of 691
If runeight does not see this & offer thoughts, maybe you could PM him referencing this thread/post.  Something like you describe could be beneficial to many CTH users.
 
A less sophisticated way to go could be finding a low/lower-mu tube that you enjoy.  Or using something like a black beauty that works better across its range.  My next, and probably last, CTH will have one.  Think it'll be the 5th one for me & will be even fancier than my "fancy" one:)
 
Jul 18, 2010 at 11:48 PM Post #540 of 691
I have been alerted!!
 
It is very hard to add NFB to this amp or I probably would have done it in the original design. As mudgecoffee has pointed out the output is at 12VDC while the input is at ground. You will need a big capacitor in the feedback loop to get a good frequency response so I don't think this is practical.
 
Also, because of the phase inversion the NFB can only go back into the grid of the input tube as the NFB in the Bijou. This technique has drawbacks and in particular it drops the Zi of the amp quite a bit.
 
So, I don't think this is possible. If you want an amp that can still roll a lot of tubes and has the NFB to more or less equalize the gain among the tubes you might have to try the SOHA II.
 

Users who are viewing this thread

Back
Top