Hi
I posted some thoughts and photos on this amp on show and tell page. Overall pleased but, as I mentioned in that post, the gain is way to high for my source. Looking for suggestions to tame it.
The amp is a 300B design. The first section is a single 12AX7 split between channels feeding a pair of 12AU7s paralleled to each channel. My source is a Denafrips DAC with RCA output at 2.0Vrms, 625 Ω
Options:
1. Lower gain tube in V1. I tried a 5751 (which has 30% less gain than the 12AX7) in the front. Their was an almost indiscernible decrease in level. I also tried an 12AT7 with similar results but a slight loss of quality. No help.
2. Any thoughts on changing out resistors or caps to lower voltage/gain in pre section while keeping tubes in designed operating parameter?
Don’t know if it helps. Here’s the spec idle voltages from schematics: 12AX7: Cathode: 0.8V. Anode: 110V. Grid: ?? 12AU7: Cathode 8.7V. Anode 175V, Grid 0V. 300B: cathode???, anode 360V, grid -70V
3. Voltage divider in front of volume pot. Looking at something similar to Goldpoint website. They have tables with VERY specific values based on pot impedance. Is there a reason for this? If I step outside of these values will there be impedance matching issue? These are not common values and I do not want a gaggle of resistors. Any recommendations for about a 12 dB decrease? Current pot is 50 Kohm. Source output is RCA at 2.0Vrms, 625 Ω
Thanks
Thanks for the updates!!
More “final thoughts“
I absolutely love this amplifier now.
When I first finished it I thought it was pretty good, but felt it did not live up to the hype...Not a keeper. While very enjoyable, the sound could best be described as slightly filtered, maybe constrained. I thought I would enjoy it for a while then move it on. No longer. After one more simple change, it is a keeper!
I originally installed Audio Note Copper Foil coupling caps in the fist stage and V-Caps CuTF in the second. Haven’t read much about both, I thought this would be a perfect blend.
The Audio Notes were described as smooth and rounded while V-Caps were called precise and airy.
On an expensive whim I decided to switch to all V-Caps. I removed the AN Caps in the first stage and replaced with V-Caps.
Absolute shock! The difference was profound! In all my years of experimenting with different cables, power conditioners and such I really never heard much difference. Maybe if I tried hard enough I could convince myself there was a subtle change. Even tube rolling changes can be subtle. This was not subtle in any way shape or form.
I don’t know that it was because the V-Caps were so much better, or that the AN caps were incompatible with my system. You may not need to spend as much as I did, but I would definitely not use AN caps in this amplifier. Keep in mind, the V caps sounded much better immediately, however there was an improvement with time.
Also, unless you’re running a very low output source, I highly, highly recommend the volume pot mod. I have the upgraded 50k ohm pot. I used 62K Ohm resistor inline with the signal and 24K Ohm to ground. See the Goldpoint audio website DIY for details. As well as a more functional volume knob, I honestly do think there was a subtle improvement in sound quality…maybe (but not even close to swapping out the capacitors).
This amplifier now lives up to the hype. It’s a keeper.
Final thoughts. This amplifier simply has too much gain. 0.270Vrms input is too sensitive for most sources. I suspect the manufacturer knows this. They have ditched the 12AX7 tube in the newer 3900 model giving it a much more reasonable 1 Vrms input which is compatible with typical 2 volt sources.
Modification complete! I’m calling it a “partial volume pot bypass“. Okay, not my idea. It’s a simple old school voltage divider mod I found on Goldpoint’s website. But the name is appropriate. You are literally bypassing a chunk of low quality potentiometer resistance with uber high quality z-foils. An improved signal path. Audible? Eh. Increase range of volume control. Definitely.
It’s easy to think in this in terms of smashing the signal down twice only to amplify it more or inserting more components into the path. As an absolute novice, that’s what I thought, so I shunned the idea. Truth is you can “smash” the source signal down entirely with a “plastic carbon composite“ potentiometer, or you can ” gently lower” it with with ultra high end vishay z foils and fine tune with the pot. The voltage allowed into the amplifier and amount of amplification is identical either way for a given listening level. As for adding components, yes you are adding a resistor, but this allows you to use less of the pots resistor. A wash at minimum. A win if higher quality resistor.
This mod is not necessary and of no benefit to most. Keep in mind my purpose for the mod. This amp has way too much gain for my source. My DAC has 2 Volts output. This is common of many modern sources—Some are higher. The input sensitivity of this Amplifier is 0.270 volts. The volume control was “zero to blast you out of the room” with with 2 degrees of rotation.
Here we go:
I first experimented with roon “headroom management” with various music to determine the appropriate dB reduction. Then I did some calculations using online voltage divider and parallel resistor calculators (based on Goldpoints modification dB tables) to figure out the voltage reduction vs dB reduction. Did some “breadboard” measurements using cheap resistors and alligator clips to volume pot. Looked the at results, read some tea leaves, spit in the wind, and decided on resistance values. Bought the vishay Z foils.
A look at the mod:
Top of volume pot module. On the right, Vishay Z foils added to left and right channel input pins (pins cut). The pot on left is for comparison and left stock as a spare.
Bottom of volume pot board. Vishay Z foils added from input to ground after the input resistors.
Reassembled. Time for a listen. Fingers crossed.
Nothing! No sound. Fu@k!!!
Whoops. Forgot to install jumper between pot board and main board. Easy fix. Installed. Try again.
Yes! Sound!!! But wait. Pretty soft. Turn it up. 12 o’clock. Better. But let’s try more. Oh no. Full volume for satisfactory levels. Damn. The the numbers were all good. I think. Did I eff up the math??? Too much reduction? Crap!
Light bulb moment. I still had 14 dB digital reduction set in roon!
Digital reduction removed…Perfection! The volume control is now fully functional from 0 to 2 o’’clock with “normal” levels around 10-11 o’clock.
Sounds sweet!!! Zero degradation and I swear there is a little extra sparkle to the high hats! Success
Give it a shot! Your logic makes sense to me.
To clarify my thought process:
Top Drawing WITHOUT voltage divider the signal travels further through pot “lower quality“ resistor vs lower drawing.
The 2 volt input signal sees the exact same resistance in both circuits. Just more “high quality“ resistance and less “low quality“ resistance for the exact same 1 Volt to the amplifier output in bottom circuit.
Not saying this will sound better, but, contrary to intuition, that no harm comes to the signal doing this. In fact, even if inaudible, an argument can be made, at a there is a benefit. IF your amplifier gain is too much for your source this will certainly give you increased “range” on the volume control.
(There are effects on the overall impedance seen by the source, but this can be negated, or even used to advantage by careful selection of resistor values vs dB reduction and source output impedance. That’s for another post.)
Thoughts?
Thanks for input.
No bypass cap. Tried subbing 12AU7 which helped a little with gain but sounded like crap.
For now, I’m using roon software to reduce gain which works but I still feel quality suffers. There is an entire debate about this on the Roon website. In theory it should not affect the quality, but I swear it does. At first, I was amongst the “it can’t possibly affect the quality” crowd, but the more I listen, I swear it does. Maybe it’s in my head. But I don’t like it.
I’m going to implement a voltage divider mod ”in front“ of the pot per Goldpoint website. Gain better matched to source would be ideal, however, there’s an argument to be made for high quality resistor voltage divider in front.
Here is my newbie understanding. Please let me know if I’m wrong.
For a given listening level, say 70dB from seat, you need a given input voltage to the amp. Let say it’s 1 Volt. If your source is providing 2 Volts that has to be reduced (divided), in this case, by half. That’s what the volume pot does. It’s a variable voltage divider.
At lower volume setting, the signal passes through the more of the pot’s “low quality” resistor. Inversely, at higher the volume settings the signal goes through less of the pot’s resistor.
Reducing the signal with a voltage divider made with high quality resistors (vishay Z foils?) allows you to use higher volume settings.
In effect, you are bypassing part of volume pots “low quality” resistor with higher quality resistors.
To be clear, I doubt putting the divider in front of the pot will cause an audible increase in quality, however it should NOT degrade the signal. The signal, in theory, should be slightly better. So there is no downside to this mod and a possible benefit.
Keep in mind the original premise. Too much gain for the output of my DAC. This would give me the increased range on the volume control I seek with no degradation and a slight, if inaudible, benefit.
Is there a Cathode bypass cap on the 12AX7? removing that will lower the gain. Maybe use a much lower gain tube like a 12AU7?