Dunno if anyone's said it so far, but the “right answer” from an electrical/electronic engineering point of view is “attenuate the output”.
In your original postings, you mused that the 10× gain was too much, and 5× would have been better. As you can imagine, there is nothing in the world simpler than a 2-equal-resistors voltage divider. Nothing.
The reason why it is sound engineering is because (we learned early that) the contribution of noise in an overall circuit is most critical at the first gain stage. Whatever noise it introduces continues on thru the rest of the circuit. However … if its gain is reduced 50% by a 50:50 voltage divider … so is the output noise.
Just Saying,
GoatGuy ✓
In your original postings, you mused that the 10× gain was too much, and 5× would have been better. As you can imagine, there is nothing in the world simpler than a 2-equal-resistors voltage divider. Nothing.
The reason why it is sound engineering is because (we learned early that) the contribution of noise in an overall circuit is most critical at the first gain stage. Whatever noise it introduces continues on thru the rest of the circuit. However … if its gain is reduced 50% by a 50:50 voltage divider … so is the output noise.
Just Saying,
GoatGuy ✓
Before you make a nice oscillator with Jon's proposal, try it this way.
Mona
Thank you, Mona!
I ask though, why 100k/1M instead of say 3k3/33k? I figured the lower the resistors within reason, the less noise they would add?
As long as the source can drive them, and with no early bass roll off, the smaller the better.
The 3.3k/33k attenuator would require a 2.2 uF output capacitor in the source.
Last edited:
Originally Posted by kodabmx
why 100k/1M instead of say 3k3/33k? I figured the lower the resistors within reason,
the less noise they would add?
With the 3k3/33k proposal the input sees an impedance of 33/12+3,3=6k (gain without FB =~12x).
So "within reason" indeed.
Mona
why 100k/1M instead of say 3k3/33k? I figured the lower the resistors within reason,
the less noise they would add?
That's right.But the input impedance goes down too.
With the 3k3/33k proposal the input sees an impedance of 33/12+3,3=6k (gain without FB =~12x).
So "within reason" indeed.
Mona
This thread is kinda painful to watch, so I'm gonna jump in with the folding chair:
Worrying about noise in this circuit is kinda redundant since it's not a low-noise circuit in the first place. Just eyeballing it, the equivalent input noise resistance of this circuit is going to be at least 15k, and that's ignoring any power supply hum (the half-mu stage only has 6dB PSRR, remember). You can easily put a 100k volume pot at the input with only a minor noise penalty. It's also a reasonable input impedance for a tube circuit. Doing this brings a headroom advantage and will make the circuit more 'useful' since the output follower will actually be driving the load with it's low output impedance.
If you still have too much gain then you can add a cathode bypass capacitor to the upper input triode. However, this will also change the distortion in a non-predictable way. See my book for more on this.
Worrying about noise in this circuit is kinda redundant since it's not a low-noise circuit in the first place. Just eyeballing it, the equivalent input noise resistance of this circuit is going to be at least 15k, and that's ignoring any power supply hum (the half-mu stage only has 6dB PSRR, remember). You can easily put a 100k volume pot at the input with only a minor noise penalty. It's also a reasonable input impedance for a tube circuit. Doing this brings a headroom advantage and will make the circuit more 'useful' since the output follower will actually be driving the load with it's low output impedance.
If you still have too much gain then you can add a cathode bypass capacitor to the upper input triode. However, this will also change the distortion in a non-predictable way. See my book for more on this.
An externally hosted image should be here but it was not working when we last tested it.
Thanks, Merlin. Perhaps I have unrealistic expectations. Effectively, with this circuit, or another built from scratch, I'm trying to have a silent output at full volume. There is no hum already, just waterfall. It's not very loud, but I know my old solid state preamp will do it. You would think the system was turned off until you blow your ears off when you start the music (if you forgot it was full blast). My power amps have virtually no noise, ear to the speaker so I know the noise is coming either from the source, or the preamp.
Yesterday, I was experimenting with making an anode follower from a 6J1 strapped as a triode... It worked quite well. I even got a 10Hz square wave to stay level, and 10kHz square looks excellent, too.
Yesterday, I was experimenting with making an anode follower from a 6J1 strapped as a triode... It worked quite well. I even got a 10Hz square wave to stay level, and 10kHz square looks excellent, too.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.