Amplify early, attenuate late. So ideally you'd bring up the voltage fast, although without clipping the amplifying stage. Then at the output of the preamp you'd have a volume control, effectively attenuating signal and noise together. Without a buffer stage you'll end up with a high and volume-dependent output impedance though.
but let me introduce you to a law that I established, but not publish publicly:
My law of self bias:
in a self bias circuit, the dc voltage at the cathode represent the input's signal envelope that the ac waveform rides inside of it. The cathode circuit waveform will always represent the signal in the amplifying stage as "X". And "X" input signal's potential will always be one times measured at the cathode for an un-bypassed cathode resistor circuit. And X is the factor of 1 that the amplifier amplifies by its amplification factor.
Not so exactly. The cathode follower (This part of the circuit effectively behaves like cathode follower although the signal actually is taken from plate) never is equal to the grid circuit, because amplification is less than 1 Otherwise, there will nor be signal to the grid the tube can amplify.
More over, with irregular waveforms, the cathode departs very much from the grid signal, so high distortion can be encountered in those conditions.
the cathode signal represents 1 in the signal amplification. For a grid signal to be the same, the input impedance must be matched with no losses. So ideally, if you want no losses, then direct coupling would be preferred. So no reactive coupling parts such as transformers and caps would be out of the question if you are trying to achieve this. Grid stoppers are not considered a loss since they are used to trim the input impedance.
well in pro audio signal chains, gain staging by attenuation before and after the circuit is used on a fix gain circuit. The other method that is established is the external input pad, with an adjustable gain stage and output attenuation.
1 as 'X" is the measurement at the cathode, which at the plate is "X" times the amplification factor.
imput impedance is looking from the grid back to the input that returns to the signal's ground (A.K.A. AC common for people like me and wavebourne)
I've done what you describe in a preamp I built last year. It was directly heated, and just a little too noisy, so I upped the bias voltage to 8V and put a low impedance control at the output.
I will say that when you do this, you will be balancing loading and output impedance at the output of your preamp.
I will say that when you do this, you will be balancing loading and output impedance at the output of your preamp.
You put the volume control out front because you (the designer) never knows WHAT kind of source the user is gonna hook up.
And a rule from older hi-fi games is that if the sensitivity at the Vol pot wiper is 50mV-100mV, the hiss of this stage will not cause much complaint. (This may need re-reckoning with hot horns.)
If you can refute this (you know what user will hook-up, and who to complain-to if it distorts) then sure: boost up big and waste it down to taste. Not the cheapest way to go but in DIY hi-fi that is often not an issue.
An extreme case: school and office speaker systems often trim the main power amp to deliver 60-69V on a "70V" power speaker line, and use individual user-operated passive turn-downs at each speaker.
And a rule from older hi-fi games is that if the sensitivity at the Vol pot wiper is 50mV-100mV, the hiss of this stage will not cause much complaint. (This may need re-reckoning with hot horns.)
If you can refute this (you know what user will hook-up, and who to complain-to if it distorts) then sure: boost up big and waste it down to taste. Not the cheapest way to go but in DIY hi-fi that is often not an issue.
An extreme case: school and office speaker systems often trim the main power amp to deliver 60-69V on a "70V" power speaker line, and use individual user-operated passive turn-downs at each speaker.
DavesNotHere said:
You appear to be trying to say something about cathode bias and maximum signal amplitude, but the sense of it got lost somewhere between you typing it and me reading it.
Why not just say that the cathode bias must exceed the signal peak amplitude by a big enough margin to ensure no grid current clipping, while also avoiding grid cutoff at the opposite peak?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.