You'll get negligible improvement in performance so there's little point doing that. Nevertheless, if you did somehow have 1mA in the secondary it likely wouldn't cause a problem anyway since the secondary has few turns, but you could put a blocking cap in the feedback path.Case 2 I pump additional external current through the LED and resistor for better LED performance.
Last edited:
Cathode resistor in the tube circuit does provide a little local immediately (not loop) corrective feedback, albeit at a loss in gain. The LED does not allow much in the way of this kind of feedback since it fixes the cathode voltage. If you need all the gain you can get, then the diode is the way to go.
Hi, I'm just testing now the bias led on an AR SP3/6 clone, on the V2 (ecc83) gain tube, before the CF.
So, in your opinion, for the "best sound" a resistor is always better than a pair of 1n4148? And diode (like 1n4148) is better of Led?
The slightly less gain of the resistor doesn't matter to me.
I also think I have read the option of putting an R in series with the diodes (or with the LEDs), in my case a pair of 1n4148.
Did I read that right?
In doing so, you have the advantage of the diodes (or LEDs) and also of the resistor?
What appropriate value should the additional resistor have? Always if I understand correctly!
PS: In the photo an ir led with added current (4.5mA) from b +. Total current about 5.1 / 5.2mA. But I think they sound better 2x 1n4148!

Thanks
So, in your opinion, for the "best sound" a resistor is always better than a pair of 1n4148? And diode (like 1n4148) is better of Led?
Did I read that right?
In doing so, you have the advantage of the diodes (or LEDs) and also of the resistor?
What appropriate value should the additional resistor have? Always if I understand correctly!
PS: In the photo an ir led with added current (4.5mA) from b +. Total current about 5.1 / 5.2mA. But I think they sound better 2x 1n4148!
https://ibb.co/vqxGjZq
Thanks
interesting...... well to translate between a resistor and a lamp, you only need to know how much the normal voltage drop is across the resistor ; and line that up with the normal [voltage-forward] of the lamp or other diodes...
now, as far as what combo sounds better, just use what ever sounds best to your ear.........
..you are correct what you say, the method. but I thought I read that a bit of local feedback given by a resistor in series with a LED (or several diodes) .. is better circuitally, that is, the final result certainly benefits.
Any comments on this?Is there a good argument against using 2 or 3 Si diodes in series instead of the LED? 1N4148 x2 and 100k plate load looks better on even harmonics than LED.
I can't argue because I have had very good results with 1 or 2 1N4148 with 12AX7A and D3A.. Depending on the bias needed I have combined conventional silicon diodes with schottky diodes. Infrared leds with a Vf around 1.2V may also be of interest.
I've been trying C3D02060F SIC diodes, for example 2 with a 53R resistor in series. I can't really live with them, useful as they are, since I hear a slight edge which grates on me. Orchestral strings are grainy and voices a little edgy. However slight the edge, and I think adding a resistor helps, it's not as good as an unbypassed resistor so that remains the ideal. Haven't tried red LED or 1N4148.
- Home
- Amplifiers
- Tubes / Valves
- Cheap Red LED bias on ecc83/12ax7