The main big flash, was around 1.00am, very early morning.
The latest one was around 9.00pm.
In the two years it was running though this never occured, irrespective of what time of day it was.
My thoughts are the PS has been damaged, the caps. What other reason could there be for a 14v increase on one channel, only, of the 6n1p plate? But that cap, if shorted, is after the main filter caps so not too sure why one could be affected and others not?
damaged resistors might give incorrect voltages, too. This might be worth starting a new thread on flashing rectifiers. are they JJ 5AR4?
Not JJ, Svetlana winged C 5U4G.
The same ones I've been using for 2yrs without issue.
The damage could have occured from the original catastrophic failure of the output tube that destroyed the bias resistor.
If advisable I/we can start a new thread?
The same ones I've been using for 2yrs without issue.
The damage could have occured from the original catastrophic failure of the output tube that destroyed the bias resistor.
If advisable I/we can start a new thread?
AKlein,
Correct me if I am wrong, please.
You have two 6N1P triode sections in your circuit:
The same B+ voltage goes to individual DC plate loads of 4.7k + 39k = 43.7k
You have Combination: Fixed Bias plus Self Bias: 1.4V to 1.5V IR LED, plus 180 Ohms.
One possibility:
The transconductance, Gm, of the 6N1P is 4,400 MicroMhos (4.4mA/Volt)
Two IR LEDs with 1.4V versus 1.5V is a 0.1V difference. At 4mA/Volt, that causes a difference of 0.44mA plate current.
0.44mA
Unless the IR LEDs have the same voltage drop, that will vary the bias voltage and the plate current, even if the two triodes are exactly matched.
The difference of the two plate voltages is 14V.
14V / 43.7k = 0.32 mA difference in plate current.
0.32mA
The plate currents are about 5mA.
0.32mA / 5 mA = 0.064
That is 6.4%.
I would not worry about a 6.4% difference in DC plate current for two triodes that are used in two channels of a single ended amplifier.
Example: Suppose the gain of the two channels also is 6.4% different, that would be a 0.54dB channel difference . . .
That is as good or better, than many things such as loudspeakers, phono cartridges, etc.
Additional food for thought:
The IR LED has very low dynamic impedance, good. No bypass cap needed across the LED.
But the 180 Ohm resistor is not bypassed.
The mu (u) of the 6N1P is 33.
The plate impedance, rp, of the 6N1P is 7500 Ohms; but with an unbypassed 180 Ohm cathode resistor the plate impedance is increased:
180 Ohms x 33 = 5940 Ohms.
7500 Ohms + 5940 Ohms = 13,440 Ohms, the new plate impedance, rp.
As has been stated before, the Svetlana 6N2P data sheet had an error, it switched the transconductance value, and the plate resistance value.
(even though the Svetlana lists Gm as 7500 uMhos, and rp as 4400 Ohms, that is wrong; those of us who have designed that tube into our amplifiers and run the tests, have proven the data sheet swapped the Gm and rp values).
Have fun listening, it sounds like you are ready to go!
Correct me if I am wrong, please.
You have two 6N1P triode sections in your circuit:
The same B+ voltage goes to individual DC plate loads of 4.7k + 39k = 43.7k
You have Combination: Fixed Bias plus Self Bias: 1.4V to 1.5V IR LED, plus 180 Ohms.
One possibility:
The transconductance, Gm, of the 6N1P is 4,400 MicroMhos (4.4mA/Volt)
Two IR LEDs with 1.4V versus 1.5V is a 0.1V difference. At 4mA/Volt, that causes a difference of 0.44mA plate current.
0.44mA
Unless the IR LEDs have the same voltage drop, that will vary the bias voltage and the plate current, even if the two triodes are exactly matched.
The difference of the two plate voltages is 14V.
14V / 43.7k = 0.32 mA difference in plate current.
0.32mA
The plate currents are about 5mA.
0.32mA / 5 mA = 0.064
That is 6.4%.
I would not worry about a 6.4% difference in DC plate current for two triodes that are used in two channels of a single ended amplifier.
Example: Suppose the gain of the two channels also is 6.4% different, that would be a 0.54dB channel difference . . .
That is as good or better, than many things such as loudspeakers, phono cartridges, etc.
Additional food for thought:
The IR LED has very low dynamic impedance, good. No bypass cap needed across the LED.
But the 180 Ohm resistor is not bypassed.
The mu (u) of the 6N1P is 33.
The plate impedance, rp, of the 6N1P is 7500 Ohms; but with an unbypassed 180 Ohm cathode resistor the plate impedance is increased:
180 Ohms x 33 = 5940 Ohms.
7500 Ohms + 5940 Ohms = 13,440 Ohms, the new plate impedance, rp.
As has been stated before, the Svetlana 6N2P data sheet had an error, it switched the transconductance value, and the plate resistance value.
(even though the Svetlana lists Gm as 7500 uMhos, and rp as 4400 Ohms, that is wrong; those of us who have designed that tube into our amplifiers and run the tests, have proven the data sheet swapped the Gm and rp values).
Have fun listening, it sounds like you are ready to go!
Last edited:
Thank you for the excellent, informative post!
There is the point though, that before the rectifier flashed and popped, the plates were exactly the same. New rectifier, plus new 6n1p, and the 14v difference arises (plus another flash and pop from the new rectifier after a few days of use).
Just have this issue of flashing/popping rectifier to resolve, which i believe, is to replace all my power supply caps which may have been damaged from the initial flash.
Then, I will be good to go!
Last edited: