You might want to carefully compare the ac performance of the channel with the LED and RC based cathode bias since you are using output transformers. I've used LEDs with choke loaded triodes and with triodes driving transformers and have not had this problem.
Drive both channels in turn with square waves and using a scope make sure there isn't excessive ringing on the transformer outputs.
Check for not so obvious wiring errors, the effects you describe sound just like what I used to do for fun with a difference signal deliberately generated from a stereo pair.
Drive both channels in turn with square waves and using a scope make sure there isn't excessive ringing on the transformer outputs.
Check for not so obvious wiring errors, the effects you describe sound just like what I used to do for fun with a difference signal deliberately generated from a stereo pair.
One time I had the sound affect you are describing with a mic input transformer driving a LED biased anode follower. I chalked it up to a bad set of transformers as the problem went away when I took them out. I never thought to try a resistor/cap on the cathode instead of the LED.
Probably not a "break-in" issue at all, especially as the sound changes again when you reinstall the RC bias network.
The 0.3V difference could be shifting the operating point enough to change the sound.
Since we don't know what tube you're using, or what the primary Z is, or what the circuit looks like, or the OPT specs, we're all just speculating on very little information.
What circuit runs at 12mA into an OPT? Sounds like a headphone amp or linestage. Also sounds like an SE amp.
Based on the available scant information you have a SE headphone amp or line stage with sensitive high-Z OPTs. When you decrease the grid bias too much the idle current increases and your OPT is saturating on DC. That certainly would explain the echo-chamber characteristic sound.
How are you measuring the idle current in the circuit? Alternatively, if you can supply the details mentioned above, someone on here can make an estimate...
Cheers,
Michael
PS If it's something like a D3A in triode mode (matches your OP sort of) then the 0.3 V bias difference could cause an increase of 8mA in idle current. I suggest the next thing you do is try a schottky diode in series with the LED, or an LED that has the correct voltage drop for yoour circuit.
Last edited:
I draw a blank on that tube. Do you have a datasheet or plate curves?
What's the DC current spec of your OTs?
You can't derive current from the cathode-ground voltage for the diode... do you put an ammeter in series?
I recommend either trying to get the grid voltage to the right value or at least measure the DC current and see if you're exceeding the spec on your OTs.
Cheers,
Michael
Sorry. 5965 (typo). I don't have the DC of the transformers. The specs are alittle unusual. Bud Purvine did them from O-netics. Bias current is measured by me doing as stated and I have a excel template that I plug in the numbers taking in account other reading, resistance load, etc.
These OTs are built like tanks. I forget what Bud said but they can handle more than 30mA of current. It was a crazy number.
These OTs are built like tanks. I forget what Bud said but they can handle more than 30mA of current. It was a crazy number.
The plate current should only increase by about 2mA (0.3V * 6500mA/V) due to the LED, which should not account for such a profound change. Just for thoroughness, what is the plate current with the LED?
By the way, I have used LED and other diode bias a lot and it never impacts the sound as you describe.
The only way to know for sure is get a diode or series string with 2.4V drop at 11mA
I'd have to put the LED back in and measure again. When I mentioned breaking in the circuit I just went back to the 5965 so the sockets are new and all the solder points. I'll give it another try. I also bought some diodes to try. Which did you wind up preferring? LED or diodes?
Also did you notice a drop off in dynamics with a solo LED opposed to a R/C bias?
Also did you notice a drop off in dynamics with a solo LED opposed to a R/C bias?
It could though with a small gaped transformer?
I always buy several different brands of LEDs with the Vf I want near the current I want, that way you can usually get one that mates well with the circuit.
Smart as in harmless ? Yes.
By all means check out this post and the enclsoed results.
If you amplifier has rather high gain of 100 (say line in and 100+W into 8R load out), you'll be encountering somewhere in the vicinity of whopping 50 microvolts of amplified LED noise when using it for first stage bias if you omit that capacitor (and correspondingly less if used in later stages).
Can you hear 50 uV in presence of bunch of other noises usually present when playing music ? I can't.
By all means check out this post and the enclsoed results.
If you amplifier has rather high gain of 100 (say line in and 100+W into 8R load out), you'll be encountering somewhere in the vicinity of whopping 50 microvolts of amplified LED noise when using it for first stage bias if you omit that capacitor (and correspondingly less if used in later stages).
Can you hear 50 uV in presence of bunch of other noises usually present when playing music ? I can't.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.