If you guys are having luck with it than more power to you. Regarding the statement, I experimented with LED bias about a year ago and encountered a few things:
-Noise
-Prone to oscillation
-Non-linear frequency response
-Tubes biasing themselves in places far from where they were supposed to be on the load line.
Perhaps it works in some situations, as some of you guys seem to have luck with it. But those experiences were enough for me to stick with the resistor or resistor&cap for biasing.
My apologies to anyone who thinks I’m jacking this thread, but since LED bias is relevant to the topic:
If I recall correctly this was the circumstance:
Subminiature pre-amp based on the 6205. Grounded cathode stage direct coupled to a cathode follower stage both stages being 6205's.
LED bias was on the first stage(obviously). I wanted to be able to switch between Pentode and Triode-strapped without the plate voltage changing. A VGC of 2V with a 200V supply and a 20K plate load would give me an aprox 100Vplate operating point in both pentode and triode. Being concerned with cathode current changing VGC when switching from pentode to triode using a resistor, I found both a blue and a red LED that had a drop of aprox 2V @ 5mA (which was the expected idle current) and decided to give LED bias a try.
I tried each color LED in circuit, with it powered on, the voltage across the LED was about what I expected but problems arose. Also, tried stacking 1N4148’s in triode and pentode mode on two separate channels and experienced the problems mentioned earlier. Ended up just using a resistor W/ Cap. Switching between Triode and Pentode turned out to not change the biasing point significantly after all.
If I recall correctly this was the circumstance:
Subminiature pre-amp based on the 6205. Grounded cathode stage direct coupled to a cathode follower stage both stages being 6205's.
LED bias was on the first stage(obviously). I wanted to be able to switch between Pentode and Triode-strapped without the plate voltage changing. A VGC of 2V with a 200V supply and a 20K plate load would give me an aprox 100Vplate operating point in both pentode and triode. Being concerned with cathode current changing VGC when switching from pentode to triode using a resistor, I found both a blue and a red LED that had a drop of aprox 2V @ 5mA (which was the expected idle current) and decided to give LED bias a try.
I tried each color LED in circuit, with it powered on, the voltage across the LED was about what I expected but problems arose. Also, tried stacking 1N4148’s in triode and pentode mode on two separate channels and experienced the problems mentioned earlier. Ended up just using a resistor W/ Cap. Switching between Triode and Pentode turned out to not change the biasing point significantly after all.
A couple things jump out. A blue LED should have a Vf of something like 3.5V, a red should be 1.5-1.8V depending on the material. For a 2V drop, you'd want to use a cheap Radio Shack green LED.
Also, for a 100V/5mA operating point in triode, you'd want more like 1.5-1.7V Vgk, which is right in range of cheap red LEDs.
5mA is a good current- the cheap red and green LEDs I use show extremely low noise at that forward current.
Also, for a 100V/5mA operating point in triode, you'd want more like 1.5-1.7V Vgk, which is right in range of cheap red LEDs.
5mA is a good current- the cheap red and green LEDs I use show extremely low noise at that forward current.
sgerus said:
Still working on it. I have all the parts, I just need to find some time to solder it together ...
I am actually doing it as a parafeed/DRD with cheap Hammond chokes (156G) as the driver plate choke and Hammond 119DA's as the output transformers. I'll use a 10m45 based CCS as the output tube's load. Tubes are (triode strapped) 5879 driving a 12b4, and the driver will be LED biased. This will be for 32R headphones (Grados) only, as the 119DA's are 8:600, so with the 32R load they should have just high enough a primary (2400) for the 12b4's.
To answer your original question, I think a CCS can be used, but it limits you a bit. As I understand it, a choke can swing more than a CCS, so with a CCS you lose gain. If that doesn't matter, then it should be fine as you'll still get the DC coupling aspect.
-d
dsavitsk said:
Actually you don't loose gain, you loose voltage swing which you can compensate for provided you have sufficient supply headroom. The upshot is that the pk-pk driver swing will be reduced and if the supply voltage is insufficient then output stage power capability will be reduced.
The inductor allows the plate to swing far above the nominal supply voltage, in theory this can approach 2X Vsupply as the driver is driven into cut off.
(Fast and loose the inductor stores energy in its magnetic field which as plate current drops the voltage across the inductance rises in an "effort" to maintain the current.)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.