Alastair E makes a good point. Some further info on this can be found here http://members.aol.com/sbench/humbal.html
If I assume you've put your schematic here for critique, I see you appear to have attempted direct coupling the first and second stages but you now have a large (electrolytic?) capacitor in the signal path unless I'm mistaken.
If I assume you've put your schematic here for critique, I see you appear to have attempted direct coupling the first and second stages but you now have a large (electrolytic?) capacitor in the signal path unless I'm mistaken.
This is just an idea,-as Ive never used the 300B....
How about having a small cathode-follower, say 6J5 or similar to directly couple the 300B to the previous stage....
That way, the coupling-cap from your previous stage, can be made smaller, without losing anything. With a direct-coupled cathode-follower to the 300b you'll need another supply, say, -200V, and can set the bias on the 300b via the grid of the driver to the 300B, and delete your large 300B cathode-resistor and by-pass cap altogether
Cathode-bias while nice/simple and 'safe' does in my opinion detract from the sound especially on more 'demanding' music Grid bias is somewhat more efficient than cathode-bias, but you'll need to reduce your +B to around 400 odd volts instead of your 500 odd.....
How about having a small cathode-follower, say 6J5 or similar to directly couple the 300B to the previous stage....
That way, the coupling-cap from your previous stage, can be made smaller, without losing anything. With a direct-coupled cathode-follower to the 300b you'll need another supply, say, -200V, and can set the bias on the 300b via the grid of the driver to the 300B, and delete your large 300B cathode-resistor and by-pass cap altogether
Cathode-bias while nice/simple and 'safe' does in my opinion detract from the sound especially on more 'demanding' music Grid bias is somewhat more efficient than cathode-bias, but you'll need to reduce your +B to around 400 odd volts instead of your 500 odd.....
A couple of quick thoughts..
As others have indicated using a pair of electrolytics on the filament of the 300B is going to result in unwanted filament current flowing through those caps. A single cap would be better in this instance.
I have built a lot of 300B based SE amplifiers and have found it impossible to get rid of all of the filamentary hum.. (I use high efficiency speakers.) I would recommend dc heating using current sources.
Note that some 300B use more filament current than others. Example is the JJ300B at 1.5A vs. WE300B at 1.2A - the JJ is non compliant with the original design spec, but mine sound good and have lasted a long, long time.
It looks to me like you are running around 100mA of idle current through your 300B, there are not too many reasonably priced current production tubes that will actually run at this current level with >400V effective Vp.. Your choices might be as limited as the WE and KR labs versions. I have popped Valve Arts (various versions) at as low as 70ma/400V and the cheapest Shuguangs at 60mA/400V - this is a few years ago now.
Fixed bias imo sounds better, probably due to eliminating a electrolytic capacitor in the signal path.
The electrolytic bypass across the second voltage amplifier cathode resistor if not chosen carefully will strongly color the sound. Consider a mu stage or ac coupled configuration in its place. I generally use an SRPP to drive 300B which is in turn driven by a simple rc coupled voltage amplifier stage. (6J5/6SN7 in that order) Another very good approach is to use an interstage transformer based driver.
On the driver stage you have elevated the filaments above ground which is a good thing, but you also need to provide a low impedance at ac so you need to add a cap to ground. (10uF should be more than enough here.)
edit: clarify some things, fix a couple of typos
As others have indicated using a pair of electrolytics on the filament of the 300B is going to result in unwanted filament current flowing through those caps. A single cap would be better in this instance.
I have built a lot of 300B based SE amplifiers and have found it impossible to get rid of all of the filamentary hum.. (I use high efficiency speakers.) I would recommend dc heating using current sources.
Note that some 300B use more filament current than others. Example is the JJ300B at 1.5A vs. WE300B at 1.2A - the JJ is non compliant with the original design spec, but mine sound good and have lasted a long, long time.
It looks to me like you are running around 100mA of idle current through your 300B, there are not too many reasonably priced current production tubes that will actually run at this current level with >400V effective Vp.. Your choices might be as limited as the WE and KR labs versions. I have popped Valve Arts (various versions) at as low as 70ma/400V and the cheapest Shuguangs at 60mA/400V - this is a few years ago now.
Fixed bias imo sounds better, probably due to eliminating a electrolytic capacitor in the signal path.
The electrolytic bypass across the second voltage amplifier cathode resistor if not chosen carefully will strongly color the sound. Consider a mu stage or ac coupled configuration in its place. I generally use an SRPP to drive 300B which is in turn driven by a simple rc coupled voltage amplifier stage. (6J5/6SN7 in that order) Another very good approach is to use an interstage transformer based driver.
On the driver stage you have elevated the filaments above ground which is a good thing, but you also need to provide a low impedance at ac so you need to add a cap to ground. (10uF should be more than enough here.)
edit: clarify some things, fix a couple of typos
widowmaker said:The filament supply for 300B is DC, but in other schematic, these two caps in this cause are parralel between ground and middle point of heating transformer...
Caps are silver-thantalum, russian military production -
K52-2 100uF
Sorry
Good to see you've gone for a DC supply, and it may be an idea to regulate this too....
Personally, I would still dump the cathode-bias, and go fixed bias,
as no matter how superduper expensive cap you have in the cathode circuit, the fact is, there's Still a cap there, which can charge under music transients and partially un-bias the stage, untill it discharges through the cathode-resistor.....
I'm with Alastair E on the fixed bias versus cathode bias with the 300B.
For the last word in transparency you really want fixed bias. It will also allow you to easily fine tune the operating point to the load and particular brand of 300B you are using.
IMO there is no comparison - I've done both and repaired lots of commercial SE amplifiers over the years, and the best ones always had fixed bias. If it's good enough for Kondo (sad, he's gone
) then it's good enough for mere mortals like me.
Edit:
Properly designed a fixed bias 300B SE amplifier may use up to 15% - 20% less energy than a comparable cathode biased design.
For the last word in transparency you really want fixed bias. It will also allow you to easily fine tune the operating point to the load and particular brand of 300B you are using.
IMO there is no comparison - I've done both and repaired lots of commercial SE amplifiers over the years, and the best ones always had fixed bias. If it's good enough for Kondo (sad, he's gone
) then it's good enough for mere mortals like me.Edit:
Properly designed a fixed bias 300B SE amplifier may use up to 15% - 20% less energy than a comparable cathode biased design.
hi kevin, i've seen a couple of schematics from ASL (Antique Sound Labs) whose amps are mostly fixed bias, but the fixed bias circuits are as simple as taken from a negative rail with some voltage divider and a cap. Do you think this is sufficient or you would suggest a better negative bias supply?
ps.
would you recommend, based on you experience, fixed bias on other triodes such as 2A3 or 45?
ps.
would you recommend, based on you experience, fixed bias on other triodes such as 2A3 or 45?
Hi Arnoldc,
This should be adequate with sufficient filtering. ASL oem'd amps for me back in the late 1990's and early 2000's and I had no problem with a simple bias setting scheme.
I recommend something like 10K 2W pot, 100K resistor to wiper on hot end, resistor from hot end to -V supply. Resistor from cold end of pot to ground. (common)
The bias supply should be well filtered, but not regulated unless the plate voltage is also regulated.
-Vbias--- "top" resistor--- pot---- "bottom" resistor--- gnd
Wiper to 300B grid resistor and bypass cap..
Depending on 300B you would want something like a 20V adjustment range of say -80 to -100V assuming 2mA current through the pot. Assuming a -120V bias supply then top resistor would be 10K, and bottom resistor would be about 40K with the pot in the middle. I recommend a 200V 0.1uF - 0.47uF film cap installed right at the wiper..
Different bias supply voltages would only require you to adjust that one top resistor value. Note that you need a minimum bias supply voltage of -100V to adequately cover the range of 300B I have run across.
This should be adequate with sufficient filtering. ASL oem'd amps for me back in the late 1990's and early 2000's and I had no problem with a simple bias setting scheme.
I recommend something like 10K 2W pot, 100K resistor to wiper on hot end, resistor from hot end to -V supply. Resistor from cold end of pot to ground. (common)
The bias supply should be well filtered, but not regulated unless the plate voltage is also regulated.
-Vbias--- "top" resistor--- pot---- "bottom" resistor--- gnd
Wiper to 300B grid resistor and bypass cap..
Depending on 300B you would want something like a 20V adjustment range of say -80 to -100V assuming 2mA current through the pot. Assuming a -120V bias supply then top resistor would be 10K, and bottom resistor would be about 40K with the pot in the middle. I recommend a 200V 0.1uF - 0.47uF film cap installed right at the wiper..
Different bias supply voltages would only require you to adjust that one top resistor value. Note that you need a minimum bias supply voltage of -100V to adequately cover the range of 300B I have run across.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.