|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.22.235.48
In Reply to: RE: BB Proof 2A3 DC posted by Garg0yle on December 05, 2014 at 20:22:08
Driver Stage
The driver stage has the most constraints placed on it. It must have a low enough source impedance and high enough current capability to drive the Miller capacitance of the 2A3 plus the plate choke self-capacitance; these alone limit the plate resistance to about 20Kohms. Being choke-loaded, the choke inductance must be high enough relative to the plate resistance to retain satisfactory bass extension as well. At 20K plate resistance, a -3dB point of 20Hz requires 159 henries. High-inductance plate chokes are few and far between, but Hammond does make one rated 150 henries at 8mA. And finally, the grid bias must be high enough to avoid driver grid current on signal peaks. The late John "Buddha" Camille recommended a bias of -2v plus the peak input signal (2.25v for this design, total 4.25v). In my own experiments I have not seen a grid current when the bias is at least 0.8v, so I believe 3.05v is safe. The design has about 3.4v bias. Any higher gain in the driver would produce a lower bias voltage, so we are limited to around a mu of 20.
The nominal pate resistance of the 6J5 is 8K ohms, so there is some margin. I elected to use this margin to obtain cathode bias without bypassing the resistor. Since this is a direct coupled amplifier, whose main purpose is to get rid of an audible coupling capacitor, that seems worthwhile. If we run the tube at high current, we can use a small cathode resistance to get the bias, and it will not add too much equivalent plate resistance. In this case, a resistance of 560 ohms gives 6mA current and adds about 11K ohms for a total of 19K ohms equivalent plate resistance. As a side benefit, the tube is very linear at this condition.
The remaining driver parameter is headroom – the plate voltage swing from zero bias to the quiescent point should exceed the required drive voltage. I have here chosen a 50% margin (70v swing vs. 45v required) as a compromise between linearity and keeping the total voltage low.
The actual voltages are adjusted by making the cathode resistance a potentiometer which can be trimmed as needed. This allows for drift of tube parameters with aging and power supply voltage variations. This pot should be wirewound or other bulk metal to ensure quietness, much like the hum pot on the 2A3.
Within the above constraints, it is possible to use other tubes – types 26, 27, 37, 56, 76, 6N7, 6N1P, 955, and 5670 come to mind. I do not offer detailed designs for any of these.
Notice that the driver plate current will drop some DC voltage across the plate choke, in this case 22 volts. That is subtracted from the 2A3 bias of 45 volts to obtain 23v drop across the upper 2A3 cathode resistor. 23v at 60mA (2A3 plate current) is 383 ohms; I chose 390 as very close. The lower resistor drops the remaining voltage, but carries less current (60mA – 6mA = 54mA) – calculate 2907 ohms, specified 3000 ohms. These calculations will have to be re-done if any different tube is used.
Follow Ups:
Belay that.
I just read the manual for a diy ccs board. The Hammond choke will do fine.
Big speakers and little amps blew my mind!
Only the choke has a low DC resistance, preventing the driver plate from exceeding the voltage at the choke positive attachment.
CCS does have sonic advantages but they bring out all the problems mentioned in the earlier thread.
PJ,
I looked at some better quality plate chokes, and I realized that the BH C4S would be cheaper. Are those kits supplied with any tables or equations, so they could be adapted to this amplifier?
I hope this question doesn't make you uncomfortable. I don't mean it to.
Jamie
Big speakers and little amps blew my mind!
Hi Paul,
I was reading the following from Gary P about using two of the inductors in series. See post #5 in the link.
The 22v drop of the second inductor could account for almost the whole drop of the 23v across the 390 ohm resistor. Perhaps the 390 resistor could be replaced by a 16 or 17 ohm resistor to measure the current.
ray
In the startup interval, as the 2A3 is heating but the 6J5 is still cold, the 390 ohm resistor provides a negative bias to the 2A3, limiting its current as well as the voltage across the cathode bypass capacitor.
Treading close to the spitball-proof amp? :^)
It took me a while but finally I can see why the cold 6J5 would be a problem.
For educational purposes, if I used the dual switch arrangement that turns on the heaters first, would the second inductor have been a good idea?
I realize that the dual switch would not be in the BB spirit of the amplifier design but I've used it a lot in the past.
ray
That would resolve this particular problem as long as the 6J5 is functioning. As an engineer I am always looking for ways to fail safe, such as when the 6J5 dies or is accidentally removed from the amp.
I'll just mention that there are commercially available plate chokes with more inductance and less DC resistance, which would be a better approach IMHO. But, you know, if you build something yourself, you get to make the decisions!
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: