|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
79.9.218.44
In Reply to: RE: Quest for DC coupled circlotron posted by 6AS7_6SN7 on January 09, 2015 at 12:09:29
At least in sim world.
The DC adjusting pots are R42, R26, etc. etc....
As expected the output spectrum signature does not change.
Even though 'the trick' has been taylored to circlotron it will likely work with all tube amplifiers (PP & SE) with trivial adaptations.
Integrated current sources like TL783 and zeners are apparently much more economic than high quality capacitors and DC coupling is perceived as a better thing by most audiophiles, so, why not?
ecc230
Follow Ups:
Why and how does it work, actually?
Other remarkable information:
http://www.pmillett.com/current_source.htm
ecc230
This component would suit fine....
http://ixapps.ixys.com/DataSheet/98704.pdf
ecc230
In the standard RC coupling (between the plate of the preceding stage and the grid of the next stage) you use an electrical component (the capacitor) that has infinite DC impedance and negligible impedance at audio frequencies (w.r.t. the grid resistor, which in turn has high impedance w.r.t. to the plate resistance).
A current sink is a device that has theoretical 'infinite' DC impedance and very high dynamic impedance (in the audio band) w.r.t. the plate resistance (and the feeding resistor).
When a current sink replaces the grid resistor, a fixed current (1 mA for example) is spilled from the plate load of the preceding stage (via a feeding resistor).
Provided that the sink current is lower than 1/5 of the cathode current of the preceding stage, the plate voltage swing is not significantly affected.
When a signal is applied, the plate voltage swing is then 'transmitted' to the grid of the next stage because both the tube grid AND the current sink (that are in parallel) have 'infinite' dynamic impedance.
In order to calculate the voltage across the feeding resistor and the current sink the math is the same as with any voltage divider.
You have however to carry out two separate calculations: one for DC and a second one for AC.
Once you have set the DC grid voltage to the target value (usually about 0V) by the proper feeding resistor you have only to check that the voltage swing across the current sink does not exceed the maximum value (40V for LM334Z may be too low and hence the 125V of TL 783 appear more suitable).
LTSPICE simply speeds up the calculus.
ecc230
for having the patience to explain and for introducing me to this novel idea. I will have to read more about it. So, the main reason to use capacitor-coupling is to avoid excessive DCV on the grid of the following tube. This method allows blocking DC without much affecting AC (signal) voltage. I am curious how it actually sounds. Does it sound as "good" as conventional direct-coupling compared to cap-coupling?
You are welcome.
The more I play with the sim toy, the more it looks attractive (even though I know that there are no free launches...).
This is the latest schematic with active plate load too.
Clearly the commercial availability of a two bucks 450V current sink/source like this completely changes the scenario, as before 2004 such a development was hardly conceiveable.
What is amazing IMHO is that in sim world you can vary HT voltages +/- 10% and the DC bias remains rock stable.
Luca
ecc230
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: