direct coupling

[2a3maniac]

I think that the bypass capacitor simple limit or erase the negative f.b. in cathode resistor [ anode follower] . anyway it has effect increasing the total gain in this stage, [and not only this] but the capacitor is not in signal path. And the PS capacitor works like a store of energy, eliminate the 50 or 100 hz noise, diminish the total Z of PS , and much more , but and this capacitor is not in the signal path. :

[fdegrove]

Hi,

I think that the bypass capacitor simple limit or erase the negative f.b. in cathode resistor [ anode follower] .

In a way, yes...Perfect caps don't exist but I see your point.

but the capacitor is not in signal path.

But it is.
Read through your post and you may see why.

and much more , but and this capacitor is not in the signal path.

With SE amps the whole PS is in series whith the signal path...No free lunch here either...

Sorry to burst the bubble...

Cheers,

[runeight]

maniac, i'm with frank on this one.

think about it this way. in a simple grounded cathode amp with a bypass cap, if the cap is doing its job, any AC current flowing through the tube is flowing through the bypass cap not the cathode resistor. that's what it's supposed to do.

now this very same current is flowing through the load resistor and defining the variation in plate voltage in response to grid voltage. so, anything that the bypass cap adds to the signal is reflected directly in the plate voltage.

here's a simple way to see this. start with a large bypass cap and check the frequency response of the circuit. then progressively make the bypass cap smaller. if it had no effect on the signal, then the frequency response would stay the same. but it doesn't. as the cap gets smaller, low frequencies are progressively lost. how could this be if the cap were not in the signal path somehow?

PS caps do more or less the same thing. in a class A circuit, progressively make the last filter cap smaller and smaller. what happens to the frequency response of the circuit? same as above (approximately).

many folks on this forum swear by getting power supplies right to get the best sonics from a given circuit. although i am not quite as ardent as many of folks here who are better at this than i am, the PS must be having an effect on the "signal".

make sense?

[angel]

Of course, using a diode (e.g. LED) to drop the voltage would offer DC coupling at the expense of potential nonlinearities in the diode.

Those could then again be traded back in for a new cap

Seriously, though, the only way to minimize distortion, is to consider the current flow through the entire circuit, counting the mains, and operating without such "magic" problem solvers (carpets) as 'ground' etc...

I think Analog Devices has an application note on grounds that could be useful in explaining this.

[Gary P]

Yep, Frank's right on here. In regular SE circuits both the cathode bypass and final capicitor in the power supply are directly in the signal path.

The thing to keep in mind is AC signal paths are loops. I can't remember if it was Kirchoff or Norton that had the "all currents sum to zero" statement in their equations.

For more input, look at the link "CCS's and signal current control" on my webpage. It's the handout I put together for my presentation at the San Francisco Triode festival last June. It has good info on how to look at signal current loops.

Gary

Gary P's DIY page

[jan.didden]

I agree to a point with the above stuff, but there is a difference in the cathode decoupling cap effect on the signal and the ps caps. The signal current does pass through both, but observe: if the cathode decoupling cap is non-linear, that will directly modulate the gain, and therefore lead to distortion (which is nothing else then gain variation during the signal cycle).

If the ps cap is non-linear, this does not affect the gain, so there is not a comparable effect on distortion. It is true that in the end ANY non-linearity, including that of the ps caps, leads to distortion, but it should be clear that the effect is orders of magnitude less. So, it makes sense to go for the best decoupling cap you can afford, but not to spend a fortune on ps caps.

Jan Didden

[coffin]

Hi
Colin's Lab
There are some Direct coupled/ Transformer coupled schmetics here.
The website is writen in Chinese, a Taiwan website.

cheers

Coffin

[Antonio Tucci]

The best direct coupling schematic (in my opinion) is this (http://home.att.net/~chimeraone/axiom300bschematic.html) modified ... in a 2 stages model.
I believe that it is possible to eliminate also other components without substantially modify the projectual philosophy of the amplifier.

[ThorstenL]

Konnichiwa,

Originally posted by 2a3maniac
I think that the bypass capacitor simple limit or erase the negative f.b. in cathode resistor [ anode follower] .

Nope. It is there to place the cathode at reference (often mistakenly called "ground") level for AC only, while allowing a DC Voltage to be developed on the cathode that s different from the reference.

Originally posted by 2a3maniac
anyway it has effect increasing the total gain in this stage,

No, it merely allows the stage to function at maximal AC gain, it cannot increase stage gain over that with the cathode connected to reference and a fixed grid bias voltage.

Originally posted by 2a3maniac
[and not only this] but the capacitor is not in signal path.

Well, the current through the capacitor changes with signal and any voltage developed across the capacitor (including that from microphonics and nonlinearities) is amplified by the stage whose cathode is bypassed with a gain that is about the same as the signal gain. In my personal definition this clearly qualifies as "being in the signal path".

Originally posted by 2a3maniac
And the PS capacitor works like a store of energy, eliminate the 50 or 100 hz noise, diminish the total Z of PS , and much more ,

That is a view that is somewhat limietd.

Originally posted by 2a3maniac
but and this capacitor is not in the signal path. :

Well, the current through the capacitor changes with signal, in fact the Load current loop is closed through this capacitor for AC. Any voltage developed across the capacitor (including that from microphonics and nonlinearities) is impressed on the output signal of the stage. In my personal definition this clearly qualifies as "being in the signal path".

All that said, your point was? And what has this got to do with either direct coupling or the price of skunk in Amsterdam?

Sayonara

[Positron]

I think I understand what you mean, 2a3, correct me if I am wrong.

" I think that the bypass capacitor simple limit or erase the negative f.b. in cathode resistor [ anode follower] ."

If I understand you correctly, you are saying that without the bypass cap, the actual output signal will be less than with the cap installed. This is true. But, the actual gain of the tube stage is the same. Let me explain, hopefully not too simplistic.

The bypass cap is actually common to both the input and output "circuits". Without the cap, a portion of the output signal is "fed", out of phase (I use phase and not polarity as some phase shifting is produced at the frequency extremes) with the input. So it is a matter of the stage having full gain but with a portion of the output signal fed back into the input, out of phase. Thus the actual output signal will be less than IF the cap were installed.

If the capacitor were perfect, and infinitely large, the reference that Thorsten mentions would be exactly the same as Ground AC signal wise, but and except for the DC bias.

The problem is that this cap, or any part for that matter, used in a cathode circuit, affects both the input signal, since it is in the input signal path, and the output signal, as it is also in the output signal's path.

Thus great care needs to be taken with any part used in the cathode circuit.

I think alot of time we kinda short cut our thinking and don't get this involved, so I hope I correctly understood what you were trying to say 2a3.