As the title says, is there any amplifier that uses a mu-stage on output?
I mean, I'm prototyping this SE amplifier; if I connect the output transformer (through C3) over the CCS, instead of below it, the parafeed output stage becomes a mu-stage.
This should ensure your lower tube works in every condition at a fixed current (in my case 100mA).
Running the lower tube in this way its linearity will be maximum.
Why nobody uses it? where's my error?
Ciao,
Giovanni
I mean, I'm prototyping this SE amplifier; if I connect the output transformer (through C3) over the CCS, instead of below it, the parafeed output stage becomes a mu-stage.
This should ensure your lower tube works in every condition at a fixed current (in my case 100mA).
Running the lower tube in this way its linearity will be maximum.
Why nobody uses it? where's my error?
Ciao,
Giovanni
No, if you connect the output transformer to the anode of the lower tube, the primary impedance will be in parallel to the load (upper pentode), this is not a µ-follower.
Andreas
Andreas
Sorry, I've been not clear enough. What I mean is to connect the output transformer to the upper pentode cathode, so above the solid-state CCS.
The SS CCS shown in the schematic is used as a "servo-CCS" to drive the upper pentode; it changes the bias voltage in manner to keep the current through itself constant.
If you connect the output at the lower tube anode the current flowing into the upper tube is constant, because it is exactly the same flowing into the CCS; but, if you connect the output to the upper pentode cathode, the current through the CCS will be again constant, but not anymore the one flowing into the pentode, cause not you have, at its pentode, the node woth the output.
In this way the upper pentode will not work anymore as a CCS, but will change its current to keep the lower tube's current constant.
It is not a real mu-stage, but something similar, isn't it?
This is the SS CCS schematic; it worked fine on my bench in a different topology (parafeed), I slightly modified it adding R7, R8 and R9 in manner to remove the CCS current from Q2 (Q2 was previously placed in place of R7; this would give you a variable voltage between the cathode and the output trafo, which I don't like).
Ciao,
Giovanni
The SS CCS shown in the schematic is used as a "servo-CCS" to drive the upper pentode; it changes the bias voltage in manner to keep the current through itself constant.
If you connect the output at the lower tube anode the current flowing into the upper tube is constant, because it is exactly the same flowing into the CCS; but, if you connect the output to the upper pentode cathode, the current through the CCS will be again constant, but not anymore the one flowing into the pentode, cause not you have, at its pentode, the node woth the output.
In this way the upper pentode will not work anymore as a CCS, but will change its current to keep the lower tube's current constant.
It is not a real mu-stage, but something similar, isn't it?
This is the SS CCS schematic; it worked fine on my bench in a different topology (parafeed), I slightly modified it adding R7, R8 and R9 in manner to remove the CCS current from Q2 (Q2 was previously placed in place of R7; this would give you a variable voltage between the cathode and the output trafo, which I don't like).
Ciao,
Giovanni
The CCS holds the current constant, so the bias for the upper pentod also will be constant...
Yes, but for output power you need chances in voltage and current, why would you hold the current constant?
Andreas
Well, how to explain in English, for me is quite difficult...
This is the revised schematic (Actually the upper triode is connected as pentode, not strapped as triode).
Premises:
1. you know that an ideal load for a triode is an infinite impedance, and in the real world the best load for it is a well-done CCS.
2. In the parafeed topology the output tube current is NOT constant, it varies with the AC signal (through the load).
Actually in the parafeed topology the output tube works exactly as in a SE stage, with the only difference that the load is DC insulated from the triode.
So, I'm trying to develop a different topology.
If I place the CCS under the load, the lower tube become a pure voltage amplifier, cause its current is absolutely constant: this means maximum linearity.
In idle (no signal) condition the pentode current is exactly the same than the lower triode: if there is an AC signal, the pentode current will vary in manner the |lower triode|+|load| current sum is always the CCS current.
An example:
Let's say the CCS current is 100mA.
If we have a positive signal on output triode grid, its plate voltage drops down, so we have a current through the output capacitor from the transformer down to the triode (let's say, 50mA).
The current in the pentode will be 100mA (CCS) - 50 mA (load) = 50mA.
If we have a negative signal on input, the triode voltage will raise, so we'll hae to load the output capacitor with a current from the upper pentode to the load (50mA).
In this case the pentode current will be 100mA + 50mA = 150mA.
The upper pentode, in other words, is a "current modulator", used to keep the triode current constant and to feed, when necessary, the load with the needed current.
Hope it is clear.
Ciao,
Giovanni
This is the revised schematic (Actually the upper triode is connected as pentode, not strapped as triode).
Premises:
1. you know that an ideal load for a triode is an infinite impedance, and in the real world the best load for it is a well-done CCS.
2. In the parafeed topology the output tube current is NOT constant, it varies with the AC signal (through the load).
Actually in the parafeed topology the output tube works exactly as in a SE stage, with the only difference that the load is DC insulated from the triode.
So, I'm trying to develop a different topology.
If I place the CCS under the load, the lower tube become a pure voltage amplifier, cause its current is absolutely constant: this means maximum linearity.
In idle (no signal) condition the pentode current is exactly the same than the lower triode: if there is an AC signal, the pentode current will vary in manner the |lower triode|+|load| current sum is always the CCS current.
An example:
Let's say the CCS current is 100mA.
If we have a positive signal on output triode grid, its plate voltage drops down, so we have a current through the output capacitor from the transformer down to the triode (let's say, 50mA).
The current in the pentode will be 100mA (CCS) - 50 mA (load) = 50mA.
If we have a negative signal on input, the triode voltage will raise, so we'll hae to load the output capacitor with a current from the upper pentode to the load (50mA).
In this case the pentode current will be 100mA + 50mA = 150mA.
The upper pentode, in other words, is a "current modulator", used to keep the triode current constant and to feed, when necessary, the load with the needed current.
Hope it is clear.
Ciao,
Giovanni
croccodillo said:
Hi Giovanni,
I understand, but this is not a µ-follower, because the load for the lower triode only for DC is high; for AC the triode "is seeing" the resistance of the upper pentod in parallel to the output load.
Looks like a push-pull-topology.
Regards Andreas
The output transformer and speaker load is variable over the frequency range. Your CCS is very sensitive to load and by attempting to use the Mu stage you are likely to introduce more distortion into the output. The triode is better able to control the load of the output transformer.
Mu stages work better into a fixed load.
Thats my understanding.
Shoog
Mu stages work better into a fixed load.
Thats my understanding.
Shoog
Well, I called this stage as mu-stage for simplicity, but it isn't actually.
It is a sort of "Active Current Source".
My inspiration has been the Nelson Pass Aleph series, where a solution like this one is used with mosfets.
Trust me, if you are so lucky to listen to an Aleph amplifier, you will never forget it.
You can see what I mean on his sites, this and this .
Nelson Pass has also a patent on such a topology, US 5'710'522.
Actually the upper active current source develop about half of the load current, not its totality, but it should work also with my topology.
So, my thought is: why not to try the same with tubes? a Nelson Pass "clone" made with tubes?
Ciao,
Giovanni
It is a sort of "Active Current Source".
My inspiration has been the Nelson Pass Aleph series, where a solution like this one is used with mosfets.
Trust me, if you are so lucky to listen to an Aleph amplifier, you will never forget it.
You can see what I mean on his sites, this and this .
Nelson Pass has also a patent on such a topology, US 5'710'522.
Actually the upper active current source develop about half of the load current, not its totality, but it should work also with my topology.
So, my thought is: why not to try the same with tubes? a Nelson Pass "clone" made with tubes?
Ciao,
Giovanni
"So, my thought is: why not to try the same with tubes? a Nelson Pass "clone" made with tubes?"
My first amp was the Zen Vol 3 with the Aleph current source. It uses a load sensing resistor to test the output demand and then feeds that to the current source which varies its current to deliver some of the current to the load. This may be your intention, but your initial diagram doesn't seem to have the required components to implement it in a tube version.
The Aleph current source is a very elegant and sophisticated design, but the bass was never as good as my 807 constant current source loaded parafeed output stage - with the load taken from the anode of the 807. I would be confident that the proper Alephs would sound better though.
The bass with an Anode chokes don't sound as good as a good current source.
Shoog
My first amp was the Zen Vol 3 with the Aleph current source. It uses a load sensing resistor to test the output demand and then feeds that to the current source which varies its current to deliver some of the current to the load. This may be your intention, but your initial diagram doesn't seem to have the required components to implement it in a tube version.
The Aleph current source is a very elegant and sophisticated design, but the bass was never as good as my 807 constant current source loaded parafeed output stage - with the load taken from the anode of the 807. I would be confident that the proper Alephs would sound better though.
The bass with an Anode chokes don't sound as good as a good current source.
Shoog
Really; my last working amplifier was an Aleph 30 clone built by me, now sold to a friend.
The sound was amazing (it took months only to match all the mosfets), and thus I would like to realize something similar with tubes.
I know exactly the Aleph topology, I read and re-read the US patent and had some e-mail contact with nelson Pass.
My first circuit was simply a CCS moved down, under the load: I'm now trying to adapt the Pass circuit to the tube.
I will post my schematic as soon as it is ready.
Ciao,
Giovanni
- Status
- Not open for further replies.