Parts and eagerness abound,
But knowledge is lacking.
Sadly, I wait.
Gents,
I'm trying to teach an old dog (me) new tricks. But I'm trying to learn a bit of knowledge first before I embark on a path than may lead to danger.
To this end, I've been trying to grasp some of the Zen variations by reading various tutorials available on the net, and I thinks I've got a bit of a handle on Son of Zen (SOZ) and the variation 6 (ZV6).
But I've been looking at variation 7 (ZV7-R), where an active current source is used and I'm back in the dark ages again.
So, is there a good place to read about active current sources and why they work? I've looked for this on the site, and it generally assumes knowledge and is asking for improvements (or showing them).
Also, is it possible to inductively load ZV6 without resorting to an active current source? Would this also provide efficiency benefits along the lines of variation 7 (ZV7-T), without the added complications (for me)?
But knowledge is lacking.
Sadly, I wait.
Gents,
I'm trying to teach an old dog (me) new tricks. But I'm trying to learn a bit of knowledge first before I embark on a path than may lead to danger.
To this end, I've been trying to grasp some of the Zen variations by reading various tutorials available on the net, and I thinks I've got a bit of a handle on Son of Zen (SOZ) and the variation 6 (ZV6).
But I've been looking at variation 7 (ZV7-R), where an active current source is used and I'm back in the dark ages again.
So, is there a good place to read about active current sources and why they work? I've looked for this on the site, and it generally assumes knowledge and is asking for improvements (or showing them).
Also, is it possible to inductively load ZV6 without resorting to an active current source? Would this also provide efficiency benefits along the lines of variation 7 (ZV7-T), without the added complications (for me)?
Cloth Ears said:
The latest copy of AudioXpress has a great article on current
sources written by Walt Jung. Lesser examples can be found
in pieces at www.passdiy.com
You're familiar with the idea of a voltage regulator, right? An ideal voltage regulator would deliver any arbitrary amount of current from 0 amps to infinite amps while maintaining an absolutely steady voltage. That's why it's a voltage regulator--the voltage remains constant, no matter what, but the current varies.
Now...call the current source by its alias: current regulator. What's a current regulator do? It locks the current and lets the voltage vary. Suppose you want a steady 1 amp. If you give a 1A current source a 1 ohm load, it will develop exactly 1V across the load. Simple application of Ohm's Law: I*R=E...1A*1 Ohm=1V. But what happens if you give it a 2 Ohm load? It will do whatever it has to do to force 1A through the load. In this case, it will develop 2V of output. 1A*2 Ohms=2V. An ideal current source could develop 1kV across a 1k resistor, simply because you told it to deliver 1A, no matter what.
Naturally, real current regulators have limitations, just as voltage regulators do. They can only deliver current up to a certain voltage, after which they run out of steam.
It's all a matter of perspective. Think of a current regulator (aka current source) as an "upside down" voltage regulator and all will be clear.
Grey
Now...call the current source by its alias: current regulator. What's a current regulator do? It locks the current and lets the voltage vary. Suppose you want a steady 1 amp. If you give a 1A current source a 1 ohm load, it will develop exactly 1V across the load. Simple application of Ohm's Law: I*R=E...1A*1 Ohm=1V. But what happens if you give it a 2 Ohm load? It will do whatever it has to do to force 1A through the load. In this case, it will develop 2V of output. 1A*2 Ohms=2V. An ideal current source could develop 1kV across a 1k resistor, simply because you told it to deliver 1A, no matter what.
Naturally, real current regulators have limitations, just as voltage regulators do. They can only deliver current up to a certain voltage, after which they run out of steam.
It's all a matter of perspective. Think of a current regulator (aka current source) as an "upside down" voltage regulator and all will be clear.
Grey
Zen Mod said:
Zen Mod, I think that after 43 years of only building passive circuits (x-overs) that my intuitive understanding of what a FET or bp transistor does, or what an active circuit does, is just not what it should be. Take this bit from the original Zen article on the current source:
OK, I understand that 2 amps through the .33ohm R1 is a voltage drop of .66V - but why does that turn on the junction of the Q3 (is that just how the MPSA92 works)? And why does Q3 limit the voltage appearing at the gate of Q2 to 4V? It seems like a water-flow system, where the water is flowing through Q2, but when it reaches a certain cfm rate, another element (Q3 - like a tap, maybe) comes into play that ensures that only a certain amount goes through Q2.
The problem is that I'm still in kindergarden, and you guys (even dumbing it down for me) are still talking at high-school level.
Nelson Pass said:
Nelson, now I'm really embarrassed. I've read all the articles from A40 to ZV9 - some of which are still over my head, and some like SOZ which are almost completely understood by my little grey cells. "Lesser examples" indeed!
GRollins said:
Grey, thank you. I am familiar with the idea, and could follow your post. And, I can understand that the current source from Zen would deliver what it says it will. I think I need to look for some tutorial on the Internet (and there must be one out there) that will explain why these transisters do what they do. And how.
I'm pretty sure I could get some parts together and copy what someone else has already done. But I'm not really into "monkey see, monkey do" - so even if I do copy someone's work, I'd like to know why I'm doing it.
I'll come back to this when I know a bit more about what questions to ask. When? Well, hopefully before "..., the sun will cool to a white dwarf, and after that the universe will experience heat death."
Actually, pretty much every part has an analogous water flow system part. A mosfet is like a valve - with a "check valve" (backflow preventing valve) attached to one side of it. The mosfet gate is analagous to the valve handle, which turns your flow on and off.
Voltage is pressure, amperage is the water flow rate. The mathematical equations for the two systems are largely identical.
I found the DIY op amps article to be quite helpful in describing exactly how that current source works. For bipolar transistors, the .65 volts is a constant, it is the voltage gate to source required to turn the device on.
As long as you are willing to do a lot of reading, and ask questions, you will get this stuff no problem. But don't hesitate to build some of this stuff even though you don't understand everything that is going on -- you will probably learn a lot more if you build some simple circuits and look at what is going on with them--via a DC meter and an oscilloscope if you can get your hands on one. I can only learn so much by just reading.
JJ
Cloth Ears said:
Agree that you shouldn't just build it without knowing how does it work. You will not be able to do troubleshooting without a thorough understanding.
Let me try to answer your question:
As you may already know that a diode will start conduct when a forward voltage of 0.66 voltage is applied to it. The B-E junction of Q3 acts like a diode. When the voltage across B-E of Q3 gets to about 0.66V, then current start to flow through the B-E junction and the current is amplified and causes a current between C-E. Since the current at C of Q3 is supplied from two fixed resistors R11 and R12, if current passes through Q3, the voltage will drop. When the voltage drops, the gate voltage of Q2 will also drop, current passing through Q2 will drop too, so does the current passing through R1. This will cause Q3 to conduct less current in turn. So Q3 is keeping the DC current of Q2 to a constant level. The value of R1 will determine the current of Q2. The voltage across R1 is always 0.66V. The current of Q2 is equal to 0.66V / R1.
(The part no. are based on Zen-V4.pdf)
I would say that this forum is better than any book can supply. Feel free to list your questions here. (That will also benefit other people who are too shy to ask.)
Cloth Ears said:
It is completely OK to substitute the 2 large 8R resistors with an inductor. But that is something I'll avoid 'cause it is difficult to get (you need to pass high current through it without making it saturated).
BTW, I cannot see any current source for ZV6 and ZV7. May be you are referring to F1 which uses current sources instead of the two large 8R.
Re: Re: Zv6-t?
Thanks Roboman,
Going backwards here...
The current source I'm referring to is Q3 in ZV7. It works one way in figure 1 and then is biased differently in figure 8. As I'm still sitting and staring (refer Grey's post), I'm not exactly sure how this change works yet - but I'm staring as hard as I can.
I'm interested in the ZV7 inductor stuff (with or without the current source) as I happen to have four 625VA 40-0-40 240v transformers lying around. I was planning to cobble together another amplifer, but it's too complex and I'd rather be able to understand what I'm doing.
No-one can ever say I'm shy. I'm going to read Zen Mod's stuff (and then delete it) that he has kindly sent me in an attempt to be able to formulate more questions.
And do a lot of sitting and staring. Especially at your answer to my questions...
RoboMan said:
Thanks Roboman,
Going backwards here...
The current source I'm referring to is Q3 in ZV7. It works one way in figure 1 and then is biased differently in figure 8. As I'm still sitting and staring (refer Grey's post
), I'm not exactly sure how this change works yet - but I'm staring as hard as I can.I'm interested in the ZV7 inductor stuff (with or without the current source) as I happen to have four 625VA 40-0-40 240v transformers lying around. I was planning to cobble together another amplifer, but it's too complex and I'd rather be able to understand what I'm doing.
No-one can ever say I'm shy. I'm going to read Zen Mod's stuff (and then delete it
) that he has kindly sent me in an attempt to be able to formulate more questions.And do a lot of sitting and staring. Especially at your answer to my questions...
I'm short on time, but I'll start with Fig. 1 and see how far I can get before I turn into a pumpkin.
Okay, contemplate R5 and R6. The node between them is like the fulcrum for a see-saw. As the output goes up and down on each side, that center point remains the same.
Back up and look at the circuit as a simple differential for a moment. The input, whether single-ended or balanced (a differential with enough impedance under its rump [either resistance or a currrent source] can synthesize its own balanced output) will cause the signal at the Drains of Q1 and Q2 to go up and down. By definition, they go up and down in opposition. If Q1 goes up one volt, then Q2 goes down one volt. The sum of the two AC voltages always equals zero if you happen to add them together...which is exactly what R5 and R6 do. Given that those two resistors have the same value, they define a voltage mid-way between the two outputs.
Ideally speaking, R5 and R6 have nulled out the AC from the signal and all that's left is an idealized DC voltage that can be used as part of a voltage divider (with P3) to set the voltage present at Q3's Gate.
There's a DC feedback mechanism involved. If, due to temperature, the voltage at the outputs drifts upwards, this indicates low bias current. However, the rise also trickles down through the voltage divider to the Gate of Q3, which also rises. As that Gate goes more positive, Q3 conducts harder, which in turn brings the outputs back down. Think of the increased current going through the output load resistors, R8 and R9. More current means those two resistors have more voltage across them, which in turn pushes the Drains of Q1 and Q2 back down, thereby correcting the original bias drift.
I'm out of time, but if you look at the center tap on the inductance in Fig. 8 as being equivalent to the node between R5 and R6, I think you'll find that it's pretty similar in principle.
Grey
Okay, contemplate R5 and R6. The node between them is like the fulcrum for a see-saw. As the output goes up and down on each side, that center point remains the same.
Back up and look at the circuit as a simple differential for a moment. The input, whether single-ended or balanced (a differential with enough impedance under its rump [either resistance or a currrent source] can synthesize its own balanced output) will cause the signal at the Drains of Q1 and Q2 to go up and down. By definition, they go up and down in opposition. If Q1 goes up one volt, then Q2 goes down one volt. The sum of the two AC voltages always equals zero if you happen to add them together...which is exactly what R5 and R6 do. Given that those two resistors have the same value, they define a voltage mid-way between the two outputs.
Ideally speaking, R5 and R6 have nulled out the AC from the signal and all that's left is an idealized DC voltage that can be used as part of a voltage divider (with P3) to set the voltage present at Q3's Gate.
There's a DC feedback mechanism involved. If, due to temperature, the voltage at the outputs drifts upwards, this indicates low bias current. However, the rise also trickles down through the voltage divider to the Gate of Q3, which also rises. As that Gate goes more positive, Q3 conducts harder, which in turn brings the outputs back down. Think of the increased current going through the output load resistors, R8 and R9. More current means those two resistors have more voltage across them, which in turn pushes the Drains of Q1 and Q2 back down, thereby correcting the original bias drift.
I'm out of time, but if you look at the center tap on the inductance in Fig. 8 as being equivalent to the node between R5 and R6, I think you'll find that it's pretty similar in principle.
Grey
<Off Topic>
Q2 was a mad dog because naughty Papa gave it a strong injection. The mad dog was totally out of control, barking and biting all pedestrians. So, Papa provided a rein, Q3, one end of which was fastened around the dog’s neck and another end was tightened to a concrete column so that the dog’s activity zone was well controlled within the circle of the rein length radius. By the way, one day Papa saw the dog bleeding due to the stiff rein. You know Papa’s mercy? Papa pondered for a while and decided to make the rein flexible like a spring. And, finally, Papa made a flexible rein growing up to 50% longer in length depending on the dog’s active craziness. So called Aleph active current source . . . Warl, warl, warrrrrr . . . !
<End Off Topic>
Q2 was a mad dog because naughty Papa gave it a strong injection. The mad dog was totally out of control, barking and biting all pedestrians. So, Papa provided a rein, Q3, one end of which was fastened around the dog’s neck and another end was tightened to a concrete column so that the dog’s activity zone was well controlled within the circle of the rein length radius. By the way, one day Papa saw the dog bleeding due to the stiff rein. You know Papa’s mercy? Papa pondered for a while and decided to make the rein flexible like a spring. And, finally, Papa made a flexible rein growing up to 50% longer in length depending on the dog’s active craziness. So called Aleph active current source . . . Warl, warl, warrrrrr . . . !
<End Off Topic>
Re: Re: Re: Zv6-t?
I thought you were talking about ZV4 as I couldn't see any 0.33 ohm resistor and MPSA92 in ZV6 and ZV7. However my posting 10 did explain "Constant Current Source" as used in ZV4 and all Aleph series.
Grey explained the operation of Q3 in Fig 1 (ZV7) in detail. What I think need to be clarify are some terms.
"Active Current Source" is the use of an active component (together with some other components) to act like a resistor in terms of DC. However in terms of AC, it is a high impedance component. So you can adjust that active component and it is equivalent to changing a resistor.
"Constand Current Source" is an "Active Current Source" that keeps the current constant.
Q3 in Fig 8 of ZV7 is an "Active Current Source". Z1 provides a stabilised 9.1V from the +ve supply. By adjusting P3, you change the bias of Q3 and you also change the current that passes through Drain and Source. Q3 is basically acting as a resistor as far as DC is concerned.
For Fig 1, it is necessary to keep the Drain DC voltage of Q1 and Q2 steady. So the Gate of Q3 is connected to the Drain pins through R5 and R6.
Here is what Mr. Pass explained two para. below Fig 1:
"If the DC value goes up, then the voltage on Gate of Q3 goes postitive, it conducts more current and the outputs of Q1 and Q2 go back dow, and are thus held fairly constant."
This is not necessary for Fig 8 as DC resistance of the inductor should be very low and the voltages of the Drain pins of Q1 and Q2 are pretty close to the +ve supply voltage.
Hope this helps.
Cloth Ears said:
The current source I'm referring to is Q3 in ZV7. It works one way in figure 1 and then is biased differently in figure 8. As I'm still sitting and staring (refer Grey's post
I thought you were talking about ZV4 as I couldn't see any 0.33 ohm resistor and MPSA92 in ZV6 and ZV7. However my posting 10 did explain "Constant Current Source" as used in ZV4 and all Aleph series.
Grey explained the operation of Q3 in Fig 1 (ZV7) in detail. What I think need to be clarify are some terms.
"Active Current Source" is the use of an active component (together with some other components) to act like a resistor in terms of DC. However in terms of AC, it is a high impedance component. So you can adjust that active component and it is equivalent to changing a resistor.
"Constand Current Source" is an "Active Current Source" that keeps the current constant.
Q3 in Fig 8 of ZV7 is an "Active Current Source". Z1 provides a stabilised 9.1V from the +ve supply. By adjusting P3, you change the bias of Q3 and you also change the current that passes through Drain and Source. Q3 is basically acting as a resistor as far as DC is concerned.
For Fig 1, it is necessary to keep the Drain DC voltage of Q1 and Q2 steady. So the Gate of Q3 is connected to the Drain pins through R5 and R6.
Here is what Mr. Pass explained two para. below Fig 1:
"If the DC value goes up, then the voltage on Gate of Q3 goes postitive, it conducts more current and the outputs of Q1 and Q2 go back dow, and are thus held fairly constant."
This is not necessary for Fig 8 as DC resistance of the inductor should be very low and the voltages of the Drain pins of Q1 and Q2 are pretty close to the +ve supply voltage.
Hope this helps.
Hi
I don't want to complicate matters here, but, since we are talking about current sources...did anyone try to build the original ZEN amp with the BJT current source ? (a simple BJT current source, as described in DIY Opamps acticle should do)
I think somebody metioned that BJT CCS may sound better than a Mosfet CCS?
What does your experience tell ya?
Regards
Vix
I don't want to complicate matters here, but, since we are talking about current sources...did anyone try to build the original ZEN amp with the BJT current source ? (a simple BJT current source, as described in DIY Opamps acticle should do)
I think somebody metioned that BJT CCS may sound better than a Mosfet CCS?
What does your experience tell ya?
Regards
Vix
Choky--if you can get the audio express article Nelson mentioned--check it out! -- it is very interesting. The gist of it -- not all current sources are created equal. I think the j-fet current source measured something like 10 or 20 dB at noise rejection.
I was not too surprised to see a current source using a diode as a reference voltage measured better than an one with an LED to provide a reference voltage...
However, I am a firm believer in listening to a circuit, not just looking at a meter output.
JJ
jupiterjune said:
I agree completely with you-that's why I even mention noise in first place;
but- in this case it is not critical and you can't hear it.........as I always sez about Papa's approach - shooting at "good enough" solution, without diminishing returns.
besides - you can see that I didn't wait long to change CCS type on Babelfish input from Papa's original one. I mean original as in A30.
anyway-now I'll read article,tnx to Damp Ears
I'll read it slowly.......just because I have few things on my mind.......hehe
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.