It depends on what you meant by "it": an ideal CCS has an infinite output impedance - or it wouldn't be ideal.
Constant current shunt regulator
- Thread starter coinmaster
- Start date
At DC but what about AC?
That is a nice circuit, but unfortunately it is not digitally controllable. From reading that article it seems what I want is a high impedance at AC rather than low impedance at AC I guess. I think my fundamental understanding of the meaning behind the term output impedance is flawed causing confusion. Maybe I'm still confused and I actually want a low impedance.
In any case LTspice is showing a 0 ohm output impedance from my circuit
In any case LTspice is showing a 0 ohm output impedance from my circuit
Last edited:
Here's the AC impedance at 100ma which is the maximum I'm going to allow to the load
If a constant current source is supposed to have a high output impedance but this current source has a low output impedance but with seemingly stable current then isn't that a paradox?
If a constant current source is supposed to have a high output impedance but this current source has a low output impedance but with seemingly stable current then isn't that a paradox?
I think what you are after is a fixed quiescent current but then a large voltage swing for the signal.
Do you want to drive it with a d to a or several steps?
PS. Did you notice it uses your fancy FETs?
Isn't the point of a regulator to suppress things like AC voltage swings?
I have no idea what you mean
Yeah but no digital control is a killer. Besides I'm going to need a voltage regulator to switch between as well so I'm killing two birds with my CCS shunt reg.
Last edited:
Yes, normally, but in this case the AC represents your music. The DC just sets the load current for the tube.
You said you wanted digital control. That means it comes in steps.. How many steps?
I'm no tube guy, but if you read the article you can see the advantages. I'm pretty sure I couldn't hear it, but some people might.
You cannot have constant current and constant voltage in the same changing load.
Except during the notes on those transients the load line is effectively altered because the voltage is altered which alters the sound as well. The AC voltage should remain fixed. The current...I don't know. I don't think I understand what it means for an AC signal to meet a "infinite" or low impedance and how/why it matters.
Actually that reminds me, since AC signals are going to be suppressed then I can't use the regulator as a current source to the anode, only the cathode. In which case this only applies to a negative supply.
I need control of microvolts on the opamp input so I'm guessing a 16 bit DAC should do it perhaps.
Constant current source loads are very noticeable to hear. I have a gyrator design that performs even better than the CCS you linked though, which I will also have on my "switch".
I'm not looking for constant voltage. I'll have a separate feedback loop for the constant voltage version of the shunt regulator.
So far I'm looking at 4 dacs per power supply and I need 2 power supplies, I'll need to start looking at an MCU that can handle the dac control and device switching as well as software interface, and that's just in the power supply. The tube loading circuits are going to require another 10 or so dacs and a few adcs. I wonder if an arduino would work.
Last edited:
I think that is what it is all about. If you put a resistor in the plate as is normal it is not linear crossing the load lines. If the tube current is constant all the voltage swing is reproduced across the CCS.
If you put the shunt in you will need to add back the plate resistor. So it's like having a much higher power supply voltage.
Keep in mind I know nothing about tube amps.
I'm sure there is a way to interface a D to A to it.
I've heard of these.
Hmmmm, I don't think that applies for the cathode since a sag in cathode voltage would affect the bias since the grid is referenced against the cathode voltage but perhaps for the plate you are right. Although I need to have the ability to switch between constant current and constant voltage, meaning I'll need to switch in and out the bypass capacitor in that case, maybe via relay or something.
I'm still wondering what exactly you mean by this.
I guess it depends on where you take the signal from - plate or cathode. See examples here:
http://www.tubecad.com/2012/02/blog0225.htm
In the constant current plate circuit you said you wanted to digitally control it. That is what I mean by interfacing a D to A to it.
Bordodynov
- Joined May 20, 2015
- 3,181
I came up with a suitable electronic circuitry with digital control. It consists of an operational amplifier, a bipolar transistor, the transistor depletion mode, DAC, optocouplers, a timer, a simple transformer, a few diodes, resistors and capacitors. If you wait a two days I will send circuit and simulation file for LTspice.
Whoa whoa, I don't know what you mean by two days but I don't want you to spend a lot of effort on it, especially when I might not need it. What advantage does your design offer vs mine?
I was going to use an MCU to control a dac which controls the opamp reference voltage.
Or in other areas of the design I would use the MCU to control a dac which controls a common emitter amplifier for adjustable voltage up to B+/-.
I scanned through it but I couldn't figure out what "d" and "a" is.
Yeah that's what I figured. But I can't understand what he means by implementing a D/A to it, especially since I already explained my plan for D/A.
Let us know how it turns out!
I shall, the problem is the price tag on the capacitors at these voltages . Not to mention whatever it's going to cost to get a 12-1300v transformer made.
. Not to mention whatever it's going to cost to get a 12-1300v transformer made.
I said 1200 to 1300v
It's a bipolar supply so it's still going to be 600v. I might make it 650v per rail for headroom.I wish. Don't you need like a million volts to get a plasma speaker to do 20hz-20khz?
