I am pretty new to electronic design. Most of my experience has just been in troubleshooting and maintenance. I am trying to figure out the differences and benefits of CCS, CVS, and regulated B+, and have wound up with some questions that i could not get answered while searching.
Would a well regulated supply across a plate resistor not act like a CCS? I understand that as the resistor heats up, its resistance value changes, hence so does the current across it. But, wouldn't this eventually stabilize, or is the voltage swing across the grid force enough variation in current accross the plate resistor that it can not longer be considered constant?
How much difference in (audible) linearity would there be between a CCS and a well regulated B+?
Would a well regulated supply across a plate resistor not act like a CCS? I understand that as the resistor heats up, its resistance value changes, hence so does the current across it. But, wouldn't this eventually stabilize, or is the voltage swing across the grid force enough variation in current accross the plate resistor that it can not longer be considered constant?
How much difference in (audible) linearity would there be between a CCS and a well regulated B+?
Actually yes, if the value of the AC resistance to ground (the tube circuit's effective AC output resistance)
is much smaller than the plate resistor in value. Triodes work best (with the lowest distortion) in this way.
That is, with both a high supply voltage and a high plate resistor value.
Last edited:
I am working with a power supply regulated by a bootstrapped LM10 opamp controlling the bias of two FET devices ( which I guess is a quasi CCS in a way). This is straight out of the TI LM10 design book p.21.
I am getting 227 volts before the anode resistor. I am using a 220ohm plate load resistor. Currently the cathode is running unbypassed with a 3.3k resistor, but I am pondering trying LED bias with an anode bypass resistor. Since everything I have seen says it works best with a CCS, this is where I came up with my line of questioning.
I am getting 227 volts before the anode resistor. I am using a 220ohm plate load resistor. Currently the cathode is running unbypassed with a 3.3k resistor, but I am pondering trying LED bias with an anode bypass resistor. Since everything I have seen says it works best with a CCS, this is where I came up with my line of questioning.
Not sure if this is what you're asking but if you understand how to draw loadlines for your 12AX7, guys here can help answer your question better.
Read about loadlines here: How to "Screw Around" Your Tube Load Line
Basically, the answer to your question:
1. CCS = the loadline will be horizontal
2. CVS = the loadline will be vertical
3. Regulated B+ with Resistor Load = the loadline will be tilted.. the tilt degree depends on your resistor load value. Higher resistor load value will tilt more towards horizontal. Very high resistor load value will eventually become a CCS. Of course B+ will have to be adjusted to keep sufficient plate current flowing.
Read about loadlines here: How to "Screw Around" Your Tube Load Line
Basically, the answer to your question:
1. CCS = the loadline will be horizontal
2. CVS = the loadline will be vertical
3. Regulated B+ with Resistor Load = the loadline will be tilted.. the tilt degree depends on your resistor load value. Higher resistor load value will tilt more towards horizontal. Very high resistor load value will eventually become a CCS. Of course B+ will have to be adjusted to keep sufficient plate current flowing.
I did a bunch of research as I developed the design of a Hi-Fi tube amp I built (Two 6SN7s and two EL34s per channel), and ended up with a front end that consists of a 6SN7 with a CCS in place of the plate resistor. Everybody knows that using a CCS there improves linearity and increases gain somewhat, but what many apparently don't realize is that it also improves power supply rejection ratio a bunch (so significantly less hum caused by pwr supply) ( a good reason to use CCS by itself), and also if the tube ever shorts out internally, it limits current, so the burn out may well be less damaging to peripheral parts. My second stage is actually a follower with a CCS off the cathode to Gnd in place of that R. Single ended stages have very little Pwr supply rejection in their basic form. IXYS makes some CCS parts that are worth looking at.
Ballpencil: I have played with load lines quite a bit and have read that article. To me it seams that, at least for the 12??7 valve types, a regulated supply would be the most linear. This is unless you don't need very much gain, your signal doesn't require much voltage swing, you can use maximum voltage, and use low current that is completely lacking in variance.
I can see in the latter case why an LED for cathode bias could be preferred over a resistor and bypass cap. With low gain and low signal (like in a phono stage) there is a higher chance for noise from the PS and components (at least audible noise).
Bob: I can see the benefit of replacing the plate resistor entirely. My only question is, if you are able to do so and still maintain a high voltage. Most low noise devices, at least that I have seen mentioned in my searches, can not support high voltage. I guess this does for CCS's in general.
I can see in the latter case why an LED for cathode bias could be preferred over a resistor and bypass cap. With low gain and low signal (like in a phono stage) there is a higher chance for noise from the PS and components (at least audible noise).
Bob: I can see the benefit of replacing the plate resistor entirely. My only question is, if you are able to do so and still maintain a high voltage. Most low noise devices, at least that I have seen mentioned in my searches, can not support high voltage. I guess this does for CCS's in general.
I don't think you've completely understood the function of a plate load, and load lines.
You wrote you have 220 ohms on the plate side, and 3.3k on the cathode side. This would make your tube stage predominantly a cathode follower, if anything.
For a 12AX7 this would mean a practically vertical load line using the anode resistor.
You wrote you have 220 ohms on the plate side, and 3.3k on the cathode side. This would make your tube stage predominantly a cathode follower, if anything.
For a 12AX7 this would mean a practically vertical load line using the anode resistor.
A few observations from experience:
CCS loads on a 12AX7/ECC83 will (assuming the next stage doesn't load down the circuit) minimize distortion IF the cathode resistor is bypassed or you use LED bias. For minimum distortion, the next stage's input impedance ought to be 500k or greater.
CCS loads on a 12AX7/ECC83 are not a good idea for phono stages since the plate resistance is high and the current noise from the CCS will be translated into voltage noise. If this is a power amp, that's much less of an issue.
CCS loads will maximize gain which may or may not be a good thing, depending on the overall circuit. They will also maximize input capacitance because of Miller effect, which may or may not be a problem depending on the overall circuit.
Hope this is helpful.
CCS loads on a 12AX7/ECC83 will (assuming the next stage doesn't load down the circuit) minimize distortion IF the cathode resistor is bypassed or you use LED bias. For minimum distortion, the next stage's input impedance ought to be 500k or greater.
CCS loads on a 12AX7/ECC83 are not a good idea for phono stages since the plate resistance is high and the current noise from the CCS will be translated into voltage noise. If this is a power amp, that's much less of an issue.
CCS loads will maximize gain which may or may not be a good thing, depending on the overall circuit. They will also maximize input capacitance because of Miller effect, which may or may not be a problem depending on the overall circuit.
Hope this is helpful.
Sy: I am glad you chimed in. You are the reason I started thinking about this. Your statements about LED biasing on other threads put me on this path.
This is for an instrument preamp. It may be overkill, but why not.
The input capacitance may affect the loading of passive pickups, but my bass is active, so it should be ok.
This is for an instrument preamp. It may be overkill, but why not.
The input capacitance may affect the loading of passive pickups, but my bass is active, so it should be ok.
While this is very true, I have found that adding a CCS fed HV shunt reg. ahead of the CCS plate load will improve the sonics, sometimes in not so subtle ways.
You just need to allow for the signal voltage swing expected and the voltage required by the CCS. The CCS will let the tube's plate voltage be what's required. If the tube is operated "fixed bias" for example LED or resistor with a constant fixed current through it, then as emission falls of as the tube ages the plate voltage will rise so it's good to leave some leeway in the B+ overhead for that.
As this is a 12AX7 the current will be on the low side. Under 10mA a cascode Power MOSFET CCS, while very easy to implement, will not perform at its best. They do better with more than 20mA of current. I would consider using a BJT CCS like ones written up in Morgan Jones 4th edition.
I have read a few threads on here where people have had good experience operating the IXYS CCS's with 1-5mA and some below 1mA. I have considered possibly changing to 12AU7's, running a higher current, then I wouldn't have to run a bypass resistor for LED cathode bias. I just don't know if I would have enough gain (even though I am currently running an unbypassed cathode load).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.