Hi there,
I came up with the idea using the IRFP9240 P-channel MOSFET as adjustable Shunt Regulator for my 300B SE amp. I came up with a prototype circuit and built it with a small PCB.
With use of LT Spice simulation, the minimum holding current required is only 1mA. I just need to use a 200K resistor connects to B+ and cathode to provide this holding current.
Bypass capacitor is not required.
After live testing for 2 days, I found how much I have missed from this 300B amp for years using RC self bias. The sound is incredible clean with details that I never heard of.
For details and schematic, please check my Facebook Group "vancouverdiyaudio".
Great to be a DIYer.
Johnny
I came up with the idea using the IRFP9240 P-channel MOSFET as adjustable Shunt Regulator for my 300B SE amp. I came up with a prototype circuit and built it with a small PCB.
With use of LT Spice simulation, the minimum holding current required is only 1mA. I just need to use a 200K resistor connects to B+ and cathode to provide this holding current.
Bypass capacitor is not required.
After live testing for 2 days, I found how much I have missed from this 300B amp for years using RC self bias. The sound is incredible clean with details that I never heard of.
For details and schematic, please check my Facebook Group "vancouverdiyaudio".
Great to be a DIYer.
Johnny
kmtang,
I do not go on Facebook.
I think you said you use a CCS and no bypass cap (in your Post # 21).
A CCS (Constant Current Sink) in series from a 300B Filament ("Cathode") Either to ground, Or to a negative voltage, the following effects will happen:
When a signal is sent to the 300B grid, the Filament ("Cathode") voltage will vary just like the grid voltage.
You will get Less 300B stage Gain; and you get Very High 300B plate impedance, rp.
Power output will be less
Damping factor will be low
Perhaps I mis-understood your circuit as you have described it.
I do not go on Facebook.
I think you said you use a CCS and no bypass cap (in your Post # 21).
A CCS (Constant Current Sink) in series from a 300B Filament ("Cathode") Either to ground, Or to a negative voltage, the following effects will happen:
When a signal is sent to the 300B grid, the Filament ("Cathode") voltage will vary just like the grid voltage.
You will get Less 300B stage Gain; and you get Very High 300B plate impedance, rp.
Power output will be less
Damping factor will be low
Perhaps I mis-understood your circuit as you have described it.
Hi 6A3sUMMER,
Yes, you have mis-understood my words. The circuit is not CCS. I thought about using variable CCS as cathode circuit for the 300B. It can be done but the CCS has very high AC impedance. The circuit would require a Bypass capacitor to provide the AC signal path for the 300B.
With use of shunt regulator, the AC impedance is close to Zero and no phase shift issue. No bypass capacitor is required is the very best thing.
I am still searching for IC which can stand 150Vdc with very minimal hold current which would be very nice.
Johnny
Yes, you have mis-understood my words. The circuit is not CCS. I thought about using variable CCS as cathode circuit for the 300B. It can be done but the CCS has very high AC impedance. The circuit would require a Bypass capacitor to provide the AC signal path for the 300B.
With use of shunt regulator, the AC impedance is close to Zero and no phase shift issue. No bypass capacitor is required is the very best thing.
I am still searching for IC which can stand 150Vdc with very minimal hold current which would be very nice.
Johnny
This is the schematic of the prototype shunt regulator circuit I used in my 45 and 300B power amp
Be careful, I have measured higher THD with Vitrous enamel WW resistsors as well as white box ones, compared to MF and MOF resistors. They never sound better if the measure worse.
kmtang,
For discussion clarity: I will use the words 'plate current' to mean the filament 'cathode current'.
They are exactly the same until you have grid current (Class A2).
For Newbies, I do not mean the current to heat the filament (1.25 Amps).
Regarding your circuit in Post # 24. If it causes the voltage across its terminals to be constant, even when the 300B plate current varies from minimum to maximum versus the music signal (or versus a test tone), then true, you do not need a bypass cap across that circuit.
In that case, the 300B bias will not vary.
But if the voltage across your circuit varies versus plate current, you Do require a bypass cap.
But there is another cause of 300B bias shift.
If you use RC coupling from the driver stage to the 300B, then the 300B bias voltage will shift if you draw grid current. The C in RC coupling will charge, and that causes the bias to shift, until eventually the charge is returned, and the voltage across the cap goes back to the quiescent state.
If you do draw grid current, then turn the volume down.
That fixes the bias shift problem.
If you have to draw grid current to get the volume you need, it is time to purchase a more powerful amplifier.
The 'Rule of Thumb' is to get an amplifier with double the power (3dB more power).
For discussion clarity: I will use the words 'plate current' to mean the filament 'cathode current'.
They are exactly the same until you have grid current (Class A2).
For Newbies, I do not mean the current to heat the filament (1.25 Amps).
Regarding your circuit in Post # 24. If it causes the voltage across its terminals to be constant, even when the 300B plate current varies from minimum to maximum versus the music signal (or versus a test tone), then true, you do not need a bypass cap across that circuit.
In that case, the 300B bias will not vary.
But if the voltage across your circuit varies versus plate current, you Do require a bypass cap.
But there is another cause of 300B bias shift.
If you use RC coupling from the driver stage to the 300B, then the 300B bias voltage will shift if you draw grid current. The C in RC coupling will charge, and that causes the bias to shift, until eventually the charge is returned, and the voltage across the cap goes back to the quiescent state.
If you do draw grid current, then turn the volume down.
That fixes the bias shift problem.
If you have to draw grid current to get the volume you need, it is time to purchase a more powerful amplifier.
The 'Rule of Thumb' is to get an amplifier with double the power (3dB more power).
Last edited:
Hi6A3sUMMER,
Thanks for your clarifications.
First of all, with the LTspice simulation and live test of the 300B amp with sine wave and music, both shows the cathode voltage is being hold constant. I added a 200K resistor between the B+ and 300B cathode in order to feed to the shunt regulator circuit in order to maintain the regulated voltage when the 300B being turned off to 0mA when it is driven to the extreme. I am rather confident that NO bypass capacitor is required for the circuit.
When the 300B is driven to maximum which will cause the grid current to flow. This will affect the bias of the 300B no matter it is with RC self bias or shunt regulator. Of course, the RC self bias would have less effect in such scenario. I am using Tannoy Edinburg speakers which has 95dB efficiency. I don't need full power from the 300B amp. Therefore, this is not an issue to me.
Anyway, the Shunt Regulator circuit is so wonderful that I enjoy the music from the 300B amp so much. Of course, the Fix Bias arrangement may have the same performance. However, it is rather difficult to modify the SE amp with self bias to fix bias. Shunt Regulator seems the easiest way to do.
Johnny
Thanks for your clarifications.
First of all, with the LTspice simulation and live test of the 300B amp with sine wave and music, both shows the cathode voltage is being hold constant. I added a 200K resistor between the B+ and 300B cathode in order to feed to the shunt regulator circuit in order to maintain the regulated voltage when the 300B being turned off to 0mA when it is driven to the extreme. I am rather confident that NO bypass capacitor is required for the circuit.
When the 300B is driven to maximum which will cause the grid current to flow. This will affect the bias of the 300B no matter it is with RC self bias or shunt regulator. Of course, the RC self bias would have less effect in such scenario. I am using Tannoy Edinburg speakers which has 95dB efficiency. I don't need full power from the 300B amp. Therefore, this is not an issue to me.
Anyway, the Shunt Regulator circuit is so wonderful that I enjoy the music from the 300B amp so much. Of course, the Fix Bias arrangement may have the same performance. However, it is rather difficult to modify the SE amp with self bias to fix bias. Shunt Regulator seems the easiest way to do.
Johnny
Sorry, I don't understand how it works well.
If the 300B produces more current (for example indefinite grid voltage, damaged grid leak resistor etc. ), the voltage on cathode resistor will growing.
The shunt tries to adjust it, the shunt FET start to conducting to stabilize voltage on cathode resistor.
If the growing of anode current of tube continues, the FET conducting more, stabilizing the voltage on cathode resistor, but current of tube continues to grow, because of current via cathode resistor AND current of FET flows through the tube.
The voltage on cathode resistor potentially constant, but dissipation of tube can growing even up to melting.
Last edited:
euro21,
Congratulations!
You are correct.
Now, for a few Generalizations that may prove helpful:
For this discussion, when I say 300B "cathode", I mean the mid voltage of the filament (filament with Either 5Vrms AC; Or 5VDC).
Un-protected circuits, versus 300B plate dissipation, in regards to certain types of circuit and tube failure modes:
1. Shunt constant voltage "cathode" bias
2. Zener "cathode" bias
3. Constant voltage fixed bias to the grid
Some kinds of circuit failures, and tube thermal run-away are not protected in these circuit topologies.
Protected 300B plate dissipation circuits:
1. Self Bias ("cathode" resistor to ground with a bypass capacitor across the resistor).
Some kinds of circuit failures, and mis-calculated circuits are not protected. (Example use a 50V rated bypass cap when the 300B tube you have runs at 65V bias everyday (watch out for the electrolytic to dry out, or watch out for excess capacitor leakage current, or even to become a short).
Those different circuits, like all circuits have lots of other tradeoffs, but I am limiting the statements above, so as Not to include the other tradeoffs (those other tradeoffs deserves, and doe have many many different threads).
I have seen a 'perforated' plate 300B go into thermal run-away (even with the quiescent plate dissipation at only 1/2 of the maximum rated value).
Yes, those are generalizations.
But there you have it, for whatever it is worth.
Congratulations!
You are correct.
Now, for a few Generalizations that may prove helpful:
For this discussion, when I say 300B "cathode", I mean the mid voltage of the filament (filament with Either 5Vrms AC; Or 5VDC).
Un-protected circuits, versus 300B plate dissipation, in regards to certain types of circuit and tube failure modes:
1. Shunt constant voltage "cathode" bias
2. Zener "cathode" bias
3. Constant voltage fixed bias to the grid
Some kinds of circuit failures, and tube thermal run-away are not protected in these circuit topologies.
Protected 300B plate dissipation circuits:
1. Self Bias ("cathode" resistor to ground with a bypass capacitor across the resistor).
Some kinds of circuit failures, and mis-calculated circuits are not protected. (Example use a 50V rated bypass cap when the 300B tube you have runs at 65V bias everyday (watch out for the electrolytic to dry out, or watch out for excess capacitor leakage current, or even to become a short).
Those different circuits, like all circuits have lots of other tradeoffs, but I am limiting the statements above, so as Not to include the other tradeoffs (those other tradeoffs deserves, and doe have many many different threads).
I have seen a 'perforated' plate 300B go into thermal run-away (even with the quiescent plate dissipation at only 1/2 of the maximum rated value).
Yes, those are generalizations.
But there you have it, for whatever it is worth.
Last edited:
I modified my 845 SE amp from RC self bias with this shunt regulator circuit today. I set the idle current at 75mA with 975Vdc B+.
I have a front panel 100mA meter to monitor the 845 current. Therefore, just in case there's anything happens, I can see it easily
The regulated voltage is 125Vdc.
Johnny
I have a front panel 100mA meter to monitor the 845 current. Therefore, just in case there's anything happens, I can see it easily
The regulated voltage is 125Vdc.
Johnny
kmtang,
The more I think about it, the more intrigued I am about your shunt regulator for biasing "cathodes" and cathodes.
Listen to the shunt regulated 845 circuit for about 1 week, then . . .
Please report back to us about any sound differences you hear for the shunt circuit versus the previous RC self bias circuit.
Thanks!
Have Fun!
The more I think about it, the more intrigued I am about your shunt regulator for biasing "cathodes" and cathodes.
Listen to the shunt regulated 845 circuit for about 1 week, then . . .
Please report back to us about any sound differences you hear for the shunt circuit versus the previous RC self bias circuit.
Thanks!
Have Fun!
Last edited:
Hello 6A3,
I modified my 45, 300B and now the 845 SE amps with the shunt regulator in the cathode circuit instead of RC.
The sound with shunt regulator at the cathode opens up the sound like lifting the curtain in front of the speaker. So lively, dynamic, clear and powerful.
I am now working to add the current limiting circuit built into the regulator circuit to safe guard the power tube.
Johnny
I modified my 45, 300B and now the 845 SE amps with the shunt regulator in the cathode circuit instead of RC.
The sound with shunt regulator at the cathode opens up the sound like lifting the curtain in front of the speaker. So lively, dynamic, clear and powerful.
I am now working to add the current limiting circuit built into the regulator circuit to safe guard the power tube.
Johnny
This is the schematic that I have tried using the LTspice. I added the current limiting feature and will build the board for testing later.
The current limiter may affect the signal peaks, unless the current limit is quite high or you add an RC filter to react to average current.
I am thinking setting the current limit at the maximum swing at full power. That means at least double of the idle current.
R6 is the current sensor, right?
Let's say you have 70mA Iq. 30 ohms x 0.070 = 2.1V, which will fully open Q2, reducing the output current. The shunt regulator is for the output stage cathode, right?
I experimented using the AZ431L shunt regulator IC in my 6SN7 line amp. Too bad that it causes so much pink noise which is totally unacceptable. However, the quality of the sound was so nice that made me to search for a better low noise shunt regulator.
I used the LTspice to work out a two transistor shunt regulator circuit. The dynamic impedance is approximately 8 ohm which should be good enough for amplification and driver stages.
I had this installed in the 6SN7 line amp. Wow..... noise level is low. The sound is so dynamic, detailed, and open.
I love shunt regulators!
I used the LTspice to work out a two transistor shunt regulator circuit. The dynamic impedance is approximately 8 ohm which should be good enough for amplification and driver stages.
I had this installed in the 6SN7 line amp. Wow..... noise level is low. The sound is so dynamic, detailed, and open.
I love shunt regulators!
Hi Jaclvarez,
Well, the circuit diagram is just for trial with the LTspice simulation. It is not specific for 300B or other tubes yet. It will be my project later on.
By the way, my buddy said having the shunt regulator for the SE amp sounds like between fixed and self bias one. Wondering if you have experience in SE amps with fixed bias???
Well, the circuit diagram is just for trial with the LTspice simulation. It is not specific for 300B or other tubes yet. It will be my project later on.
By the way, my buddy said having the shunt regulator for the SE amp sounds like between fixed and self bias one. Wondering if you have experience in SE amps with fixed bias???
Unfortunately I don't. My only SE prototype amp used self-bias, the usual resistor + cap cathode to ground.
My push-pull amp does use fixed bias. It sounds great, but again I have nothing else to compare it, well, other than my Onkyo SS amp.