Hi folks,
I am restoring a vintage Criterion MkVII receiver which I saved from the land fields. With this post, I have two goals:
1. Get help with setting the correct bias in the power amp section
2. Get help with understanding the power amp circuit.
Little bit about the Criterion…It was the last receiver developed and sold by Lafayette. As a matter of fact, the company was closing doors when it came on the market. There is no information available other than the attached schematic that came weirdly, with the owner's manual. No budget for service manual or service afterwards. Personally, I think that it is a very clean sounding receiver and one of the prettiest of the era. It is very simple and elegant looking one.
With regards the first question above, I need to setup the DC bias, but I have no idea to what value and where to measure it. Normally, for class AB amp it is measured across either or both R21 and R23. Am I correct and if so, how do I calculate the proper bias voltage?
With regard understanding the circuit…here is what I think I know about it and a few questions at the end:
We are going to look at the left channel only.
Signal travels though C1 and is applied at the base of the left transistor of Q1 (Q1 is a supermatched pair of transistors in one enclosure) of the longtail pair comprised of Q1 and Q3 which forms the IPS. Overall negative feedback/closed loop gain is provided at the base of the right transistor of Q1 through the R11/C7.
The signal is taken from the collector of the left transistor of Q1 and applied to the base of Q5 which forms the VAS. after that one half of the wave (lets assume sinusoidal signal for simplicity) is applied to the base of Q11 and the other half through C11 is applied at the base of Q9. Afterwards it is easy, one half goes through Q11, Q15 and Q19, while the other goes through Q9, Q13 and Q17. Q7 is a voltage source and along with the STV4H's diodes and R17 form the bias spreader of the amplifier.
Here are the questions:
1. Is my general understanding of the circuit correct?
2. How come that the feedback is applied to both: the base of the right T of Q1 and Q5? I would expect that it is applied only to right T of Q1.
3. What is the function of C9 and C31?
4. Can someone explain the dynamics between IPS, VAS and the feedback gain loop?
Thanks in advance.
Best
Soundork
P.S. Here is a little interesting fact as a reward for reading this far: The only thing wrong with it is a bulging capacitor in the PS. The rail voltage is 50V and they have a cap rated at 50V max. Wicked weird…😊
I am restoring a vintage Criterion MkVII receiver which I saved from the land fields. With this post, I have two goals:
1. Get help with setting the correct bias in the power amp section
2. Get help with understanding the power amp circuit.
Little bit about the Criterion…It was the last receiver developed and sold by Lafayette. As a matter of fact, the company was closing doors when it came on the market. There is no information available other than the attached schematic that came weirdly, with the owner's manual. No budget for service manual or service afterwards. Personally, I think that it is a very clean sounding receiver and one of the prettiest of the era. It is very simple and elegant looking one.
With regards the first question above, I need to setup the DC bias, but I have no idea to what value and where to measure it. Normally, for class AB amp it is measured across either or both R21 and R23. Am I correct and if so, how do I calculate the proper bias voltage?
With regard understanding the circuit…here is what I think I know about it and a few questions at the end:
We are going to look at the left channel only.
Signal travels though C1 and is applied at the base of the left transistor of Q1 (Q1 is a supermatched pair of transistors in one enclosure) of the longtail pair comprised of Q1 and Q3 which forms the IPS. Overall negative feedback/closed loop gain is provided at the base of the right transistor of Q1 through the R11/C7.
The signal is taken from the collector of the left transistor of Q1 and applied to the base of Q5 which forms the VAS. after that one half of the wave (lets assume sinusoidal signal for simplicity) is applied to the base of Q11 and the other half through C11 is applied at the base of Q9. Afterwards it is easy, one half goes through Q11, Q15 and Q19, while the other goes through Q9, Q13 and Q17. Q7 is a voltage source and along with the STV4H's diodes and R17 form the bias spreader of the amplifier.
Here are the questions:
1. Is my general understanding of the circuit correct?
2. How come that the feedback is applied to both: the base of the right T of Q1 and Q5? I would expect that it is applied only to right T of Q1.
3. What is the function of C9 and C31?
4. Can someone explain the dynamics between IPS, VAS and the feedback gain loop?
Thanks in advance.
Best
Soundork
P.S. Here is a little interesting fact as a reward for reading this far: The only thing wrong with it is a bulging capacitor in the PS. The rail voltage is 50V and they have a cap rated at 50V max. Wicked weird…😊
That is a pretty receiver! Nice save!
You've described the PA circuit well. Re questions 2 & 3, C9 is a so-called Miller cap in the VAS that is probably the dominant mechanism rolling off the open-loop gain of the amp--- the VAS gain typically rolls off at 20 dB/decade, sort of an integrator stage. The stages after the VAS have a voltage gain of about 1 and lots of current gain. These power stages have accumulating phase shift with increasing frequency. C31 helps compensate the loop by bypassing the growing phase shift of the power stages by "intervening" with feedback from the VAS--- keeps the amp stable.
Schematic is a bit marginal. Are R21 and R23 both 0.47 ohm? If so, bias drop across R21 of about 25mV is probably about right.
Enjoy!
You've described the PA circuit well. Re questions 2 & 3, C9 is a so-called Miller cap in the VAS that is probably the dominant mechanism rolling off the open-loop gain of the amp--- the VAS gain typically rolls off at 20 dB/decade, sort of an integrator stage. The stages after the VAS have a voltage gain of about 1 and lots of current gain. These power stages have accumulating phase shift with increasing frequency. C31 helps compensate the loop by bypassing the growing phase shift of the power stages by "intervening" with feedback from the VAS--- keeps the amp stable.
Schematic is a bit marginal. Are R21 and R23 both 0.47 ohm? If so, bias drop across R21 of about 25mV is probably about right.
Enjoy!
Q17 and Q19 are SOAR protection transistors, the output stage is Q13 and Q15.
C9 and C31 must be part of the frequency compensation. Instead of the usual heavy Miller compensation, a scheme with a rather small Miller compensation and a feedforward path (C31) is used.
Theoretically there must be a distortion optimum close to the bias point BSST just mentioned. Without indications to the contrary, I would also assume that 25 mV across each 0.47 ohm emitter resistor is about right.
C9 and C31 must be part of the frequency compensation. Instead of the usual heavy Miller compensation, a scheme with a rather small Miller compensation and a feedforward path (C31) is used.
Theoretically there must be a distortion optimum close to the bias point BSST just mentioned. Without indications to the contrary, I would also assume that 25 mV across each 0.47 ohm emitter resistor is about right.
Thanks for replying to my post. You are absolutely correct, the resistors are 0.47 Ohm each. I even checked what's on the board. I have additional questions with regards the circuit operation, but will form it below, because I need to attach stuff...
Best
Soundork
Very good catch and thank you for your reply. I have additional questions and it would be nice if you have the time to chime down.
Best
Soundork
Thanks for all the replies so far.
The quick and dirty schematic diagram above is my interpretation of the circuit between IPS and the VAS.
If I understand how LTP's work in PA's, the input signal is applied at the base of Q1 left and from the collector of it and then it is passed to the base of Q5 which is the VAS transistor. The overall feedback is applied to the base of Q1 right through R11/C7. I do understand that C9 is there for stability, but what I do not understand is how C31 also contributes to the stability. The way I see it, it provides path for the overall feedback to the base of Q5, which defeats the purpose of the LTP.
In my understanding of circuits, the overall feedback is applied to the base of Q1 right and the signal path is from collector of Q1left to base of Q5. The overall feedback isn't fed to the base of Q5.
Best
Soundork
Hi,
I'll try to answer some of your questions, adding some detail to your schematic below. I hope it's legible.
I've filled in some of the component values, drawn in a symbol of the output stage, and show it having a gain of about 1. In fact it will have a voltage gain of slightly less than 1 and will have a low-pass characteristic. As the gain falls with increasing frequency, there will associated increasing phase shift and falling gain.
Assume for a moment that C31 is not present, i.e. 0pF. The amp is essentially an opamp and at mid audio frequencies, the gain will be 1+R11/R3 = 1 +47k/3k = 16.67 At 0Hz, C5 is an open circuit, so the gain is 1; in principle, the amp output is 0V, following the 0V bias at the Q1 left input.
Amplifier compensation is a complex subject, but I'll attempt a crude, overly simple explanation. The input differential stage, in combination with the VAS/Miller integrator, behaves much like a unity-gain-stable opamp. If resistor feedback is taken from the VAS stage, the circuit will be stable with a 90 degree phase margin. But we want to drive the power output stages from the VAS so that the overall amp can benefit from negative feedback. At high frequencies, the output stages will exhibit increasing phase shift, and when accumulated phase shift contributes an additional 90 degree shift, the amp can oscillate.
Now, note how C31 connects back from the VAS to the inverting input. If C31 were the only feedback element, the circuit would be stable, since we know opamp integrators with capacitor feedback are stable. This is how C31 serves to stabilize--- at high frequency, C31's falling reactance begins to dominate when its equals R11 resistance, roughly at 1/(2 PI C31 R11) or about 282kHz, and the amp becomes stable.
I'll try to answer some of your questions, adding some detail to your schematic below. I hope it's legible.
I've filled in some of the component values, drawn in a symbol of the output stage, and show it having a gain of about 1. In fact it will have a voltage gain of slightly less than 1 and will have a low-pass characteristic. As the gain falls with increasing frequency, there will associated increasing phase shift and falling gain.
Assume for a moment that C31 is not present, i.e. 0pF. The amp is essentially an opamp and at mid audio frequencies, the gain will be 1+R11/R3 = 1 +47k/3k = 16.67 At 0Hz, C5 is an open circuit, so the gain is 1; in principle, the amp output is 0V, following the 0V bias at the Q1 left input.
Amplifier compensation is a complex subject, but I'll attempt a crude, overly simple explanation. The input differential stage, in combination with the VAS/Miller integrator, behaves much like a unity-gain-stable opamp. If resistor feedback is taken from the VAS stage, the circuit will be stable with a 90 degree phase margin. But we want to drive the power output stages from the VAS so that the overall amp can benefit from negative feedback. At high frequencies, the output stages will exhibit increasing phase shift, and when accumulated phase shift contributes an additional 90 degree shift, the amp can oscillate.
Now, note how C31 connects back from the VAS to the inverting input. If C31 were the only feedback element, the circuit would be stable, since we know opamp integrators with capacitor feedback are stable. This is how C31 serves to stabilize--- at high frequency, C31's falling reactance begins to dominate when its equals R11 resistance, roughly at 1/(2 PI C31 R11) or about 282kHz, and the amp becomes stable.
Great explanation!
I incorrectly thought it was some sort of feedforward bypass for the input stage, but that made no sense: the polarity was wrong and input stage transistors are usually fast anyway. Your explanation makes perfect sense.
I incorrectly thought it was some sort of feedforward bypass for the input stage, but that made no sense: the polarity was wrong and input stage transistors are usually fast anyway. Your explanation makes perfect sense.
It bypasses the OUTPUT stage. The lead compensation capacitor (C7) is normally used in most circuits. But when trying to design for every bit of loop gain one can get (ie, making C5 small in this case) the accumulated phase shift in the output stage can get to be too much, even with fast outputs. C7 takes the high frequency feedback ahead of the output stage, eliminating its phase shift. One might see this and determine that it’s a bad idea, as distortion correction (from the output stage) at high frequency cannot occur. But what does happen is that the output impedance of the VAS is reduced at high frequency because HF feedback is taken from there. Its as if the miller cap were larger, without the reduction of slew rate that would occur if it were. The lower output impedance of the VAS makes the output stage easier to drive and reduces the crossover distortion it would otherwise produce.
That trick is used most often in very high power vintage amps, with the output triples and hometaxial output transistors - along with the very high gain front ends. It was the only way to get these things stable. However, it’s quite possible to get it wrong and end up with a stability nightmare.
That trick is used most often in very high power vintage amps, with the output triples and hometaxial output transistors - along with the very high gain front ends. It was the only way to get these things stable. However, it’s quite possible to get it wrong and end up with a stability nightmare.