Many of you might be aware of the comments I made in my thread "Improving the Fisher 400" regarding the performance of the phase inverter circuit in this receiver. In particular, Fisher employed what I have come to call a "noose" on this stage, which effectively limits the amount of peak output the stage can produce. I have speculated that this noose was employed as a limited measure of protection for the output tubes, as it was a circuit "feature" that Fisher widely employed in many of its receiver, integrated, and dedicated console designs.
This noose was applied in varying degrees of strangulation across the various Fisher models it was used in. But none have it applied so tightly as that employed in any version of the Fisher 400 receiver. In this model, the phase inverter stage is gasping for what precious little current draw it is allowed to have (the lifeblood of any driver stage), causing it to operate right on the ragged edge of being able to produce any kind of clean, low distortion drive to the output stage, let alone drive it to its full power capabilities with such a signal. The term noose then is appropriately applied, as the stage is ultimately choked down to being able to produce no more drive than the output stage theoretically requires to produce full power.
The design goal is worthy enough I suppose, and the circuit does generally work "OK", but depends heavily on very consistent characteristics for the tubes being used -- both in the phase inverter, and even in the output stage as well. Since Fisher could control those characteristics by way of the tolerance specifications it provided to its tube manufacturers, the units employing these nooses all generally ran pretty well, with the noose always being close by, but largely remaining invisible -- as long as the original tubes were installed that is.
Enter the recent 400 receiver I just encountered. All the component values were well within specification, if not very well matched and close in tolerance to the specified value between the channels of the phase inverter stages in this unit. In the listening room, the unit sounded nice enough for generally listening. But play some more demanding material and kick it a little (who doesn't do that from time to time?), and the unit just seemed to be asthmatic. Heavy bass tracks suffered the most of course. This was not the sound I was accustomed to from a properly operating 400 receiver.
Relevant to the concern at hand, this receiver had just had all the goodies installed to provide some real protection for the output stages (Screen Stability and cathode sampling resistors), individual bias controls, and completely rebuilt power supplies. It also sported a quad of truly NOS GE bottom exhausted 7868 output tubes, which while not matched, all excelled in my power output tube tester. With power supply voltages on the mark, the output stage was clearly not the source of the audible woes.
On the scope, the displays clearly showed what was happening: The phase inverter noose was kicking in as a result of two basic issues:
1. The output tubes employed were all set to 35 ma of quiescent current, generally requiring at least -19 volts (with some over -20) to achieve that level of quiescent current draw. Against the stock unit's bias voltage of -17 volts and much higher quiescent current draw, this means that the phase inverter now had to generate more drive to cover the increased bias voltage the output stage was now operating with -- drive it already struggles to provide with the stock quiescent current levels. In essence, reducing the quiescent current to more sane levels in the output stage then only aggravates an already highly compromised condition.
2. This unit had been purchased sans any tubes, so all tubes for it had to be procured before I received it. As received then, all the 12AX7 positions sported brand new modern production Tung Sol 12AX7 tubes -- which is the real meat of this bulletin.
With these tubes installed -- that is ANY of the 9 versions of this tube that the unit employs installed in the phase inverter positions -- they all exhibited early -- and slow/lazy recovery clipping, this because the tubes simply do not draw enough current in the circuit as designed. However, when good American produced 12AX7s were installed, the effect was largely eliminated.
Upon testing the Gm and AC gain of these Russian tubes against the precise conditions published by RCA under which these parameters of a 12AX7 tube were developed, the Russian tubes get an A+, with the average Gm coming in almost exactly at 1600 micromhos, and the Mu displaying a factor of nearly 100. So what the heck is going on then? American tubes seem to work satisfactorily in the phase inverter position of the 400, but the Russian tubes don't.
I then tested both the American and Russian tubes at the very low current operating point established by the Fisher 400's design (~.5 ma). It is important to note that 12AX7 designs (including Fisher) routinely operate the tube at this current level and lower in many phono preamp and line level stage designs, which is perfectly fine. But only Fisher does so with the tube in a driver position, and that's the rub. At such low current levels, when the load on the tube increases significantly (is numerically reduced) -- AS IT DOES IN A DRIVER POSITION WHEN THE OUTPUT TUBES APPROACH THE ONSET OF GRID CURRENT -- then the Russian tubes simply fold up shop and go home. Again, the tubes are already running on the ragged edge in this design -- even with American tubes, and in that region, the consistency from tube to tube can go right out the window. In human terms, it's like trying to have all children at the age of precise 2 yrs old learn at exactly the same rate and speed. That simply doesn't happen because at that point, you're still well off of the "linear" portion of human existence. At that age, the learning curve for children is all over the map, and such is generally the case with tubes in these operating conditions as well. The Russian tubes simply do not perform as dependably in this region as the American tubes do, under the fixed bias conditions set up by the noose in the Fisher phase inverter design.
The quick solution of course was simply to remove the noose in both channels. This returns the stage to operate as a cathode bias design, which allows the tubes to settle at the operating point that is best for the individual tube installed. Bingo. This instantly resolves the issue then for ANY good tube installed. Now, the Russian tubes perform in an indistinguishable way from the American tubes. Gone is the early clipping and lazy recovery from it, with drive to the output tubes now remaining strong and balanced, right up to and beyond the onset of clipping. In the listening room, this translates into a strong dynamic presentation when pushed, replacing the anemic presentation heard earlier.
Removing the noose is easy. The resistor creating the noose has different designators on the different 400 schematics, but in all cases, it is comprised of the 150K resistor connected to pin #3 of the phase inverter tubes. Simply remove this resistor, and then reset the phase inverter adjustment using the procedure presented in the sticky information. And that's it.
The bottom line is that for best performance from your 400, remove the noose, or if not, at least use tubes of American manufacture in the phase inverter positions. Then -- enjoy!!
Dave
This noose was applied in varying degrees of strangulation across the various Fisher models it was used in. But none have it applied so tightly as that employed in any version of the Fisher 400 receiver. In this model, the phase inverter stage is gasping for what precious little current draw it is allowed to have (the lifeblood of any driver stage), causing it to operate right on the ragged edge of being able to produce any kind of clean, low distortion drive to the output stage, let alone drive it to its full power capabilities with such a signal. The term noose then is appropriately applied, as the stage is ultimately choked down to being able to produce no more drive than the output stage theoretically requires to produce full power.
The design goal is worthy enough I suppose, and the circuit does generally work "OK", but depends heavily on very consistent characteristics for the tubes being used -- both in the phase inverter, and even in the output stage as well. Since Fisher could control those characteristics by way of the tolerance specifications it provided to its tube manufacturers, the units employing these nooses all generally ran pretty well, with the noose always being close by, but largely remaining invisible -- as long as the original tubes were installed that is.
Enter the recent 400 receiver I just encountered. All the component values were well within specification, if not very well matched and close in tolerance to the specified value between the channels of the phase inverter stages in this unit. In the listening room, the unit sounded nice enough for generally listening. But play some more demanding material and kick it a little (who doesn't do that from time to time?), and the unit just seemed to be asthmatic. Heavy bass tracks suffered the most of course. This was not the sound I was accustomed to from a properly operating 400 receiver.
Relevant to the concern at hand, this receiver had just had all the goodies installed to provide some real protection for the output stages (Screen Stability and cathode sampling resistors), individual bias controls, and completely rebuilt power supplies. It also sported a quad of truly NOS GE bottom exhausted 7868 output tubes, which while not matched, all excelled in my power output tube tester. With power supply voltages on the mark, the output stage was clearly not the source of the audible woes.
On the scope, the displays clearly showed what was happening: The phase inverter noose was kicking in as a result of two basic issues:
1. The output tubes employed were all set to 35 ma of quiescent current, generally requiring at least -19 volts (with some over -20) to achieve that level of quiescent current draw. Against the stock unit's bias voltage of -17 volts and much higher quiescent current draw, this means that the phase inverter now had to generate more drive to cover the increased bias voltage the output stage was now operating with -- drive it already struggles to provide with the stock quiescent current levels. In essence, reducing the quiescent current to more sane levels in the output stage then only aggravates an already highly compromised condition.
2. This unit had been purchased sans any tubes, so all tubes for it had to be procured before I received it. As received then, all the 12AX7 positions sported brand new modern production Tung Sol 12AX7 tubes -- which is the real meat of this bulletin.
With these tubes installed -- that is ANY of the 9 versions of this tube that the unit employs installed in the phase inverter positions -- they all exhibited early -- and slow/lazy recovery clipping, this because the tubes simply do not draw enough current in the circuit as designed. However, when good American produced 12AX7s were installed, the effect was largely eliminated.
Upon testing the Gm and AC gain of these Russian tubes against the precise conditions published by RCA under which these parameters of a 12AX7 tube were developed, the Russian tubes get an A+, with the average Gm coming in almost exactly at 1600 micromhos, and the Mu displaying a factor of nearly 100. So what the heck is going on then? American tubes seem to work satisfactorily in the phase inverter position of the 400, but the Russian tubes don't.
I then tested both the American and Russian tubes at the very low current operating point established by the Fisher 400's design (~.5 ma). It is important to note that 12AX7 designs (including Fisher) routinely operate the tube at this current level and lower in many phono preamp and line level stage designs, which is perfectly fine. But only Fisher does so with the tube in a driver position, and that's the rub. At such low current levels, when the load on the tube increases significantly (is numerically reduced) -- AS IT DOES IN A DRIVER POSITION WHEN THE OUTPUT TUBES APPROACH THE ONSET OF GRID CURRENT -- then the Russian tubes simply fold up shop and go home. Again, the tubes are already running on the ragged edge in this design -- even with American tubes, and in that region, the consistency from tube to tube can go right out the window. In human terms, it's like trying to have all children at the age of precise 2 yrs old learn at exactly the same rate and speed. That simply doesn't happen because at that point, you're still well off of the "linear" portion of human existence. At that age, the learning curve for children is all over the map, and such is generally the case with tubes in these operating conditions as well. The Russian tubes simply do not perform as dependably in this region as the American tubes do, under the fixed bias conditions set up by the noose in the Fisher phase inverter design.
The quick solution of course was simply to remove the noose in both channels. This returns the stage to operate as a cathode bias design, which allows the tubes to settle at the operating point that is best for the individual tube installed. Bingo. This instantly resolves the issue then for ANY good tube installed. Now, the Russian tubes perform in an indistinguishable way from the American tubes. Gone is the early clipping and lazy recovery from it, with drive to the output tubes now remaining strong and balanced, right up to and beyond the onset of clipping. In the listening room, this translates into a strong dynamic presentation when pushed, replacing the anemic presentation heard earlier.
Removing the noose is easy. The resistor creating the noose has different designators on the different 400 schematics, but in all cases, it is comprised of the 150K resistor connected to pin #3 of the phase inverter tubes. Simply remove this resistor, and then reset the phase inverter adjustment using the procedure presented in the sticky information. And that's it.
The bottom line is that for best performance from your 400, remove the noose, or if not, at least use tubes of American manufacture in the phase inverter positions. Then -- enjoy!!
Dave