Now that I've restored proper operation in the AM and FM tuners, MPX decoder, and FM Automatic matrix switching circuits in my 800C (there were problems in all four of these sections), I've now turned my attention to the audio section.
When received, it still had the "original" Fisher 7591 output tubes installed, which were advertised as having tested fine, but in reality were so bad that although "functional", they would only allow a small fraction of the unit's performance to be produced. But, what the hey, everything else was advertised as operating properly in this unit as well, so it fit right in with the Grand Canyon of difference that existed between truth and reality by the seller. At any rate, with the unit really starting to come together, a decent set of low mileage Westinghouse 7591's was purchased as the next order of business to attend to, in the quest to make this unit be all that Fisher intended it to be.
As is so often the case today however, while the new tubes were a great improvement over the old ones (almost anything would have been!), one was out in left field with it's grid bias voltage requirements. This was aggravated by the fact that the tech who serviced the unit in readying it for sale put in a bias modification to cool down the output tubes. Boy did it ever cool them down! Good tubes were only drawing 9, 11 and 13 ma of total cathode current each, with the odd-ball pulling 22 ma, and collectively producing an abundance of crossover distortion at even modest power levels. Throttling back the bias voltage a bit brought the three tubes up into the 16-18 ma territory, and the odd-ball up to about 28 ma. Clearly, this would hardly do for the otherwise great tubes this receiver now had, and the condition it was rising to. The ability to adjust each output tube individually then became top of the list important.
When it comes to installing individual output tube bias adjustments, the approach incorporating DC Bias and Balance controls in each channel offers a number of advantages over having four separate bias controls:
1. The overall adjustment process is easier to execute, with only two bias controls requiring adjustment to match channel performance rather than four.
2. It offers greater fail-safe protection when properly designed, where if the wiper of either the DC Bias or Balance control in a given channel lifts away from its carbon track, both output tubes in that channel immediate go to a maximum negative grid bias condition.
3. It is the only configuration that allows for the complete output stage quiescent current draw to be easily adjusted up or down (both tubes in unison) to set the stage for the lowest distortion operating point.
When Fisher included a FULL complement of output tube adjustment controls, the DC Bias and Balance approach is always the approach they used. Think SA-100, X-202/X-202B, X-1000 (EL34) and K(X)-200 integrated amplifiers as some examples of this type of design. If I was going to install adjustment controls to optimize output stage operation, this is the type I wanted to employ.
Also, many of the examples of individual biasing schemes I could find had the controls and/or test points located underneath the receiver chassis, which I try to avoid at all cost. As I get older, these things get heavier and more cumbersome. Fixed umbilical cords to separate external boards or monitoring boxes can address that, but make removal of the unit a hassle from a cabinet when necessary. Therefore, anything I can do to either eliminate removal, or minimize the maneuvering of these things when removal is required is a blessing for sure!
With these design and installation goals then framing the project, I thought I'd share the approach I used to achieve them, as but one more offering to include in the stable of available options when these types of circuits are installed in Fisher receivers.
The circuit containing the 4 pots, 6 resistors, and 4 caps, along with the 5 test points (one being ground) for use when making adjustments are all contained on a small piece of perf board measuring 1" tall X 2&13/16" wide. The board requires 10 leads to connect it into the 800C, which was conveniently handled by a 9 conductor shielded cable, with the shield acting as the 10th (ground) lead. The leads do not need to be shielded, but it provided a convenient way to neatly bundle the 10 leads required to install the board.
The test points are nothing but insulated closed-end wire connectors bent into an "L" shape to form a convenient point for mounting the board to the chassis as well as a test point for the circuit, with similar connectors bent into the same shape but mounted on the back side of the board behind the test points to allow for a connection to the test point. The insulated barrel of the wire connectors forming the test points are then affixed to the chassis with hot glue. I know. Cheezy. But in practice, it works quite well, and is in keeping with my effort of being only minimally invasive when necessary, and only when necessary. This time, it wasn't necessary, and all work can easily be reversed without a trace of the installation remaining. The pics show all of this rather clearly.
The final results turned out very well indeed. Located just behind the Reverb connection jacks on the top of the chassis, the bias can now be measured and adjusted quite easily at will, even when installed in a cabinet. The wiring from the board enters the chassis through one of the ventilation holes drilled around one of the can caps, and is secured underneath by a wire restraint.
Pics include:
1. The front side of the board during construction. You can see how compact it is, yet contains everything needed to properly adjust the output tubes for optimum operation.
2. The rear side of the board during construction. The wiring is close quarters, but hardly the tightest quarters.
3. The assembly was finished by attaching the flying pigtail lead that will connect the board into the circuits of the 800C. Each twisted pair represents the wiring to a specific output tube: one lead connects to the cathode terminal to establish test point operation, while the other is the adjustable grid bias supply source from the board for that same tube. Individual 10 ohm cathode sampling resistors are installed at each output tube socket, as are individual 100 ohm Screen Stability resistors as well.
4. A close up of the board mounted in place. It was my first hot glue experience so the job is not the neatest, but does the job of mounting the board down quite admirably none the less. The center test point is a ground point for the negative meter lead, while the test points on either side of the ground test point represent a channel, with the test points electrically connected/arranged in the same order as the output tubes themselves appear on the chassis. For long term memory of the test point/output tube relationship, the two most left terminals are color coded green, with a green dot then marked on the chassis between the two most left (left channel) output tubes. Red indicates the same information for the right channel tubes and test points.
The control centered above a channel's pair of test points is that channel's DC balance control. It is first adjusted for 0.0 vdc between the two test points below it. The control next to each channel's DC Balance control is that channel's Bias control, which is then adjusted for the target cathode voltage from either of that channel's test points to ground after the DC Balance control is adjusted. The range of the controls are such that the least negative bias voltage that can be applied to either tube in each channel is -13.5 vdc, preventing any catastrophic tube damage from setting the grid bias voltage too low.
5. A rear view of the finished installation, showing how the flying pigtail lead snakes around the Reverb jacks, making them still fully usable if needed, before entering the chassis through a ventilation hole.
The design allows for a 3 volt spread over the range of the DC Balance control, and another 3 volt spread over the range of the Bias control. This nicely accommodates the rather wide spread of the tubes installed in one channel, while still allowing an easy enough adjustment of the two more closely matched tubes in the other channel. The output tubes are now all happily idling at 30 ma each.
In the final post, I'll show some underside shots of how the board's pigtail lead is connected into the wiring of the receiver, and a quick scribble of the circuit I developed for this project.
Dave
When received, it still had the "original" Fisher 7591 output tubes installed, which were advertised as having tested fine, but in reality were so bad that although "functional", they would only allow a small fraction of the unit's performance to be produced. But, what the hey, everything else was advertised as operating properly in this unit as well, so it fit right in with the Grand Canyon of difference that existed between truth and reality by the seller. At any rate, with the unit really starting to come together, a decent set of low mileage Westinghouse 7591's was purchased as the next order of business to attend to, in the quest to make this unit be all that Fisher intended it to be.
As is so often the case today however, while the new tubes were a great improvement over the old ones (almost anything would have been!), one was out in left field with it's grid bias voltage requirements. This was aggravated by the fact that the tech who serviced the unit in readying it for sale put in a bias modification to cool down the output tubes. Boy did it ever cool them down! Good tubes were only drawing 9, 11 and 13 ma of total cathode current each, with the odd-ball pulling 22 ma, and collectively producing an abundance of crossover distortion at even modest power levels. Throttling back the bias voltage a bit brought the three tubes up into the 16-18 ma territory, and the odd-ball up to about 28 ma. Clearly, this would hardly do for the otherwise great tubes this receiver now had, and the condition it was rising to. The ability to adjust each output tube individually then became top of the list important.
When it comes to installing individual output tube bias adjustments, the approach incorporating DC Bias and Balance controls in each channel offers a number of advantages over having four separate bias controls:
1. The overall adjustment process is easier to execute, with only two bias controls requiring adjustment to match channel performance rather than four.
2. It offers greater fail-safe protection when properly designed, where if the wiper of either the DC Bias or Balance control in a given channel lifts away from its carbon track, both output tubes in that channel immediate go to a maximum negative grid bias condition.
3. It is the only configuration that allows for the complete output stage quiescent current draw to be easily adjusted up or down (both tubes in unison) to set the stage for the lowest distortion operating point.
When Fisher included a FULL complement of output tube adjustment controls, the DC Bias and Balance approach is always the approach they used. Think SA-100, X-202/X-202B, X-1000 (EL34) and K(X)-200 integrated amplifiers as some examples of this type of design. If I was going to install adjustment controls to optimize output stage operation, this is the type I wanted to employ.
Also, many of the examples of individual biasing schemes I could find had the controls and/or test points located underneath the receiver chassis, which I try to avoid at all cost. As I get older, these things get heavier and more cumbersome. Fixed umbilical cords to separate external boards or monitoring boxes can address that, but make removal of the unit a hassle from a cabinet when necessary. Therefore, anything I can do to either eliminate removal, or minimize the maneuvering of these things when removal is required is a blessing for sure!
With these design and installation goals then framing the project, I thought I'd share the approach I used to achieve them, as but one more offering to include in the stable of available options when these types of circuits are installed in Fisher receivers.
The circuit containing the 4 pots, 6 resistors, and 4 caps, along with the 5 test points (one being ground) for use when making adjustments are all contained on a small piece of perf board measuring 1" tall X 2&13/16" wide. The board requires 10 leads to connect it into the 800C, which was conveniently handled by a 9 conductor shielded cable, with the shield acting as the 10th (ground) lead. The leads do not need to be shielded, but it provided a convenient way to neatly bundle the 10 leads required to install the board.
The test points are nothing but insulated closed-end wire connectors bent into an "L" shape to form a convenient point for mounting the board to the chassis as well as a test point for the circuit, with similar connectors bent into the same shape but mounted on the back side of the board behind the test points to allow for a connection to the test point. The insulated barrel of the wire connectors forming the test points are then affixed to the chassis with hot glue. I know. Cheezy. But in practice, it works quite well, and is in keeping with my effort of being only minimally invasive when necessary, and only when necessary. This time, it wasn't necessary, and all work can easily be reversed without a trace of the installation remaining. The pics show all of this rather clearly.
The final results turned out very well indeed. Located just behind the Reverb connection jacks on the top of the chassis, the bias can now be measured and adjusted quite easily at will, even when installed in a cabinet. The wiring from the board enters the chassis through one of the ventilation holes drilled around one of the can caps, and is secured underneath by a wire restraint.
Pics include:
1. The front side of the board during construction. You can see how compact it is, yet contains everything needed to properly adjust the output tubes for optimum operation.
2. The rear side of the board during construction. The wiring is close quarters, but hardly the tightest quarters.
3. The assembly was finished by attaching the flying pigtail lead that will connect the board into the circuits of the 800C. Each twisted pair represents the wiring to a specific output tube: one lead connects to the cathode terminal to establish test point operation, while the other is the adjustable grid bias supply source from the board for that same tube. Individual 10 ohm cathode sampling resistors are installed at each output tube socket, as are individual 100 ohm Screen Stability resistors as well.
4. A close up of the board mounted in place. It was my first hot glue experience so the job is not the neatest, but does the job of mounting the board down quite admirably none the less. The center test point is a ground point for the negative meter lead, while the test points on either side of the ground test point represent a channel, with the test points electrically connected/arranged in the same order as the output tubes themselves appear on the chassis. For long term memory of the test point/output tube relationship, the two most left terminals are color coded green, with a green dot then marked on the chassis between the two most left (left channel) output tubes. Red indicates the same information for the right channel tubes and test points.
The control centered above a channel's pair of test points is that channel's DC balance control. It is first adjusted for 0.0 vdc between the two test points below it. The control next to each channel's DC Balance control is that channel's Bias control, which is then adjusted for the target cathode voltage from either of that channel's test points to ground after the DC Balance control is adjusted. The range of the controls are such that the least negative bias voltage that can be applied to either tube in each channel is -13.5 vdc, preventing any catastrophic tube damage from setting the grid bias voltage too low.
5. A rear view of the finished installation, showing how the flying pigtail lead snakes around the Reverb jacks, making them still fully usable if needed, before entering the chassis through a ventilation hole.
The design allows for a 3 volt spread over the range of the DC Balance control, and another 3 volt spread over the range of the Bias control. This nicely accommodates the rather wide spread of the tubes installed in one channel, while still allowing an easy enough adjustment of the two more closely matched tubes in the other channel. The output tubes are now all happily idling at 30 ma each.
In the final post, I'll show some underside shots of how the board's pigtail lead is connected into the wiring of the receiver, and a quick scribble of the circuit I developed for this project.
Dave
Attachments
Last edited: