Dave451; Here's a detailed explanation of the Major mods to 400-500/800C (incl the 1800 now) from Dave G. It's about 5 years old so disregard the .1uf coupling cap size, as .068uf is actually closer to the ideal of .072uf (he determined this later) when used with 220K ohm resistors. So just substitute .068uf for .1uf.
START EXPLANATION
400-500C-800C MODS for BIAS and POWER SUPPLY
1.) replace the 4 coupling caps with .1uf/630V caps (OP Choice of cap)
2.) replace 300k grid resistors with 220K R 1/2Watt.
3.) install 10oHm 1/4W resistors to ground @ pin 5 on each output tube.
4.) install 100oHm 1/4W screen stabilization resistors to jumper @ pin 3&8 on each tube. ( I'll give them a try)
5.) install 10K variable resistor (pot) in BIAS POWER lead to facillitate some adjustment in the Bias circuit.
Explanation by Dave Gillespie;
If I may, here is some detail behind the recommendations for these modifications:
1. "Reducing the grid resistors to 220K": The published maximum DC resistance for grid #1 of 7591 class tubes is 300K (by RCA, Sylvania, and others). This is the maximum value allowable when the tube is operated under conditions of maximum dissipation -- which many designs nearly did -- so yes, using 330K is using a value that is right on the ragged edge.
The reason for the maximum resistance specification is to limit the value of reverse grid current voltage that can build up under conditions of maximum dissipation (read that, maximum heat). But there are other concerns that can aggravate this issue as well. Higher line voltages push the heater voltages higher, which means that today, even more heat is generated. The basic concern is that excessive heat (however it's generated) can cause cathode material to boil off and deposit itself onto the grid. This means the grid has now become a partial emitter, with any current flow it provides developing a voltage across the grid resistor that is opposite in polarity to the negative voltage applied to the bottom of the grid resistor. This negative grid current voltage then works against the negative bias voltage, causing the tube to conduct more current, generating even more heat, boiling off more cathode material, and ultimately causing thermal runaway. The production of gas comes into play with these events as well to further aggravate the issue. Therefore, to help control this condition, a maximum DC resistance is specified for the control grid to limit the amount of ill effects that can be caused by reverse grid current.
As a result, with the high temps that these units typically run at inside their cases, and importantly, because of today's higher line voltages, and because the tubes are biased "on" pretty hard to begin with, it makes very good sense to pull the value of the grid resistors back a little to help minimize the chances of developing any significant reverse grid current voltage, and the possible thermal runaway that can produce -- and this is even before the quality of modern production tubes is considered!
2. "Double the value of the coupling caps": This has nothing to do with loading on the driver stage, in fact, lowering the value of the grid resistors actually increases the load on the driver stage. The reason the cap values are increased is to keep nearly the same time constant in the RC coupling circuit between the driver stage and the output stage with the new lower value grid resistors installed.
3. "Install 10 Ohm cathode resistors": This makes for the easiest way to monitor output tube current draw, and therefore allows for the easiest way to set the bias and balance of the output stage (if such controls exist), or check to see that matched output tubes are in fact matched. These resistors are typically used for balancing currents under quiescent conditions, but can also be used to check for balance under dynamic conditions if the proper equipment is available. Balancing output tube quiescent current maximizes OPT low frequency power handling ability (by minimizing core saturation), while balancing dynamic currents minimizes overall distortion.
4. "Installing a 10K bias pot": Modern tubes are manufactured with far less precision than the best tubes from yesteryear were, and so, fixed bias amplifiers operating with tubes of modern manufacture on higher line voltages will usually need some adjustment from the original fixed setting a manufacturer provided. Installing some type of adjustable control is therefore highly desirable, and adding a DC balancing feature to that is even better. Installing these controls so that they cannot be inadvertently altered from their setting goes hand in hand with the installation of any such controls.
5. "Installing 100 ohm Screen Stability Resistors": So many folks have posted on various forums -- from all degrees of experience -- how everything is as it should be, but the output tubes will arc for no reason that they can determine. How expensive does that get when output tubes can be ruined in an instant?
Long ago, I determined that these arcs were in fact occurring from the screen grid. Many pieces of old hifi gear were designed with very little resistance in the screen circuit, which usually didn't present too many problems as the power supply caps of the day had much higher ESR values, which tended to help keep things stable. Even at that however, some designs (Eico HF-89 for example) were simply notorious for arcing output tubes because of the larger overall power supply capacitance values they employed.
Today it is all too common, with the installation of new power supply caps that have much lower ESR values, and which are so often increased in size as well (all installed on new drop in power supply boards), to have all manner of output tube arcing events occur, which leaves folks puzzled because it just never used to do that before. New manufacture tubes got the lion's share of the blame -- until even NOS tubes were arcing in these refurbished pieces of equipment as well.
With some form of transient trigger (music, an open ground, whatever), the lower ESR value caps can set up incredibly strong oscillations in what is effectively a tank circuit that's formed with the screen grid, and can cause an arc at any time. Installation of 100 ohm Screen Stability resistors (for 7591 class tubes) stops this problem virtually every time. I've never had anyone tell me that installation of these resistors did not stop the problem. Personally, I've never had an output tube arc event in nearly twenty years now, where before I had boxes of blown tubes until I could determine what the cause was, and develop an appropriate cure. In equipment such as this Fisher represents, it is very cheap insurance against such events, and produces no audible or measurable difference in performance.
Hopefully, this will help with an understanding of why these modifications can make such worthwhile improvements in so many of these wonderful instruments. With some components being virtually unobtainable now, minimizing any damaging events that might occur helps all of us to enjoy this equipment even more.
Good luck with your receiver!
Dave
END EXPLANATION
