I have a Fisher 500 mono receiver. The chassis is marked "EL37" for the output tubes. The EL37's were replaced by 6L6GB's by a previous owner. EL37's are scarce and way out of my budget, and new KT66 won't fit (glass diameter is too large). The only Sam's schematic I can find is very close, but lists EL34/KT66. I purchased some new Tungsram 6l6GC's, which seem to work fine (sound good, no red spots on the plates). Do I need to recheck the bias for the replacement tubes, and for future tube changes? There is no adjustment, just fixed resistors. I have replaced all the electrolytic and paper caps, and a couple of bad mica caps that screwed up the FM, and installed a grounded power cord with a safety cap. It sounds great, very quiet- just not sure if I need to reset the bias.
Fisher 500 mono tube bias
- Thread starter mmblh
- Start date
-
- Tags
- bias adjustement ta-500
Thee are resistor changes between EL37 and 6L6, to give a greater negative voltage. Maybe already done, since it came with 6L6s. The Sams schematic should be correct for 5881 or 6L6. The Sams is wrong with BOTH tube numbers - it never came with KT66 or EL34 - was only EL37 or 5881.
ivan100000
Active Member
I've always been partial to this incredible receiver that is still the best vintage Fisher value out there. I have bought and sold at least a dozen of these over the years, and I've eyeballed many many more. As Tom Bavis mentions above, I'm sure that the SAMS schematic claim of the EL34/KT66 is some sort of typo, as I have never seen or even heard tell of one in the wild. My advice? Bite the bullet and buy a solid pair of the Mullard EL37 from a reliable source; it will be the last time you will ever have to worry about output tubes in this amp as the mono 500 doesn't put out anywhere near the voltage that the majestic, coke-bottled EL37 can handle. The tubes will outlast you.
Sorry for the delay- thanks to both for the replies.
I am going to keep my eyes open for the EL37's (will be an interesting conversation with my accountant/ wife), but it would be reassuring to know if I have the current tubes set up correctly. Checking bias seems to be a routine for many amps- is the procedure the same as with guitar amplifiers? My skills were good enough to re-cap, and troubleshoot the FM, and I have enough respect for the voltages to avoid getting hurt, just not sure where to begin at least verifying the bias.
I am going to keep my eyes open for the EL37's (will be an interesting conversation with my accountant/ wife), but it would be reassuring to know if I have the current tubes set up correctly. Checking bias seems to be a routine for many amps- is the procedure the same as with guitar amplifiers? My skills were good enough to re-cap, and troubleshoot the FM, and I have enough respect for the voltages to avoid getting hurt, just not sure where to begin at least verifying the bias.
5881 (a 6L6 variant) was used in later production, so 6L6GC is an option as well. (Would be MY choice, as I have a dozen or more, and zero EL37s...)
This was designed with no bias adjustment or monitoring points, so you may have to add current sample resistors (i.e. 10 Ohms in series with each cathode) as means to measure idle current. And changing it will require resistor changes - might as well install a pot or two while you're at it. Biasing circuit values were changed by Fisher when the change to 5881 was made - may have been done to this unit already. Compare Sams schematic to earlier Fisher K100 at http://www.fisherconsoles.com/service manuals/fisher k100 chassis.pdf (K100 is same as 500 except that eye tubes were used instead of meters)
This was designed with no bias adjustment or monitoring points, so you may have to add current sample resistors (i.e. 10 Ohms in series with each cathode) as means to measure idle current. And changing it will require resistor changes - might as well install a pot or two while you're at it. Biasing circuit values were changed by Fisher when the change to 5881 was made - may have been done to this unit already. Compare Sams schematic to earlier Fisher K100 at http://www.fisherconsoles.com/service manuals/fisher k100 chassis.pdf (K100 is same as 500 except that eye tubes were used instead of meters)
Last edited:
I’m working on a fisher 500 mono receiver and I saw this old thread so I thought I’d ask for some help and hope that some of you experts see it. Regarding the -36v supply, I have recapped all electrolytic and wax capacitors but I can only get the -36v supply to about -32v. Since this feeds the 3-12ax7 tube filaments and is the dc bias supply for the power tubes it seems like it would be necessary to be at -36v. I replaced the 5w 400 ohm resistor with a 1000 ohm resistor and have been able to maintain -36v. Is there a better way to control this voltage? All the B+ voltages are close to spec and the amp sounds very good. Just wondering how much I should concern myself with the -36v supply. Thanks!
This circuit is a variation of Fisher's tried and true formulation of having the output tube cathode current power the heaters of certain small signal tubes. In this case however, it is the current flow of the entire receiver drawing current through the small signal tubes -- and of course that 400Ω resistor. As such then the success of the circuit depends on the bias voltage needs of the output tubes (to produce the correct quiescent current) being basically the same voltage that it takes to light the heaters of three 12AX7 tubes wired in series using their 12 volt heater configuration. The 400Ω resistor is there to prevent too much voltage rise with increasing power output from the Class AB1 output stage (an operating mode that has the output stage drawing significantly more current as power output is increased).
But of course, that's the rub -- different output tubes draw different current levels for a give grid bias voltage. So unless you have weak output tubes installed (which would definitely cause your issue), it's a virtually certainty that you have output tubes installed that draw less current for a given grid bias voltage, than the tubes Fisher sold the unit with (large manufacturers could always specify a particular characteristic of the tubes they buy from tube manufacturers, because they purchase so many). So like any cathode biased circuit, the actual bias voltage produced (across the heaters in this case) is a function of equilibrium: current flow through the bias system increases bias voltage until the bias voltage won't allow any more current to flow (I'm targeting the output tubes since they are the largest current sink in the design).
Now if you increase the 400Ω resistor to 1000Ω, that allows for more bias voltage to be developed, but also allows for a greater rise in bias voltage with increasing power output. Short of finding a matched set of output tubes that have the same characteristic that the tubes Fisher used, the best answer is to install a bias control. You'd still need to use matched output tubes, but you'd just set the output tube cathodes to a potential of 36 volts and be done with it.
A bias control would be easy enough to install, only requiring a 5000Ω pot (linear, .5 watt), and a .5 watt 24K resistor. Connect one side of the pot to the -36 volt source. Connect the other side of the pot to one lead of the 24K resistor, and ground the other end of the resistor. Disconnect the 33K resistor supplying bias voltage to the output tubes from the -36 volt source, and connect it to the wiper of the pot -- and that's it.
This arrangement will allow as much bias voltage as before when the wiper of the pot is turned towards the -36 volt source, and as much as about 6 volts less bias as the wiper is turned towards the 24K resistor. Of course this arrangement will do nothing if you have the opposite problem (too much bias voltage with the bias control already turned full towards the -36 volt source) -- all you can do in that instance is get tubes more in keeping with those you have installed now, that don't produce enough bias voltage, and then let the bias control bring them up to speed.
Understand that the small signal tube heaters will do just fine with anything from about -33 to -36 volts across them. Warm up time will increase as the bias voltage is lowered, but the main concern is that with -36 vdc across the three heaters and 400Ω resistor, the correct current draw will then be established to set the correct operating point for the output stage, allowing it to deliver optimum performance.
BTW -- Welcome to AK!
I hope this helps!
Dave
But of course, that's the rub -- different output tubes draw different current levels for a give grid bias voltage. So unless you have weak output tubes installed (which would definitely cause your issue), it's a virtually certainty that you have output tubes installed that draw less current for a given grid bias voltage, than the tubes Fisher sold the unit with (large manufacturers could always specify a particular characteristic of the tubes they buy from tube manufacturers, because they purchase so many). So like any cathode biased circuit, the actual bias voltage produced (across the heaters in this case) is a function of equilibrium: current flow through the bias system increases bias voltage until the bias voltage won't allow any more current to flow (I'm targeting the output tubes since they are the largest current sink in the design).
Now if you increase the 400Ω resistor to 1000Ω, that allows for more bias voltage to be developed, but also allows for a greater rise in bias voltage with increasing power output. Short of finding a matched set of output tubes that have the same characteristic that the tubes Fisher used, the best answer is to install a bias control. You'd still need to use matched output tubes, but you'd just set the output tube cathodes to a potential of 36 volts and be done with it.
A bias control would be easy enough to install, only requiring a 5000Ω pot (linear, .5 watt), and a .5 watt 24K resistor. Connect one side of the pot to the -36 volt source. Connect the other side of the pot to one lead of the 24K resistor, and ground the other end of the resistor. Disconnect the 33K resistor supplying bias voltage to the output tubes from the -36 volt source, and connect it to the wiper of the pot -- and that's it.
This arrangement will allow as much bias voltage as before when the wiper of the pot is turned towards the -36 volt source, and as much as about 6 volts less bias as the wiper is turned towards the 24K resistor. Of course this arrangement will do nothing if you have the opposite problem (too much bias voltage with the bias control already turned full towards the -36 volt source) -- all you can do in that instance is get tubes more in keeping with those you have installed now, that don't produce enough bias voltage, and then let the bias control bring them up to speed.
Understand that the small signal tube heaters will do just fine with anything from about -33 to -36 volts across them. Warm up time will increase as the bias voltage is lowered, but the main concern is that with -36 vdc across the three heaters and 400Ω resistor, the correct current draw will then be established to set the correct operating point for the output stage, allowing it to deliver optimum performance.
BTW -- Welcome to AK!
I hope this helps!
Dave
Dave - I apologize for my clumsy responses. I’m trying to figure out which buttons to push on here.
You filled in several gaps for me. It makes sense regarding the tubes needing to match the circuit. I have a pair of of EL37s and 6L6EH (electro harmonix) but not the original Fisher EL37s. Both sets sound good and have a quiescent voltage of approx -35.1v at each tube with the -36v supply. Does this indicate the tubes are matched in each pair? in that they have the same idle current (voltage drop) or am I reading too much into that? Regardless of that I am gong to pursue the circuit mod you suggested. After replacing the 400 resistor with the 1k and increasing the volume the -36v rail voltage went to -40v + for periods like you predicted. One question on the modification is why doesn’t the resistance of the max pot resistance and fixed resistor meet or exceed the 33k resistor that is being removed? Is it because you’re trying to decrease the voltage drop to the tubes so you always want to be on the low side (if that makes sense)?
Again, I really appreciate your insight and help. This receiver is a really neat piece. It’s hard to believe how well built it is. The tuner pulls in stations very well and the sound is very clean when streaming to it!
You filled in several gaps for me. It makes sense regarding the tubes needing to match the circuit. I have a pair of of EL37s and 6L6EH (electro harmonix) but not the original Fisher EL37s. Both sets sound good and have a quiescent voltage of approx -35.1v at each tube with the -36v supply. Does this indicate the tubes are matched in each pair? in that they have the same idle current (voltage drop) or am I reading too much into that? Regardless of that I am gong to pursue the circuit mod you suggested. After replacing the 400 resistor with the 1k and increasing the volume the -36v rail voltage went to -40v + for periods like you predicted. One question on the modification is why doesn’t the resistance of the max pot resistance and fixed resistor meet or exceed the 33k resistor that is being removed? Is it because you’re trying to decrease the voltage drop to the tubes so you always want to be on the low side (if that makes sense)?
Again, I really appreciate your insight and help. This receiver is a really neat piece. It’s hard to believe how well built it is. The tuner pulls in stations very well and the sound is very clean when streaming to it!
Last edited:
No -- it indicates that the two pairs of tubes -- collectively as a pair -- each draw the same amount of current. But within each pair, you still have no idea if one tube is pulling more current than the other tube or not. To determine that, you would need to add small precision 10Ω, 0.25 watt resistors at pin 8 of each output tube (disconnect the ground connection at pin 8 and insert the resistors between pin 8 and ground). With the resistors installed, then you could measure the voltage across each one, and through that (and Ohm's Law) determine the amount of current each tube is flowing.
You are not removing the 33K resistor at all. You are simply disconnecting the end of it that is connected to the -36 volt point, and then reconnecting it to the wiper of the bias pot. That resistor (and the .22 uF cap appearing after it) are acting to isolate the output tube grid circuit from the ripple voltage appearing across the DC powered heaters, so as to not introduce any hum in the output.
Dave
You are not removing the 33K resistor at all. You are simply disconnecting the end of it that is connected to the -36 volt point, and then reconnecting it to the wiper of the bias pot. That resistor (and the .22 uF cap appearing after it) are acting to isolate the output tube grid circuit from the ripple voltage appearing across the DC powered heaters, so as to not introduce any hum in the output.
Dave
Thanks, Dave. I now understand what you’re saying about the tube bias/matching and the bias circuit with the 5k pot and 24k resistor. For my education, how did you determine the pot and resistor values? If I wanted to bias each tube separately (after installing the cathode current measuring resistors), what would that circuit look like? My EL37 tubes are not “matched” unless by blind luck. Thanks! Doug
