500C - 7591 killer

Discussion in 'Fisher' started by Jailtime, Jan 4, 2018.

  1. Jailtime

    Jailtime Standin' on a corner Subscriber

    Messages:
    11,375
    Location:
    Velva ND
    I restored this 500C a few years ago. http://audiokarma.org/forums/index.php?threads/not-messin-around-anymore-500c-rebuild.545832/. It was pretty happy with a set of matched EH7591s, but I wore those out after 3 solid years of listening. Decided to try a set of the new production Tung Sol 7591As, the 500C didn't really like them from the start. I could only get -16.6V negative bias on them, for a draw of 40mA per tube, best I could do. I've been running them for probably a year that way.

    I noticed recently that one of the tubes (V10) was looking a bit heat stressed on the plates. So I took the Fisher out of the rack for some testing. My line voltage is 125VAC, the receiver is seeing 110 after the pair of CL70s I installed during the restoration. Pin 3 on the outputs is a nice 435V, pin 8 is 403V, pin 6 is -16.6V (that's as good as the neg bias gets on my 10K pot). The tubes are all over the place for current draw, V8 is 20.4mA, V9 is 43.8, V10 is 25.6, and V11 is 38.3. Looks like I have a problem with balance between the tubes, since the even numbered ones are trashed. Would an IBBA fix this with a new set of tubes? Is there anything else to check out to see why this 500C is killing tubes? Phase inverter adjustment, perhaps?

     

    Attached Files:

     

    Please register to disable this ad.

  2. AlTinkster92

    AlTinkster92 AK Subscriber Subscriber

    Messages:
    3,158
    Location:
    NC
    Jailtime, I certainly would suggest a IBAM , those mA numbers sure aren't balanced at all. I believe that the mA should be within 5 percent of each other? Dave G put a IBAM on my 800C and my tubes were far from matched. Brand new never used Westies were installed and they are running fine 4 years later. Good luck with it! Al
     
  3. larryderouin

    larryderouin Turn it UP, POP? PLLUUEEEZZZZZEE Subscriber

    Messages:
    21,656
    Location:
    Glen Burnie Md.
    Jailtime. You definately need to install either the IBAM or the IBBA. 40ma on the tubes is high. Presuming a plate voltage of 435V, 40ma gives you approx 17.4w per tube( not including subtraction of 3ma for the screens). The preferred range of operation for 7591's now is 70% to 80% of maximum dissipation of 19W. This is between 13.3 and 15.2 watts per tube. All of my 7591 receivers have either EH's or Tung-Sol's. The EH's are approx 8 yrs old and bias up like new. gm is about 10% below what it was when tubes were new. They have approx 1500 hours on them.

    Understand from reading your thread on rehab of this unit you rebuilt the Power supply, and added Screen stability resistors, 200K grid return resistors and .1uf coupling caps. No mention of balancing the P.I., nor did you give final details of voltages, bias readings , etc. No problem. We can get new voltages from you.

    The Tung-Sol Tubes will, once you have the IBAM/IBBA boards in, will need more negative voltage on the BIAS Circuit. They DO run hotter than an equivalent EH7591 or an OLD STOCK 7591. For Example, a Westie or EH7591 might need -17V to have a bias current of 36ma. The Tungsol would need approx -20v to -21v to maintain the same level.

    Remember when setting bias to concentrate on the current readings from pin 5, rather than the vague setting of the voltage on pin 6. The pin 6 reading was used by the factory as they had matched tubes installed in the units and set them for maximum dissipation to meet or exceed published specs. This tended to burn out tubes in 1-2 years. Current draw is a more accurate setting and reflects the true operation of the tube. Think of a car engine. Everybody raves about Horsepower, but very few know or understand Torque. Horsepower is a function of torque x RPM. You can measure Torque directly but have to extrapolate horsepower. So in this case, you want a direct reading of power (Torque), via pin 5, and multiply that by rpm(in this case the plate voltage on pin 3). This gives you Horsepower (Watts). With the over max dissipation, they also set the grid return resistors to more than the tube manufacturers specification. So you have a factory set that is a tube hog.

    Backing off the Grid return resistors to 220K, and backing off the bias current to btwn 70% and 80% max will extend tube life, give you more than enough SPL's to Crank the volume to Ear blowing levels, and with minimal distortion. 435V x .400mv (divide by 10 ohms to get 40ma) will give 17.2W. The screen uses about 3ma. So 435 x37ma (.037) =16.095 watt per tube. It'll need to come down some and then balance the tubes (This the IBBA board is set to do, the IBAM just sets the bias individually. ) Ideally the tube current draw should be between 15.2 and 13.3W for 80% and 70%. Setting the bias for .035mv after subtracting the screen dissipation will give you 15.225W (38ma on the resistor) which is just on the 80% mark. 31ma (after screen dissipation is subtracted. 34ma before subtraction.) will give you 13.4W or just over 70%.

    The R/C network consisting of the grid returns and the coupling caps, is changed when 220K and .1uf is installed. The closest numbers to the original R/C network values is gotten with the use of a 220K resistor and a .068uf cap. Ideally .071uf would get extremely close to the original values if not spot on. Dave G. did the math. There is no change in sound quality with this combination. Leaving the .1uf caps in place will not hurt the unit.
     
    Last edited: Jan 5, 2018
  4. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    35,813
    Location:
    Southern NJ
    probably just running them too hard honestly. Phase inverter balance won't cause the tubes to age poorly but it will result in higher output distortion than if it was adjusted right. There is a relatively simple method of adjusting that based on a few measurements with an ohm meter or if you have an IMD analyzer you can just tweak it for best results on that.

    might also be worth checking your bias supply if it won't go more negative than -16.6 volts. Possibly an off-value resistor somewhere in there. Largely that will get replaced with a bias adjust board but you do want to verify the output voltage from the rectifier is proper or it may not have enough adjustment range.
     
  5. Jailtime

    Jailtime Standin' on a corner Subscriber

    Messages:
    11,375
    Location:
    Velva ND
    I did tweak the PI pots back when the old Telefunkens of unknown condition were in the inverter spots, used one of the "close enough" methods mentioned in old threads here. I should probably check that again now that I put new matched Tung Sol reissue 12AX7s in the PI spots. Why would V9 and V11 be fine, basically the 40mA I originally set, and V8 and V10 be down in the 25mA range?

    I had plenty of range in the 10K bias pot when the EH7591s were in, IIRC. A bias adjust board (either IBAM or IBBA) will get me some more range for adjustment? I'd sure like to think a totally rebuilt unit would be able to make sufficient bias voltage.
     
  6. larryderouin

    larryderouin Turn it UP, POP? PLLUUEEEZZZZZEE Subscriber

    Messages:
    21,656
    Location:
    Glen Burnie Md.
    Kyle; With the unit plugged into the wall, measure the BIAS voltage at the NEGATIVE lead of the Rectifier. If it's not between -22v and -24v you won't get enough bias voltage (away from Zero). My thinking is if you got matched EH7591's from a store they probably weren't matched for current draw. Most of the stores and other places match for gm and they call that matched. Having said that, current draw could be all over the place with equal gm. And weaker tubes could have been in V8 & V10, allowing V9 and V11 to hog all the current. Jim McShane matches the tubes for current draw at operating voltages. Everyone else uses a tube tester and matches for gm and emissions which in this case is useless.

    The IBAM/IBBA boards will allow a larger range, but you'd have to play around with resistors to get the range you want. The stock configurations are sufficient to get btwn about -24V to about -15V. The IBBA Board is a little harder to build IMO, but you get more accurate results with it.

    Larry
     

     

    Please register to disable this ad.

  7. Jailtime

    Jailtime Standin' on a corner Subscriber

    Messages:
    11,375
    Location:
    Velva ND
    Both sets of 7591s, the EH and Tung Sol, have been from McShane. They were all pretty even for current draw, I think I ran the EH at 34mA.

    I'm away for work for a few days, I'll check that bias voltage at the rectifier though.
     
  8. larryderouin

    larryderouin Turn it UP, POP? PLLUUEEEZZZZZEE Subscriber

    Messages:
    21,656
    Location:
    Glen Burnie Md.
    Ok. With McShane supplied tubes, it sounds like the setting for the BIAS was set too hot. Back them down to between 32 and 34ma. As the Plate Voltage changes the ma will change so do the checks after 15-30 minutes and then check and adjust every hour for the 1st few hours (let it run to check for drift).
     
  9. Jailtime

    Jailtime Standin' on a corner Subscriber

    Messages:
    11,375
    Location:
    Velva ND
    40mA was as cool as I could get the Tung Sols. McShane said that Fishers need what he called "cool biasing" Tung Sols, but we took a chance on a regular set he had in stock, and my 500C hated them.

    I liked the EH better anyway, and they're cheaper. I'll probably just get a new set of those, check the PIs, build an IBBA, and they'll live forever.
     
  10. larryderouin

    larryderouin Turn it UP, POP? PLLUUEEEZZZZZEE Subscriber

    Messages:
    21,656
    Location:
    Glen Burnie Md.
    I don't have any problems with the Tung-Sols in my 800C in the Executive. The Tung-Sols do tend to need more bias voltage to get to a normal bias ma. I'm running them @ 34ma and the bias voltage is between -21v and -22v, while the EH's need about -17v to -18V. I really should change the resistors in the IBAM as the pots are near the end of travel, but I've actually got enough to go to -24v for the Tung-Sols if needed.
     
  11. Jailtime

    Jailtime Standin' on a corner Subscriber

    Messages:
    11,375
    Location:
    Velva ND
    Last edited: Jan 9, 2018

     

    Please register to disable this ad.

  12. larryderouin

    larryderouin Turn it UP, POP? PLLUUEEEZZZZZEE Subscriber

    Messages:
    21,656
    Location:
    Glen Burnie Md.
    Just remember to put the Light Bulb in yer mouth 1st when you have one of those Uncle Fester moments.
     
    UncleBingo likes this.
  13. Jailtime

    Jailtime Standin' on a corner Subscriber

    Messages:
    11,375
    Location:
    Velva ND
    I'm bouncing IBBA build ideas around in my head, any reason I can't just build it in at the output stage instead of a stand alone board? I'm young enough to wrestle a 500C around every once in a while to check bias/current draw.

    Let's see if I have this straight - delete R132 (my 10K bias pot), R131, and R86. Take that -22V from the bias caps to the tabs where the grid resistors tie in with each other. Lift the grid resistors from that tab, add the new 15K IBBA resistors there and 10K DC balance pot. Tie the grid resistors to their respective places on the balance pot, add the 10uF caps from the pot to ground. Would the pin 5 ground on the outputs be suitable for the 10uF caps? Finally, add the 10 or 20K bias pot and 10K resistor from ground to DC balance pot. Whew, I'm pooped. :D
     
  14. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    35,813
    Location:
    Southern NJ
    You can, but finding a place to securely mount the components could be a challenge. Its easy to assemble it on a board just for mechanical reasons. The one I built in my TA-600 is underneath and requires flipping it over to check and adjust stuff.

    Just whatever you do make sure you don't have flying leads and splices in space. Those just ask to fail.
     
  15. Jailtime

    Jailtime Standin' on a corner Subscriber

    Messages:
    11,375
    Location:
    Velva ND
    I'm just about over tubes. I built an IBBA mod in near the output section, but I can only make -13V bias (have good -20.5V coming from the bias cap), and the tubes are pulling over 60mA each. I only ran it for like 30 seconds that way, those numbers are ridiculous. I really got a scare when the bias started to run away more positive, toward -10V. I cut power before it got any further.

    Anybody want to build an IBBA board for me? I'm clearly in over my head.
     
  16. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    35,813
    Location:
    Southern NJ
    exactly what did you do when you tied in the IBBA board? I believe there are some stock components that get replaced with the board. Someone with more specific experience can probably tell you what resistors and such are pulled.
     

     

    Please register to disable this ad.

  17. Jailtime

    Jailtime Standin' on a corner Subscriber

    Messages:
    11,375
    Location:
    Velva ND
    I did exactly what I ran through up in post 13. 15K IBBA resistors to pins 1 and 3 on the DC bal pot. Grid return resistors from pins 1 and 3. 10uF caps from pins 1 and 3 to ground. Pin 2 of DC bal pot tied to pin 2 of 20K bias adjust pot. 10K resistor from pin 3 of bias pot to ground.

    Bias pot seemed to have minimal effect, maybe .5V range total.
     
  18. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    35,813
    Location:
    Southern NJ
    maybe some pics would help? I'm fairly awful understanding verbal descriptions, schematics and pics of actual circuitry does me far better.
     
  19. Jailtime

    Jailtime Standin' on a corner Subscriber

    Messages:
    11,375
    Location:
    Velva ND
    Here you go, quite the mess now that I take a picture of it.

    The brown Dale resistors are the 15K. The light blue ones are grid return. One of the black 10uF Nichicons is grounded to the tab with the cathode resistor. Other one is grounded with the 10K bias pot resistor where you can see that little yellow film cap. You can also see where I tied the pin 2s of the DC bal and bias pots together. Maybe I just can't read a schematic?

    IMG_7835.JPG
     
  20. larryderouin

    larryderouin Turn it UP, POP? PLLUUEEEZZZZZEE Subscriber

    Messages:
    21,656
    Location:
    Glen Burnie Md.
    I think part of the problem is using 2- CL-70's and the subsequent drop in voltage after them. Normal accepted practice now is use of 1 - CL-80. 47ohms start up and 3amp load which is fine for a 500c. I use one in 800c's with no problem. Might give you a volt or 2 higher negative voltage. Even with 1 - CL-80 I get -23 to -25v depending on wall voltage of 121-123. Obviously it will drop and rise with wall voltage, but lets put it on a variac and get a steady state 117vac,( What was the wall voltage when you got 112V in post 11. ) then re-do all the voltages. Then we have a baseline to work with.
     

Share This Page