• You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.

Using mismatched sweep tubes in amplifiers

wow, thanks for that nomad!

i've read it once, and im going to read it again tomorrow so i can actually start to get it.

what you have given me here is a way to understand some of the things i'll be reading about in the ARRL handbooks of yore that i have around here.
(i like the editions before the 90's much more than the new ones)
LC
 
  • Like
Reactions: Tokin
Individual bias usually only allows you to use unmatched parts from the same manufacturer and preferably from the same lot number. Gain at DC and gain at HF are two entirely different things due to internal element size, spacing and capacitance. Don't just assume that because DC bias voltage has been set differently to achieve the same idling current, that the full operating current under RF load will still be balanced. That should be measured and confirmed just like the idling current was. If a tube was drawing low idling current because it was weak and you turn up the bias current, it still will not do its share of the workload.

Remember, this modification didn't offset the RF drive voltage like it did with the DC bias to reach balance. Balancing only the DC current and ignoring the RF drive voltages to the tube will not allow you to do something like run a GE 6LF6 with a Sylvania 6LF6. The internal differences within the tubes are enough that the harder you drive them with RF, the more out of balance they will run, even if they start out at the same idling current. The more the DC bias voltage has to be offset to obtain equal bias current, the more this problem will become an issue.

PS: I can make a weak tube draw normal idling current by adjusting its bias voltage but it will still NOT do its share of the workload when amplifying RF due to a lack of peak emissions.
 
Last edited:
Shockwave makes a good point. The individual bias pots will allow you to balance the idle current of each tube with no drive power.


But it won't affect how the tubes balance at full drive power. If one tube is 'soft' and fails to pull full current, the remaining load stress is thrown onto the other tubes. Worst case is one strong tube and three weak tubes. The strong one will pull more than its one-fourth share of the load current, overheat and fail prematurely, while the other three weak tubes simply 'loaf' carrying too little current to overheat.

How the tubes share the load under full drive is much more important than how they balance with no drive.

73
 
  • Like
Reactions: Shadetree Mechanic
Good read, lots of information here. So it sounds like the drive would need to be set individually going to each tube while driven into a load at desired output level. How would the output for each tube be measured individually to do this? Or would measuring grid current individually work?
 
How would the output for each tube be measured individually to do this? Or would measuring grid current individually work?

The current each tube draws under full output is measured in the same place as the individual bias current. The schematic in the first post shows the metering installed in the B+, 900 volt plate line. I prefer to install current meters in the cathode return circuit for safety reasons. Then you don't have 900 volts right behind the meter glass (or plastic). Metering the cathode will show plate and grid current combined but since grid current is many times lower than plate, you are essentially still measuring plate current. Grid current is more useful for tuning than it is at confirming a balanced output between tubes.
 
  • Like
Reactions: Shadetree Mechanic

dxChat
Help Users
  • No one is chatting at the moment.
  • @ Wildcat27:
    Hello I have a old school 2950 receives great on all modes and transmits great on AM but no transmit on SSB. Does anyone have any idea?
  • @ ButtFuzz:
    Good evening from Sunny Salem! What’s shaking?
  • dxBot:
    63Sprint has left the room.