Pretty sure I posted a description of this, ah, "conversion" job 8 or 10 years back.
Can't remember where, though. The original pics were lost in a hard drive crash long ago. Only the pics I saved to ImageShack remain. And they lost one of them a few years back.
This was one of those "just because you can doesn't mean you should" jobs. Did it for a friend, knowing it would never be economically worthwhile.
The amplifier started life as a Ken-Rich Eagle 525. A glass factory with seven sweep tubes. Gets a single russky GI-7BT triode installed.
This was the starting point once we removed all the irrelevant stuff. Pretty much everything but the meters, tuning caps, power transformer and fan.
The original transformer gave us about 1900 Volts DC with one of our Pride HV boards reconfigured as a full-wave voltage doubler.
The two fat resistors limit surge current. One of them is between the transformer's HV secondary and the input to the full-wave doubler. Suppresses the turn-on surge. The other one is the "glitch" resistor, in line with the 1900 Volts DC leading to the plate choke.
I wanted to make the cathode bias variable, and used this shunt-regulator setup to simulate a variable Zener diode in the tube's cathode return.
Here's a schemo.
Turns out I was too clever for my own good. The first tube that broke down blew out this circuit completely. Ended up using a series string of 30 rectifier diodes, same as we use for a 3-500Z tube. Much more robust.
It's best to start with a handful of tubes on hand for this kind of project. Testing them to see if there is a risk of breakdown is an iffy proposition, until you power it up in the amplifier and put some fire to it. We only had one of them break down in this box when it got tested.
The meters in this box were both still good. Wanted one of them for plate current and the other one for grid current, rather than the voltmeter scale that's on it.
Turns out the coils in these meters had a pretty high coil resistance. The actual voltage that's needed to drive these meters was too high to just slap two back-to-back diodes across them for surge protection. Had to stack the diodes like this so they wouldn't kick in before the meter shows full deflection.
Lost the pic showing them in place on the meters.
So here's what the final product looked like. I covered the air inlet on the rear panel with sheet metal. The fan on the left is the original exhaust fan. The one on the right blows into the cabinet. This got me enough airflow forced around the socket and through the tube's fins. Added foam around the top of the divider plate to prevent air from spilling around the path across the tube.
The socket came from a canadian seller on Ebay. Nicely made, and not much more expensive than the tube. The original Driver Tune became the input tuning control. This gave it a wide-enough tuning range to cover 10 and 11 meters both with a low input SWR.
Only had to add the one small transformer to power the tube's heater.
The final result wasn't terribly impressive. Would get 350 Watt peaks with a single-final 40-channel radio, 500 with a dual-final base like a 2517 or Saturn.
Only got 800 Watt peaks with 100 Watts peak drive. A graph showing input to output power would nto be a straight line. Pretty well disqualifies it as a "linear" in that sense. Don't remember anyone referring to an amplifier like this as a "curvear". Would be more accurate, though.
The owner was happy with the result, but it's truly one of a kind. Haven't tried this cornball trick since.
Just because you can......
73
Can't remember where, though. The original pics were lost in a hard drive crash long ago. Only the pics I saved to ImageShack remain. And they lost one of them a few years back.
This was one of those "just because you can doesn't mean you should" jobs. Did it for a friend, knowing it would never be economically worthwhile.
The amplifier started life as a Ken-Rich Eagle 525. A glass factory with seven sweep tubes. Gets a single russky GI-7BT triode installed.
This was the starting point once we removed all the irrelevant stuff. Pretty much everything but the meters, tuning caps, power transformer and fan.
The original transformer gave us about 1900 Volts DC with one of our Pride HV boards reconfigured as a full-wave voltage doubler.
The two fat resistors limit surge current. One of them is between the transformer's HV secondary and the input to the full-wave doubler. Suppresses the turn-on surge. The other one is the "glitch" resistor, in line with the 1900 Volts DC leading to the plate choke.
I wanted to make the cathode bias variable, and used this shunt-regulator setup to simulate a variable Zener diode in the tube's cathode return.
Here's a schemo.
Turns out I was too clever for my own good. The first tube that broke down blew out this circuit completely. Ended up using a series string of 30 rectifier diodes, same as we use for a 3-500Z tube. Much more robust.
It's best to start with a handful of tubes on hand for this kind of project. Testing them to see if there is a risk of breakdown is an iffy proposition, until you power it up in the amplifier and put some fire to it. We only had one of them break down in this box when it got tested.
The meters in this box were both still good. Wanted one of them for plate current and the other one for grid current, rather than the voltmeter scale that's on it.
Turns out the coils in these meters had a pretty high coil resistance. The actual voltage that's needed to drive these meters was too high to just slap two back-to-back diodes across them for surge protection. Had to stack the diodes like this so they wouldn't kick in before the meter shows full deflection.
Lost the pic showing them in place on the meters.
So here's what the final product looked like. I covered the air inlet on the rear panel with sheet metal. The fan on the left is the original exhaust fan. The one on the right blows into the cabinet. This got me enough airflow forced around the socket and through the tube's fins. Added foam around the top of the divider plate to prevent air from spilling around the path across the tube.
The socket came from a canadian seller on Ebay. Nicely made, and not much more expensive than the tube. The original Driver Tune became the input tuning control. This gave it a wide-enough tuning range to cover 10 and 11 meters both with a low input SWR.
Only had to add the one small transformer to power the tube's heater.
The final result wasn't terribly impressive. Would get 350 Watt peaks with a single-final 40-channel radio, 500 with a dual-final base like a 2517 or Saturn.
Only got 800 Watt peaks with 100 Watts peak drive. A graph showing input to output power would nto be a straight line. Pretty well disqualifies it as a "linear" in that sense. Don't remember anyone referring to an amplifier like this as a "curvear". Would be more accurate, though.
The owner was happy with the result, but it's truly one of a kind. Haven't tried this cornball trick since.
Just because you can......
73
Last edited:
