6CL6 was the common driver for a pair of parallel 6146 in linear RF service.
Still cheap, and still widely available.
12BY7 was also used. They seem to be less cheap, and less common, these days.
Both are video amplifiers. They have to be very linear, AFAIK, or the TV picture won't look right. Big names in industry used them in linear amps for a reason.
Still cheap, and still widely available.
12BY7 was also used. They seem to be less cheap, and less common, these days.
Both are video amplifiers. They have to be very linear, AFAIK, or the TV picture won't look right. Big names in industry used them in linear amps for a reason.
The driver issue notwithstanding, it's curious that a schematic drawn in 2003 specifies 6146 or 6146A.
The B may ( or may not ) interchange with the earlier tubes in RF service, but for a scratch built audio amp I would go with the B's. The Shuguang 6146B works pretty decent at AF and seems to have a bit more plate dissipation than NOS 6146B's. It also has a visible heater, so the cosmetics are a bit better for an audio amp. They have squared off pins instead of tapered, so I would NOT stick one in a vintage socket or piece of equipment, but they are fine in the cheap sockets currently available.
I used 100K grid resistors when I used 6146, and didn't seem to have any issues. 200 volts is OK for the screens - I guess this is fixed bias and we're not seeing that supply. I ran mine cathode biased.
The B may ( or may not ) interchange with the earlier tubes in RF service, but for a scratch built audio amp I would go with the B's. The Shuguang 6146B works pretty decent at AF and seems to have a bit more plate dissipation than NOS 6146B's. It also has a visible heater, so the cosmetics are a bit better for an audio amp. They have squared off pins instead of tapered, so I would NOT stick one in a vintage socket or piece of equipment, but they are fine in the cheap sockets currently available.
I used 100K grid resistors when I used 6146, and didn't seem to have any issues. 200 volts is OK for the screens - I guess this is fixed bias and we're not seeing that supply. I ran mine cathode biased.
I see 150 Watt annotated on the OT!! The 600V B+ is right at the max spec for 6146x.
Might as well use a pair of 30 or 33 or 40 Watt TV Sweep tubes instead of a quad of 6146x, fewer bias adjustments and probably longer life.
Datasheet Pdiss for 6146x are given as absolute max.,
design max would be more like 15W for 6146A and 20 W for 6146B.
Might as well use a pair of 30 or 33 or 40 Watt TV Sweep tubes instead of a quad of 6146x, fewer bias adjustments and probably longer life.
Datasheet Pdiss for 6146x are given as absolute max.,
design max would be more like 15W for 6146A and 20 W for 6146B.
Last edited:
Smoking-amp, thank you for your input, actually, my plan was to use sweep tubes, for some odd reason I fell in love with the idea of 6CL6's or equivalent tubes driving the push-pulls. With the 6CL6s and the 7054's being close, would you start with the same component values?
Thanks
Ray
Thanks
Ray
Last edited:
The 7054 has higher gm than the 6CL6 for the same current, about 1.3X. So I would increase the 80 Ohm cathode resistors to 100 Ohms as a 1st guess to get the same driver stage current.
The Sweep tubes will only need about -30V to turn them off versus -50V for the 6146x tubes (depending on the Vg2 used, assuming 75% to %60 lower Vg2 for the Sweeps). You may be able to reduce the driver current a little further and increase the 8200 Ohm driver loads some, so less heat developed there. Could try 120 Ohms for the driver cathode R and 10K for the driver load R, see if the driver is developing enough swing and is centered in the available driver V range.
The Sweep tubes will only need about -30V to turn them off versus -50V for the 6146x tubes (depending on the Vg2 used, assuming 75% to %60 lower Vg2 for the Sweeps). You may be able to reduce the driver current a little further and increase the 8200 Ohm driver loads some, so less heat developed there. Could try 120 Ohms for the driver cathode R and 10K for the driver load R, see if the driver is developing enough swing and is centered in the available driver V range.
Last edited:
The Sweep tube grid 1 bias resistors can be increased above 22K, more like 100K.
That allows increasing the driver load R some more, with a similar reduction in driver current.
Maybe go to 20K for the driver loads now, with 180 Ohm driver cathode R (for bias ). Oh yeah, the 330 Ohm driver tail R would need to be increased similarly to 750 Ohm or 820 Ohm. An adjustable CCS tail from -Vbias would allow setting the driver stage current better, with lower distortion too.
That allows increasing the driver load R some more, with a similar reduction in driver current.
Maybe go to 20K for the driver loads now, with 180 Ohm driver cathode R (for bias ). Oh yeah, the 330 Ohm driver tail R would need to be increased similarly to 750 Ohm or 820 Ohm. An adjustable CCS tail from -Vbias would allow setting the driver stage current better, with lower distortion too.
Last edited:
Ampeg SVT bass-amp, early production, used a sextet of 6146 and delivered AMPLE bass. It is a truly frightening beast.
https://el34world.com/charts/Schematics/files/Ampeg/Ampeg_svt_model_6146b.pdf
After a few years Ampeg switched to selected 6550, and these are arguably more reliable.
https://el34world.com/charts/Schematics/files/Ampeg/Ampeg_svt_model_6146b.pdf
After a few years Ampeg switched to selected 6550, and these are arguably more reliable.
With the RCA 6146x datasheets listing absolute max Pdiss (20 or 27 Watt) instead of the design max, I guess there was some tendency to push the limit in designs. The 6146 plate looks similar in size to a 6JN6/6GE5 Sweep tube (17.5 Watt).
The 6146 datasheet lists 30,000 Ohms (instead of 0.5 meg Ohms for class AB1 ) for a class AB2 grid 1 bias resistor. Since the schematic has 25000 Ohm grid 1 resistors, and around 150 Watt output (OT rating), this design must have been running class AB2. One would expect a grid 1 cooler fin up top for class AB2, but the 6146 has an RF shielding box up top instead. The grids probably get really hot in the 6146 tubes, and then runaway....
For running class AB1, the driver load resistors could be increased even further beyond 20K. (with similar changes to the driver cathode R and tail R) One needs to check the input capacitance of the Sweep tube versus the doubled 6146 to see how far that can go before HF bandwidth is an issue.
With a single set of 6JN6/6GE5 I would think 50 Watts out max would be more typical in class AB1. Two sets, 100 Watts. With some big TV Sweeps (26LX6, 36MC6, 26LW6...), maybe 150 Watts in class AB1. Class B could get more. Tubelab I think has gotten up to near 250 Watts with the _LW6 at 70% plate efficiency.
The 6146 datasheet lists 30,000 Ohms (instead of 0.5 meg Ohms for class AB1 ) for a class AB2 grid 1 bias resistor. Since the schematic has 25000 Ohm grid 1 resistors, and around 150 Watt output (OT rating), this design must have been running class AB2. One would expect a grid 1 cooler fin up top for class AB2, but the 6146 has an RF shielding box up top instead. The grids probably get really hot in the 6146 tubes, and then runaway....
For running class AB1, the driver load resistors could be increased even further beyond 20K. (with similar changes to the driver cathode R and tail R) One needs to check the input capacitance of the Sweep tube versus the doubled 6146 to see how far that can go before HF bandwidth is an issue.
With a single set of 6JN6/6GE5 I would think 50 Watts out max would be more typical in class AB1. Two sets, 100 Watts. With some big TV Sweeps (26LX6, 36MC6, 26LW6...), maybe 150 Watts in class AB1. Class B could get more. Tubelab I think has gotten up to near 250 Watts with the _LW6 at 70% plate efficiency.
Last edited:
For the original circuit, if you are going to be using say an Edcor 100 Watt 4K OT, there is no need to keep the B+ up at 600V, that will just heat the tubes un-necessarily.
500V B+ will drop the power output level to 100 Watts, and 450V B+ will drop the power level to 80 Watts, with cheaper capacitors for the power supply. A little safer to work on too.
Shifting to a 3.3K 100 Watt OT would want around 460V B+ for 100 Watt output. Not counting OT losses in these estimates.
You should not need to go grid positive (AB2) with big TV Sweeps for a 100 Watt power output level.
500V B+ will drop the power output level to 100 Watts, and 450V B+ will drop the power level to 80 Watts, with cheaper capacitors for the power supply. A little safer to work on too.
Shifting to a 3.3K 100 Watt OT would want around 460V B+ for 100 Watt output. Not counting OT losses in these estimates.
You should not need to go grid positive (AB2) with big TV Sweeps for a 100 Watt power output level.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.