I was playing in Tube CAD and came up with this White cathode follower. I seem to recall that tubes are happiest with a load of 10x the circuit's Zout. According to Tube CAD the Zout for this circuit is 3.44 ohms, almost 1/10 the 32 ohm impedance of my headphones.
Since my CD source puts out between 1 & 2 volts and the gain of this circuit is .97 and plate dissipation is 989mW it seems that it should drive my headphones to very loud levels quite handily.
What am I missing?
Since my CD source puts out between 1 & 2 volts and the gain of this circuit is .97 and plate dissipation is 989mW it seems that it should drive my headphones to very loud levels quite handily.
What am I missing?
Can't say whether or not you are missing something or not as you didn't post your white follower based design or even mention what tube type, quiescent current or voltage across the follower you intend to employ.
That said I designed and built a headphone amp using white followers driven by an SRPP driver stage. I wanted sufficient gain to use any line level source and to be able to drive it with any level from a couple of hundred mV upwards.
My white cathode follower uses a pair of 5842 running at approximately 20mA per channel with a total supply voltage of 210V. Idle dissipation is ~4W per channel. (2W per tube)
It will comfortably drive a 32 ohm load, but IIRC the actual source impedance is much higher than what your theoretical model might suggest for my circuit. I don't remember the exact source impedance of my design, but I believe it is >32 ohms. (My usual headphones are 64 ohm Sony's.)
No feedback is employed, the lower tube is fixed bias to avoid the need for a huge cathode bypass cap. The upper tube is direct coupled to the 5687 SRPP (10mA), and despite careful calculation and modeling, the plate resistor in the white CF ended up being quite a bit higher than the model predicted or I independently calculated (About 2 times higher) for good drive symmetry.
I used a 360V/220uF photoflash cap as the output capacitor, but I really don't recommend this as the safest approach because if the cap fails you might end up with half VCC across the transducer in your headphone. I think an autoformer would provide a more benign load and a much lower dcr in the event of a fault, keeping dc out of your phones. (The ear cups in my phones are all plastic construction and the transducers are more than 1/2" from my ears, still I would be cautious.)
Various edits to correct typos, refine thoughts on re-read.
That said I designed and built a headphone amp using white followers driven by an SRPP driver stage. I wanted sufficient gain to use any line level source and to be able to drive it with any level from a couple of hundred mV upwards.
My white cathode follower uses a pair of 5842 running at approximately 20mA per channel with a total supply voltage of 210V. Idle dissipation is ~4W per channel. (2W per tube)
It will comfortably drive a 32 ohm load, but IIRC the actual source impedance is much higher than what your theoretical model might suggest for my circuit. I don't remember the exact source impedance of my design, but I believe it is >32 ohms. (My usual headphones are 64 ohm Sony's.)
No feedback is employed, the lower tube is fixed bias to avoid the need for a huge cathode bypass cap. The upper tube is direct coupled to the 5687 SRPP (10mA), and despite careful calculation and modeling, the plate resistor in the white CF ended up being quite a bit higher than the model predicted or I independently calculated (About 2 times higher) for good drive symmetry.
I used a 360V/220uF photoflash cap as the output capacitor, but I really don't recommend this as the safest approach because if the cap fails you might end up with half VCC across the transducer in your headphone. I think an autoformer would provide a more benign load and a much lower dcr in the event of a fault, keeping dc out of your phones. (The ear cups in my phones are all plastic construction and the transducers are more than 1/2" from my ears, still I would be cautious.)Various edits to correct typos, refine thoughts on re-read.
No bang at switch on or insertion of phones into jack. I use 1K resistors to ground and the output to provide charging path for output caps when the phones are not connected.
The photoflash have low ESR and ESL which is why I used them in the first place, sound quite decent. HV Cerafines would be better I'm sure if obtainable.
The photoflash have low ESR and ESL which is why I used them in the first place, sound quite decent. HV Cerafines would be better I'm sure if obtainable.
kevinkr said:
Oops,
I thought the schematic post went OK. Basically it was a text book White cathode follower design using 6922 tubes.
I built the circuit last week and found the output to be severely lacking. Bass was non-existent and volume was too low. I added a grounded cathode stage in front of the White follower with a gain of about 24 and volume became usable. However the bass was still not there. I changed the output coupling caps and got the bass along with a bunch of distortion.
Bottom line- I'm first going to build someone else's proven design. I'm looking at the Optimized Morgan Jones unit on Headwize. I've got all the bits so I just need to get to it!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.