The thing is...
As Sch3mat1c pointed out, valves don't turn off sharply. so any practical ampliier will be Class AB. We are now in the realms of changing anode load (from Class A to Class B) and ringing from one side of the transformer's leakage inductance when a valve finally does switch off.
Sorry, Joel, but there was a reason why Class B was used for transmitters, PA amplifiers, and the like. Quality wasn't paramount.
As Sch3mat1c pointed out, valves don't turn off sharply. so any practical ampliier will be Class AB. We are now in the realms of changing anode load (from Class A to Class B) and ringing from one side of the transformer's leakage inductance when a valve finally does switch off.
Sorry, Joel, but there was a reason why Class B was used for transmitters, PA amplifiers, and the like. Quality wasn't paramount.
Think I'll go work on Frankenhouse.. right now it's 6V6 PP with like 400V in it.. 
(Name comes from shoehorning a Philco PT into a Westinghouse amp chassis ) It's got fixed bias in it, something like -36V IIRC...<Edit: oh.. 28V > last time I played it I thought I heard a little small-signal distortion.. i.e. x-over distortion. It's pretty much right on the edge, and even so one of the 6V6 plates has a touch of red glow.
Think I'll:
Put regulated screen & driver power in (2 0A2s should work great)
*Variable* bias
Move around the controls (the power switch is between the PT and 5Y3 rectifier!
)
Maybe add a CCS for the LTP
Etc.
Oh- and NFB.
Tim
(Name comes from shoehorning a Philco PT into a Westinghouse amp chassis ) It's got fixed bias in it, something like -36V IIRC...<Edit: oh.. 28V > last time I played it I thought I heard a little small-signal distortion.. i.e. x-over distortion. It's pretty much right on the edge, and even so one of the 6V6 plates has a touch of red glow.Think I'll:
Put regulated screen & driver power in (2 0A2s should work great)
*Variable* bias
Move around the controls (the power switch is between the PT and 5Y3 rectifier!
)Maybe add a CCS for the LTP
Etc.
Oh- and NFB.
Tim
Attachments
> the RCA tube manual calls the typical characteristics of the 6N7 "Class B Power amplifier", yet the tubes are never biased into cut-off in the example they give.
Tubes are awful hard to cut-off, and class B audio designs never try.
You propose class B1 (no grid current). The drive voltage goes way up, and as SY says the linearity is horrid. I have an old Fisher using 7189 tubes. I don't know how Fisher wanted them biased. RCA Manual says they can be run at 17mA idle; yes, but even with my bad ears and a $9 phono preamp it sounded much better idling 40mA, and you could see the warped transfer curve un-bend as you brought the idle up well above RCA's suggested point. Even with substantial NFB.
Then there is AB2 and B2. Grid current.
If you work conventional AB2 or B2, you have no grid current for small signals and large grid current for large signals. The tube input impedance drops very suddently from 100K to 1K. If you run ANY grid current, you have to build a driver capable of feeding a 1K load. And it has to have a super-low output impedance if it must handle the change from 100K to 1K without gross distortion. For lower plate voltages, the driver ends up as a power amplifier not-much smaller than the output stage.
What the 6N7 does is work Class 2 through the full cycle. It is zero-bias, draws a little grid current on both grids at idle. As you swing signal, one grid stops taking current but the other grid takes a little more. The input impedance is very nearly constant. You still need to supply significant grid current, but you do not need an over-sized driver that can handle 100K and 1K equally well: the driver always sees about 1K (times transformer ratio).
The 6N7 with a simple driver gives tolerably low distortion with very low plate current at idle and low-volume, while still being able to play fairly loud when asked to. It is a 1930s Beach Radio. Low power consumption at low listening levels with ample reserve of power is what it is good for.
In wall-powered use, you get an ugly load on the driver and probably high distortion. Since the plate power consumption of an AB or even A amp of the same power will not run up your electric bill, I don't see why you'd use 6N7 type class B for wall power. It makes sense for portable battery operation, but where do you get those HV batteries? And sand-state amps are SO much more efficient, the "sane" thing to do is to use a couple small preamp tubes for flavor and a chip to push the speaker. (Of course sanity is not always the best policy...)
Tubes are awful hard to cut-off, and class B audio designs never try.
You propose class B1 (no grid current). The drive voltage goes way up, and as SY says the linearity is horrid. I have an old Fisher using 7189 tubes. I don't know how Fisher wanted them biased. RCA Manual says they can be run at 17mA idle; yes, but even with my bad ears and a $9 phono preamp it sounded much better idling 40mA, and you could see the warped transfer curve un-bend as you brought the idle up well above RCA's suggested point. Even with substantial NFB.
Then there is AB2 and B2. Grid current.
If you work conventional AB2 or B2, you have no grid current for small signals and large grid current for large signals. The tube input impedance drops very suddently from 100K to 1K. If you run ANY grid current, you have to build a driver capable of feeding a 1K load. And it has to have a super-low output impedance if it must handle the change from 100K to 1K without gross distortion. For lower plate voltages, the driver ends up as a power amplifier not-much smaller than the output stage.
What the 6N7 does is work Class 2 through the full cycle. It is zero-bias, draws a little grid current on both grids at idle. As you swing signal, one grid stops taking current but the other grid takes a little more. The input impedance is very nearly constant. You still need to supply significant grid current, but you do not need an over-sized driver that can handle 100K and 1K equally well: the driver always sees about 1K (times transformer ratio).
The 6N7 with a simple driver gives tolerably low distortion with very low plate current at idle and low-volume, while still being able to play fairly loud when asked to. It is a 1930s Beach Radio. Low power consumption at low listening levels with ample reserve of power is what it is good for.
In wall-powered use, you get an ugly load on the driver and probably high distortion. Since the plate power consumption of an AB or even A amp of the same power will not run up your electric bill, I don't see why you'd use 6N7 type class B for wall power. It makes sense for portable battery operation, but where do you get those HV batteries? And sand-state amps are SO much more efficient, the "sane" thing to do is to use a couple small preamp tubes for flavor and a chip to push the speaker. (Of course sanity is not always the best policy...)
Well, I guess it's a little troubling when someone intrudes into a thread with nothing but negative comments. If anyone doesn't wish to work on a class B amp, they should restrain themselves from posting in this thread.
That seems obvious to me.
If they can't restrain themselves from doing so, then the moderator should. That also seems obvious to me.
That seems obvious to me.
If they can't restrain themselves from doing so, then the moderator should. That also seems obvious to me.
It's got global NFB so it shouldn't have too much distortion. I'll try to take lots of measurements...- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.