No. Slew rate limiting is not caused by negative feedback.dcbingaman said:
No. Earlier triodes were generally more linear, because their electrode structure was more nearly self-similar (e.g. all cylindrical). Most people regard the 12AU7 as one of the most nonlinear triodes ever invented.
No. Devices do not have slew rate limitations; circuits have slew rate limitations. Device bandwidth is not a problem for audio; maybe you are getting confused with UHF/microwave circuit design?
Class AB solid-state amps usually use a BJT follower output stage, biased at an optimum point to minimize crossover distortion. However, there is no simple way to bias the devices to make a smooth, constant transconductance throughout the waveform. Biasing hotter just makes gm doubling distortion as the devices transition from both conducting to only one conducting. It is well-known that this is the main source of distortion for standard-topology solid-state amps. The only way to completely avoid it is going Class-A.
NFB decreases the crossover distortion, but since it is the dominant source of distortion in the amplifier it will always stick out.
I believe that Douglas Self had a solution that injects current into the output stage that offsets the crossover point so that very low level signals can have much lower distortion. "Crossover Displacement," he called it.
And it is not the whole story. Bias voltage is controlled by Vbe multiplier that senses temperature of semiconductor dies through layers of metals, i.e. with phase shifts, so especially in class AB amps bias follows loudness of the music with phase shift, so when sounds decay spectrum of distortions gets wider and THD increase. It is contrary to how our own perception distorts and ignores, so even though a Blameless Amplifier has Optimal Bias and pretty low THD on a nominal power, it sounds worse than a simple SE amp with transmitting triode output that has on a nominal power higher THD, but when sounds decay distortions decay faster, and their order gets lower. Such distortions are ignored by our perception that got used to ignoring it's own distortions that are similar, so the amp sounds pristine clean. It sounds as if "louder" than it does.
Monotonicity as used in audio amps means that distortion decreases as signal level decreases, at all signal levels. This is actually pretty hard to do, requiring Class A operation of the valves and bias current in the output transformer. It's wasteful of electricity, but entropy must have its pound of flesh.
All good fortune,
Chris
All good fortune,
Chris
Yes true I like it.
But we, first, have to calculate Dynamic capacitance (for instance), to detemine the driver for the output tube?
Howzat?
Just make the driver such low Z that nothing else matters. My 845s are transformer driven from 300Bs.
Yes true I like it.
But we, first, have to calculate Dynamic capacitance (for instance), to detemine the driver for the output tube?
Just make the driver such low Z that nothing else matters. My 845s are transformer driven from 300Bs.
Member
Joined 2009
Paid Member
Sounds like Wretched Excess, and there's nothing I like better than Wretched Excess.
Reminds me that many DIY'ers might already have a suitable driver stage for a type 845 at home. Practically any existing single-ended-output amplifier using type EL84 or larger valve needs only a high voltage coupling cap from valve anode to added output RCA jack, safety pull-down resistor at jack, and a Zobel for valve loading. Instant driver stage with plenty of juice. 845 and its power supply in a new chassis.
All good fortune,
Chris
Last edited:
Member
Joined 2009
Paid Member
some DIY'ers, in this case the late Sakuma-san, do have a suitable driver stage at home - another 845
Attachments
look how at much grid current 845 develops, it needs a good driver.
if you don't sink the current, it loads the stage, reduces power, and heavy distortion in the 20s Watts, if you sink the current properly, drive it with a IT and such as 2A3, you can get almost 30Watts into 8 ohm with a 845, running in A2, beautifully linear. around 1% thd.
Some designs have 10%thd at 20 watts !!! unacceptable.
the IT is non-essential, you can use a regulated supply for the bias, with a current sensor on the cathode which augments the bias and protects the valve from self destruction.
Self-destruction possibility and good sound are the results of proper driving of the 845 tube.
if you don't sink the current, it loads the stage, reduces power, and heavy distortion in the 20s Watts, if you sink the current properly, drive it with a IT and such as 2A3, you can get almost 30Watts into 8 ohm with a 845, running in A2, beautifully linear. around 1% thd.
Some designs have 10%thd at 20 watts !!! unacceptable.
the IT is non-essential, you can use a regulated supply for the bias, with a current sensor on the cathode which augments the bias and protects the valve from self destruction.
Self-destruction possibility and good sound are the results of proper driving of the 845 tube.
Here I revealed all details one by one, while helped others, but the complete compiled schematic posted nowhere, it is a commercial project. I sold few prototypes to music lowers, though.
Edelweiss-3 is a budget project, for people who can not afford usual high-end with whistles and bells, but want high end sound quality. The line will include: basic power amplifier, integrated power amplifier, power amplifier for high end magnetic planar headphones, preamplifier, and a preamplifier with MM/MC inputs.
While working on design I build prototypes with different features that may go, may not to the production, and sell them after they are thoroughly tested.
Basic #Edelweiss-3 power amplifier:
Input: 0 dB (0.775V) RCA, 50K stereo
Outputs: low ohm for 4-6 Ohm and high ohm, for 8-12 Ohm load.
Power: at least 6W for <1% total harmonic distortions.
Soft asymmetric saturation (inaudible clipping),
negative output resistance for damping of woofers (servo-control of speakers).
Bandwidth: 10 Hz-90 KHz -3 dB, from 20 Hz full power.
Available versions with different output tubes:
EL34/KT77
5881/6П3С-Е
Raytheon VT127
4П10С
The driver and preamp tube is 8CB11, a twin video pentode. It's filament is powered from rectified DC for zero audible hum.
Voltages are stabilized by VR tube and MOSFET. It gave the name "Edelweiss", a nice flower living on hard rocks.
Voltage regulator tube: VR75 or VR105 (orange glow), or VR150 (purple glow), depending on output tubes.
I am not sure that amps in production would cost $800. There will be expenses added for distribution, support, marketing, certification.
The principles of Wretched Excess require cathode resistor bias for type 845. 160 DC Volts and 80mA. Works great, but use thermal grease to heatsink it to chassis. Plant a few extra trees if you're concerned about your criminality.
All good fortune,
Chris
Use of MOSFETs is allowed as long as one's heart is pure, and everyone knows yours is.
Always the best,
Chris
Wicked, wicked. Also brings back that element of danger to valve health. Just be aware that many modern religions forbid blue LED's. So there's that.
All good fortune,
Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.