I realize these parts are marketed for audio output stages, still they are not very linear below 1A idle current. As can be seen from the datasheet shown above. 
Leading to crossover distortion, (or similar type) which is *not* a friendly form of distortion.
My point is that one should find a clever way of avoiding this crossover distortion, before venturing into this type of MOSFET (as per my first post in this thread).
Leading to crossover distortion, (or similar type) which is *not* a friendly form of distortion.
My point is that one should find a clever way of avoiding this crossover distortion, before venturing into this type of MOSFET (as per my first post in this thread).
h_a said:
Hannes I have to disagree. This caused no end of trouble in the Krill thread. Crossover distortion, as is commonly understood, exists without switching due to the logarithmic relationship of Vbe vs Ic. It is related to the tanh function and its derivative.
EDIT bi-polars of course
Mr. Lars,
Necessarily, the P-channel performs more poorly. It`s extremely hard to design matched complementary FETs. This is the best vertical couple available.I realize these parts are marketed for audio output stages, still they are not very linear below 1A idle current. As can be seen from the datasheet shown above.
Very true, in the case of bipolars in particular. In the case of MOSFETs it is considerably more friendly.
There`s no other way than class A biasing.
Some of you asked for a simplified schematic of the VSOP amp. Here it is, a low power no frills version. It's not something intended to be built, but for analysis purposes only.
- no protections
- no active clamp
- no thermaltrack devices, simple diode multiplier
The AC performance is virtually identical with the full blown amp I have posed above.
- no protections
- no active clamp
- no thermaltrack devices, simple diode multiplier
The AC performance is virtually identical with the full blown amp I have posed above.
An externally hosted image should be here but it was not working when we last tested it.
Lumba Ogir said:
That really was suppose to be a joke. See below.
"Your doctor will also likely prescribe a mood stabilizer for bipolar disorder. Mood stabilizers help control mood swings, prevent recurrences of mood swings, and reduce the risk of suicide. They are usually taken for a long time, commonly for many years. Examples include lithium and certain anticonvulsant drugs."
danieljw said:
Here's the latest measured spec. Things may still slightly change in the final construction, amps of this power levels and performance are of course extremely sensitive to the construction details (case, wiring, etc...):
@Power supply: +/-60V
- Continuous RMS power output into 4ohm: >400W
- Continuous RMS power output into 8ohm: >200W
- Harmonic distortions at full output power into 4ohm, 20KHz, BW=600KHz: <0.0008%
- IM Distortions at full output power into 4ohm, 19+20KHz: <0.0009%
- Bandwidth: 10Hz-50KHz +/-0.1dB
- Phase error: 10Hz-50KHz +/-0.5 degs
- S/N ratio: >106dB ref. 1V input
- Slew rate: 100V/uS non-slewing, rise/fall times determined by the input RF filter
- Damping factor at 20KHz/4ohm: >100, determined by the output coil (1uH).
- Stable in every capacitive load I was able to test, including 2uF/1ohm at full power.
As a side note, I have tested the output stage in isolation (and optimized the bias). The open loop distortions of the output stage at 4ohm, 100W output and 20KHz is 0.05%. Don't you love matched power bipolars? I don't think it's possible to get such numbers using MOSFETs in the output stage. But about other bipolars pros and cons, later...
scott wurcer said:
A couple of questions; What did you consider the output stage in isolation? Depending on where you break it out the source impedance of the stimulus matters more or less.
First Watt THD
See attached. Source impedance is (in my measurements) always 50ohm. Optimal bias was of course around 26mV over the emitter resistors (about 110mA)
Don't have a clue about the First Watt THD
Attachments
Now you have drawn it in isolation, I notice that it's the same type of Vbe multiplier that I first saw on Shinichi Kamijo's site (the one labelled "14").syn08 said:
See attached. Source impedance is (in my measurements) always 50ohm. Optimal bias was of course around 26mV over the emitter resistors (about 110mA)
Don't have a clue about the First Watt THD
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.