Strangely enough, I'm not worried at all - and I'm not considered "resistant to advice". So what's going on?
1. the distortion factor
At -90dB to -80dB, i.e. 0.0032% to 0.01%, it is absolutely realistic and nothing else can be expected from this type of circuit.
So check it off!
2. the risk of explosion or time constants
Well, now it's going to be exciting, which adjustment screw will be turned to remedy the situation? In addition to the process of finding the most favorable dimensioning of the components involved, this is also an issue for the design of the power supply unit. Off the top of my head, I can think of PSUs that could have an active current limiter - let's say 50mAdc, so we're prepared for all eventualities.
So ticked off!
3. The control of the quiescent current by Q1 & Q2 - and the transfer distortion
Well, that's the biggest problem and
not really solved yet, is it?
4. the signal-to-noise ratio
It can't be particularly outstanding, and that doesn't surprise me. You could now extract it from @Bernhard's curves (THD(Po) & THD+N(Po)) or load the model in MC12 ... but I'm not interested in that right now because it's unlikely that the amp is noisy.
We can seemingly check off this point!
#
To sum up, I find that the L.Stellema rescue program is challenging. To sum up, I find that the L.Stellema rescue program is challenging. The solution to a weakness generates a new one.
Maybe I'll be smarter in 14 days,
HBt.
1. the distortion factor
At -90dB to -80dB, i.e. 0.0032% to 0.01%, it is absolutely realistic and nothing else can be expected from this type of circuit.
So check it off!
2. the risk of explosion or time constants
Well, now it's going to be exciting, which adjustment screw will be turned to remedy the situation? In addition to the process of finding the most favorable dimensioning of the components involved, this is also an issue for the design of the power supply unit. Off the top of my head, I can think of PSUs that could have an active current limiter - let's say 50mAdc, so we're prepared for all eventualities.
So ticked off!
3. The control of the quiescent current by Q1 & Q2 - and the transfer distortion
Well, that's the biggest problem and
not really solved yet, is it?
4. the signal-to-noise ratio
It can't be particularly outstanding, and that doesn't surprise me. You could now extract it from @Bernhard's curves (THD(Po) & THD+N(Po)) or load the model in MC12 ... but I'm not interested in that right now because it's unlikely that the amp is noisy.
We can seemingly check off this point!
#
To sum up, I find that the L.Stellema rescue program is challenging. To sum up, I find that the L.Stellema rescue program is challenging. The solution to a weakness generates a new one.
Maybe I'll be smarter in 14 days,
HBt.
Your hobbyhorse is the THD - and therefore the underlying cause is of essential importance in every new iteration.
For example:
You would like to know why (causally) a resulting characteristic value suddenly becomes larger or smaller, better or worse. I like to refer to this complex as the "adjusting screws". In other words, you have to know what you are doing.
For example:
You would like to know why (causally) a resulting characteristic value suddenly becomes larger or smaller, better or worse. I like to refer to this complex as the "adjusting screws". In other words, you have to know what you are doing.
I not only fixed the distortion, but also the noise, the input configuration and the AC coupled VAS.
The original high impedance voltage divider wasn´t a good idea... to start with.
I can not agree.
Attached .cir to play with
Prizes are not awarded
Dear Bernhard,
you and I come from completely different warrior tribes and castes
.I suggest a break
.Bye,
HBt.
My mistake:
The apparent statement was actually meant as a global headline.
My line of attack:
I stay and deal with the last circuit /schematic, see posting #306 among others. Why? Because I don't want to stray any further (than I already have) from the original proposal of the article.
With that in mind, let's all stick with this topic for now. Conversations are welcome via pm or in the lounge. From now on, it's the construction that counts, not the keyboard - sorry.
kind regards,
HBt.
The apparent statement was actually meant as a global headline.
My line of attack:
I stay and deal with the last circuit /schematic, see posting #306 among others. Why? Because I don't want to stray any further (than I already have) from the original proposal of the article.
With that in mind, let's all stick with this topic for now. Conversations are welcome via pm or in the lounge. From now on, it's the construction that counts, not the keyboard - sorry.
kind regards,
HBt.
I installed the demo version of MicroCap and loaded your cir.
There was a transistor model missing and after some attempts
I gave up doing a transient analysis as this procedure
is way to complicated for me compared to LTSpice.
It is a pity that by using exotic programs most of the forists are excluded.
So I will finalize my objections and leave this thread.
The current flow through the VAS is extremely dependent from supply voltage due to
biasing with to 24V Zeners. Which, btw have a nomimal tolerance of 5 or 10%.
Accordingly power supply rejection will be subtimal.
Will you use a ultrastable power supply - that is adjusted according to the amp and temperature?
Certainly not.
This type of driving a VAS is a call for problems that are already solved with less exotic attempts.
These are exactly the faults undiscovered by simulation
with its inherently super stable virtual conditions -
that will collide the hard way with reality.
There was a transistor model missing and after some attempts
I gave up doing a transient analysis as this procedure
is way to complicated for me compared to LTSpice.
It is a pity that by using exotic programs most of the forists are excluded.
So I will finalize my objections and leave this thread.
The current flow through the VAS is extremely dependent from supply voltage due to
biasing with to 24V Zeners. Which, btw have a nomimal tolerance of 5 or 10%.
Accordingly power supply rejection will be subtimal.
Will you use a ultrastable power supply - that is adjusted according to the amp and temperature?
Certainly not.
This type of driving a VAS is a call for problems that are already solved with less exotic attempts.
These are exactly the faults undiscovered by simulation
with its inherently super stable virtual conditions -
that will collide the hard way with reality.
Last edited:
@bucks bunny
Let me give you a hand getting started.
Please install Micro-Cap 12 full version, if you haven't already.
If you can't find MC12 for download I can help you, just let me know. This excellent simulator is now FREE!
Then download the CIR below and run Transient, Dynamic DC, AC, or Stability (from the Analysis menu). You can run these four types of analysis without bumping into any issues with the CIR file I'm attaching (latest schematic from Bernhard).
Let me give you a hand getting started.
Please install Micro-Cap 12 full version, if you haven't already.
If you can't find MC12 for download I can help you, just let me know. This excellent simulator is now FREE!
Then download the CIR below and run Transient, Dynamic DC, AC, or Stability (from the Analysis menu). You can run these four types of analysis without bumping into any issues with the CIR file I'm attaching (latest schematic from Bernhard).
Last edited: