If I can join, Hugh, I would say no distortion pattern (in isolation) is able to make sound more 'pleasing' or 'better'. There are many possibilities how to implement exactly defined distortion on wav data (Howard e.g.). I am pretty sure that distortion itself is a wrong track, and that amplifier sound is affected more by many other influences.
PMA,
Thank you for your post. Interesting comment. If indeed distortion is not the right path to pursue (and many would disagree with good reason) then what are those other influences? Surely they too must be distortive? And if they are not, what precisely are they? What is the mechanism?
This is a vexing question....
Hugh
Thank you for your post. Interesting comment. If indeed distortion is not the right path to pursue (and many would disagree with good reason) then what are those other influences? Surely they too must be distortive? And if they are not, what precisely are they? What is the mechanism?
This is a vexing question....
Hugh
andy
That is not fair.
You can like or dislike the design, but saying that steve has been lying is not right.
He has even tried to learn LTspice to do the same simulations as you and a lot of other do.
He made some mistakes, and admitted that, but if you are saying that he has been lying about his results you must really be some kind of a guy.
Mr AKSA (Hugh)
I don’t think Steve needs your”help” here. He is a respected member.
You actually remind me of another “helping hand” in another thread.
BTW the circuit and explanation you have on Steve’s web page is not right.
If you like to discuss, you have your own forum, bring it up there instead of hijacking here.
Cheers
Well said Andy.
And attached below is a my previous 4-transistor pair Krill simulation, but this time with a 20 ohms bias resistor for the driver stage.
It is also worth pointing out that a 220 ohm biasing resistor is far to small for any power output stage anyway - that will only bias the drivers with 6mA or so, which is far too low to achieve an adequately fast switch off of the output devices that is necessary for the lowest possible distortion. D.Self even recommends bypassing this resistor with a large cap to further speed up the switching.
Anyway, getting back to the attaced simulation; in the first attached picture the green trace shows the ballast resistor current for the complete Krill output stage which has been modified with a 20 ohm bias resistor for driver transistors.
The red trace, for comparison, shows the ballast resistor current of the non-krill output stage with ideal voltage source bias and no bias resistor for the drivers.
The green trace of the second attached screen shot shows the Krill output stage ballast resistor current on a much smaller scale to show that it does indeed switch completely when the drivers are adequately biased.
The current drops down to about 80nA - which is just leakage or junction capacitance current (in a seperate sim I found that, under DC operating conditions with comparable supply rails, the output transistor model generates an off collector leakage current of about 90nA).
And as before, the complete LTspice simulation file is attached for anyone to download and run for themselves
Attachments
Last edited:
Nope Stinius,
Utterly refuted. I asked Syn08 a reasonable question, I'd like a reply if he is willing. This is neither hijacking the thread, nor is it in defence of Steve, who is quite capable of defending himself, though Andy's accusation is reprehensible.
Besides, is not your intervention uncalled for, as it was a question for another?
I see GK is routinely defending his pal here too, care to comment on that?
If you would like to 'correct' my explanation of Steve's circuit, perhaps you could send it to me in a PM. I will certainly give it my time.
Utterly refuted. I asked Syn08 a reasonable question, I'd like a reply if he is willing. This is neither hijacking the thread, nor is it in defence of Steve, who is quite capable of defending himself, though Andy's accusation is reprehensible.
Besides, is not your intervention uncalled for, as it was a question for another?
I see GK is routinely defending his pal here too, care to comment on that?
If you would like to 'correct' my explanation of Steve's circuit, perhaps you could send it to me in a PM. I will certainly give it my time.
andy
That is not fair.
You can like or dislike the design, but saying that steve has been lying is not right.
He has even tried to learn LTspice to do the same simulations as you and a lot of other do.
He made some mistakes, and admitted that, but if you are saying that he has been lying about his results you must really be some kind of a guy.
Excuse me?
Compare the (contrary to reality) claim made in post 208:
http://www.diyaudio.com/forums/solid-state/151295-krill-next-generation-21.html#post2010027
With the facts of the matter provided by someone who actually posted evidence of having run a simulation with the driver biasing resistor, instead of just saying that they did:
http://www.diyaudio.com/forums/solid-state/151295-krill-next-generation-21.html#post2010053
This thread has once again moved on to feigned moral outrage by one wishing to evade the basic technical facts of the matter – that is what is “despicable” in the whole charade.
Last edited:
Hugh there is no need for this acrimony, just step back for a moment and think about it. An output stage is sitting at say an amp into 4 Ohms. It matters weather it was at 4 amps 10 usec before or not. Past history is important. Any description that walks through a circuit in a DC matter does not describe the behavior in a dynamic sense. I'm trying to make you think about the misunderstanding without being confrontational.
Last edited:
I believe you are referring to me, stinius, so who appointed you as the moderator of this thread - what gave you the notion that you can tell Hugh to post on his own forum or otherwise police this thread? Why not refer to the "helping hands" that I see posting in "group" mode again here as in the other thread. I see the venom has started again - I'm backing out of the snake pit now!
I usually respond pretty bad when summoned, but I'll make an exception, here's my answer (and I really mean it):
Maybe, I don't know.
Hi, AKSA,
Could it be something similiar, the "trapped base charge effect" that causes this soft switching? Without the driver's emitor resistor (bleeder for this base charge), the remaining trapped base charge will continue opening the output transistors, making somekind of shoot through between upper and lower output transistors. If this is true, then this effect should be more obvious at higher frequencies than lower frequencies and should vary with different type output transistor (different hfe and different base capacitance), for the same frequency.
If this effect caused by the bias chain, then it shouldn't vary much with different output transistors.
Could it be something similiar, the "trapped base charge effect" that causes this soft switching? Without the driver's emitor resistor (bleeder for this base charge), the remaining trapped base charge will continue opening the output transistors, making somekind of shoot through between upper and lower output transistors. If this is true, then this effect should be more obvious at higher frequencies than lower frequencies and should vary with different type output transistor (different hfe and different base capacitance), for the same frequency.
If this effect caused by the bias chain, then it shouldn't vary much with different output transistors.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.