Dear Hans and Andrej, thanks for the tips. Interesting geometries there. I might try some for myself. 
My last PC was a cheap OCC copper with Furukawa wires.
Furukawa PCOCC 2Meters Audio Power Cable Cord 2M 9AWG | eBay
Not bad for the price, though it has not any especial spatial arrangement to its wires.
Dear Andrej, have you felt the need to check the CLC option for your SMPS's output?
I only have tested on one SMPS whose output was suspicious to begin with and now my scope is not operative so I cannot make further experiments...
I know you prefer "the other" topology but any advise will be welcome.
Cheers,
M.
Edit: now I see there are offerings for Gold Plated OCC copper wire.
My last PC was a cheap OCC copper with Furukawa wires.
Furukawa PCOCC 2Meters Audio Power Cable Cord 2M 9AWG | eBay
Not bad for the price, though it has not any especial spatial arrangement to its wires.
Dear Andrej, have you felt the need to check the CLC option for your SMPS's output?
I only have tested on one SMPS whose output was suspicious to begin with and now my scope is not operative so I cannot make further experiments...
I know you prefer "the other" topology
but any advise will be welcome. Cheers,
M.
Edit: now I see there are offerings for Gold Plated OCC copper wire.
Last edited:
I found a limited time offer for Gold Plated OCC copper 26AWG, 60 feet: good for interconnects and internal signal.
18.3m Elf Custom Cables 24k Placcato Oro 99.9999% pure Occ Filo di Rame | eBay
Of course I will try them!
Gold is the future.
M.
18.3m Elf Custom Cables 24k Placcato Oro 99.9999% pure Occ Filo di Rame | eBay
Of course I will try them!
Gold is the future.
M.
Excellent. Welcome.
I have the four pairs of SONY 2SK60/2S18, if I remember correctly, which belong to the SONY TA-5650. Two pairs are on duty there.
I have two pairs of NEC 2SK70/2SJ20 (again, from memory), which are the ones that I intend to use as bootstraps on this amp's output since I hope they will give me the right voltages for the main BJTs. If too low, the SONY will probably be OK.
Then I have several non complimentary that I use or will use on single ended amps: TOKIN 2SK180 on use; SONY 2SK82 on use; TOKIN THF51S for units for future use.
Cheers,
M.
Those are veritable V-FETs (SIT) all right, very valuable, except the Tokin parts that are not in the same class.
Would you take a look at this please:
http://www.amplimos.it/images/hiraga2_Vfet.gif
Would you take a look at this please:
http://www.amplimos.it/images/hiraga2_Vfet.gif
Thank you. I already know that amp and of course, planned a cascoded output for it but never built it.
Maybe you did not get the spirit of this thread, which is in a broad sense, to try new paths with the excuse of the Low Thermal Memory thing, but not limited to it. We may totally fail but at least we would have tried.
And yes the Tokin are industrial parts (hard to destroy) but they can be made to sound good and they can also be submitted to cascodying to make negligible the big Cgd. That is why I bought four THF 51S
But of course I will cascode them first with a big transistor or a MOSFET.Honestly, after listening to a LTMD amp, all other (safe maybe some single ended tube amps) sound boring to me.
Which reminds me, having to compare my two amps with other amps and using different sources and with different cables make the possible permutations endless so I won't comment on sound perception, meaning the usual frequency predominance, colour ...etc
I will only insist that the persistent trait is satisfactory dynamics and relaxation factor. Now, if more braves want to try (and perfect or clean-up) the amp, you will be also welcome.
I forgot to show my new upgraded power cord for my SD card player plus TDA1543 DUAL-MONO DAC, with 12awg solid core OCC copper surrounded by 3 wires of silver plated OFC (aerospatial grade) 22AWG winded in counter-rotation on both active leads, imitating the expensive example given by LC . The ground lead is (almost) straight. Noise floor went down and the picture is more orderly but colder now, IMHO.
Sorry to insist with these horrible videos...I think the cellphone not only distorts the highs but it also fails to catch the low bass.
YouTube
Oops! It is twisted.
Ha, ha!
M.
Attachments
Last edited:
Should try the same power cord for power amp too.
Phone's mic is not so bad, freq resp is the least of a problem, compressor-limiter is the one to blame. This is the one taken by mine.
Yours sound good. Seems iPhone does it OK. Mine is Samsung something: the only requisite I asked my wife was to get me something for free
Maybe I should borrow one.
I was thinking: why don't I use inductors instead of diodes to isolate the input/VAS from the output section?
I have some cute coils here. Maybe that would improve stability in the real world.
Talking about poetry, may I post here a music from Chilean group "Los Jaivas" composed for Chilean Poet Pablo Neruda's "Canto General". Though I do not share his political views, it cannot be denied that his poetry is full of power on a level // to Homer. Surelly he received inspiration from somewhere.
YouTube
Complete album:
YouTube
Lyrics and then free translation.
I forgot to mention that Peru's Minister of Culture, after aprouval of the filming and given that heavy intruments were to be loaded there, said: -"but please, do not ruin our ruins"
Last edited:
That is clear to me. Lin/Blameless topology is not that bad.
What amazes me is that so few are willing to make a simple test.
I could not answer in a technical way your last question and I will not study further now because I am busy with urgent matters.
Cheers,
M.
Those capacitances are HUGE, so what's the point?
A memoryless transition means a transient change of state, it is like throwing shells into the Llanquihue Lake.
Consider the Hiraga circuit, your chances of coming up with something better based on dubious theories is in the close vicinity of zero.
A memoryless transition means a transient change of state, it is like throwing shells into the Llanquihue Lake.
Consider the Hiraga circuit, your chances of coming up with something better based on dubious theories is in the close vicinity of zero.
I have not discarded the Hiraga at all. I talked before about chances of success...
The basic part of the Amnesis is the novel input and VAS configuration (thanks to Hephaïstos and Peufeu) and I consider the driver/output section still experimental, so it can be configured in many ways, from typical emitter follower to more complex, like simple or cascoded Darlington or cascoded Sziklai...etc. Maybe I am completely wrong and simple EF is better, who knows, but for that I need more listeners, should they want to invest time and effort experimenting.
IF they can be made stable, then the final personal decision is subjective. IF one can detect a difference in sound emission then what is the problem to choose the (subjectively) better sounding (real dynamics and relaxation factor) configuration?
Maybe someone can even invent and Hybrid with Valve output.
What do you think about:
a) The effect of transforming the basic input differential amplifier into a cascoded (bootstrapped) Sziklai?
b) Cascodying the common emitter VAS?
c) A cascoded common source/emitter output?
(very tricky)
Cheers,
M.
PS: I am trying to bring technical aspects for the technical mates to bite into, but remember, your model of reality is not reality, so we would not know until we try, and even then there may be things hidden from us...this is DIY not PIY meaning predict it yourself, that's why I'm doing even if guided by intuition.
The input transistor should be a JFET biased at a small Vds. The grounded base transistor is needed to carry out the voltage amplification.
What would be the purpose behind the Sziklai configuration here? Did Peufeu say that?
I eschew the long tailed pair, the Darlington pair as well as the emitter-follower.
Pointless.
For what reason? The output stage should have low voltage gain obtained by FETs with a high channel (output) impedance, a small reverse transfer capacitance and preferably also a fairly low input impedance (SIT).
Thanks dear N101N. That helps. I have made a couple of inputs with combined JFET-->BJT Sziklai already and it works very well. I understand that JOSI1 also has tried it with success.
The idea is trying to achieve constant power (lowest memory) for input's (and VAS) critical active devices and Peufeu explains it here in his fourth part:
Memory Distortion Philosophies - Part 4 : Circuits
Well, we have to use a LTP, be it resistor or BJT based (modded Wilson) and make listening tests...
Cheers,
M.
A good lot of nonsense can be concatenated in something that looks like a serious posting.
But you have shown elsewhere that you can still do a lot better and funnier, so I’m a bit disappointed.
Hans
Guys, do not start a fruitless argument.
The progress of "science" starts when somebody asks: "What if?".
Hephaïstos asked "what IF" and then experimented and built. Based on the former, he managed to start his own company, which is still around BTW, however he is not...
He was gracious enough and shared his thoughts for us to try by ourselves and judge. These are simple mods to be performed in any Lin/Blameless to judge its worth in real life, real amps.
The pre-conception of worth based on artificial models is what has this World as bad as it is...
Best wishes,
M.
Last edited: