I focus more on the audibly different sound characters of complementary transistors. The resulting noise - not the commonly understood "distortions": Complementary parts push-pull sound dirty, bumpy, have a "Spanish wall" in the sound image, and more - compared to, for example, SE with, for example, the same number of stages and the same parts and power supplies (or one power supply;-)
The conceivable different gain or distortion values of complementary parts could be handled to some extent with line viewing "measuring" methods and adjustment work - which, however, would not justify a high effort, as a half-wave separated amplification by means of different sounding components is unsuitable for audio applications. We don't get together the different audible tonal characters of complementary transistors. It remains a dirty "distorted" sound: "noise".
Since pp uses same (not identical) parts in terms of sound the tonal "distortions", audible "distortions", Noises, are less than of complementary parts pp designs.
The problem of pp designs in audio, hifi, highend is that even the same components sound different. And different batches often sound very different.
But a Le Monstre is a great party story. I'll see if I can build another one from TO-126 transes as "driver" and power stage;-)
The conceivable different gain or distortion values of complementary parts could be handled to some extent with line viewing "measuring" methods and adjustment work - which, however, would not justify a high effort, as a half-wave separated amplification by means of different sounding components is unsuitable for audio applications. We don't get together the different audible tonal characters of complementary transistors. It remains a dirty "distorted" sound: "noise".
Since pp uses same (not identical) parts in terms of sound the tonal "distortions", audible "distortions", Noises, are less than of complementary parts pp designs.
The problem of pp designs in audio, hifi, highend is that even the same components sound different. And different batches often sound very different.
But a Le Monstre is a great party story. I'll see if I can build another one from TO-126 transes as "driver" and power stage;-)
Yes but that one has 4 times the active devices as the original NemesisWe measured this at your place together last summer :
https://www.diyaudio.com/community/threads/the-jean-hiraga-nemesis-quadrige.401114/
View attachment 1317011
Patrick
Jan
Taken it back to where it should belong :
https://www.diyaudio.com/community/threads/the-jean-hiraga-nemesis-quadrige.401114/post-7698926
Patrick
https://www.diyaudio.com/community/threads/the-jean-hiraga-nemesis-quadrige.401114/post-7698926
Patrick
Lots of information on the design considerations in the original article of 1983.
https://sound-au.com/tcaas/monster27.htm
Some decades before the "F5" ?
Patrick
https://sound-au.com/tcaas/monster27.htm
Some decades before the "F5" ?
Patrick
I tested quite a number of driver transistors at a supply voltage of +/-19V (that's the output voltage of my modified Walt's Improved_PN_Regs) and a current of 1.7A (I would increase that if the heatsinks allow, the optimum is 1.9-2A) some three years ago . However, the output transistor is also important, which must have as little base current as possible. Then we can use transistors that are intended for VAS and they are an excellent choice, 2SC3503/2SA1381 or similar are suitable even at +/-19V supply voltage. Currently, I have 2SC1940K/2SA915K (better then 2SC3503/2SA1381) in that position and the output transistors are 2SC3284Y/2SA1303Y.
Attachments
Last edited:
One if the purpose of our build was to make you of active components as much as possible, or at least ones that can be obtained from reliable source at reasonable price, so that others can also build.
The KSA1381/KSC3503 we used were obtained from the forum at very fair price :
https://www.diyaudio.com/community/threads/fairchild-ksc3503-estu-ksa1381-estu.397239/#post-7599004
The PNP in E grade is active.
The NPN in E grade is in stock at Rochester.
So they can easily be sourced.
The output transistors are active at Mouser and DK.
The JFETs are the only ones difficult.
But there are just no alternatives.
And we'll see how they perform in listening by a panel, however subjective.
Cheers,
Patrick
The KSA1381/KSC3503 we used were obtained from the forum at very fair price :
https://www.diyaudio.com/community/threads/fairchild-ksc3503-estu-ksa1381-estu.397239/#post-7599004
The PNP in E grade is active.
The NPN in E grade is in stock at Rochester.
So they can easily be sourced.
The output transistors are active at Mouser and DK.
The JFETs are the only ones difficult.
But there are just no alternatives.
And we'll see how they perform in listening by a panel, however subjective.
Cheers,
Patrick
Last edited:
What would be the main difference in performance or in sound if I use Toshiba TTC004b/TTA004b and 2SC5200/2SA1943? For CCS I use J-Fet CPH3910 with ADR510 for the 5V version until I get new WJ 1V reference. For higher voltages I use 6.8V GLED431.
Are there more easily available parts in Mouser or DK with some better parts then Toshibas and the CPH3910 J-Fet?
Hi all,
I'm the lucky guy getting the chance to work on the Le Monstre adaptation for 2024 and I have just had my first chance with some friends with educated ears to have a listen to one of the Le Monstre versions mentioned by EUVL in the first post/PDF. The version available was the Le Monstre 12V, with modern semis. This was no more than a first listen, more structured listening will take place as other versions are completed, chassis prepared etc. The setup was also hardly fair to the amp - I was using my 01a DHT preamp for the task, and the speakers are my stacked quad 57s, certainly not the easiest load for an amp with higher level output impedance. But past experience tells me that the electrostatics are not as tough a drive as the specs might say. The Le Monstre was powered from a lab supply set at +/- 12V, feeding 150kuF caps and then with 1000uF decoupling on the amp boards. So I think all you can say from this is that it was a taster of initial impressions of an amp under development - the other amps heard during the evening were a J2, and later my own build of the F5Pi (https://www.diyaudio.com/community/...rfectly-integrated.406024/page-6#post-7657850)
The comments from the panel here were very complimentary towards the amp - they were very impressed with the sound. In particular, the soundstage was mentioned, and that bass seemed tighter than with the J2. They were amazed that the low output power could produce such a robust sound, and we were listening reasonably loud - I did not measure SPL, but you would have to speak up quite noticeably to talk to each other, so probably as loud as I think most people would listen. I want to note that the panel are all dedicated/seasoned listeners with excellent audio systems, and so are well used to listening to excellent sound. They are also quite brutal in their comments, no punches are pulled.
So what does this rambling post actually contribute here? I think it serves to give a brief glimpse of this adaptation, it certainly encourages the further work, testing, building and development of the versions mentioned in the PDF, and that it is highly likely that the investment of time and money will be well worth it.
2x 70Ah batteries have now been sourced to replicate the original!
Fran
I'm the lucky guy getting the chance to work on the Le Monstre adaptation for 2024 and I have just had my first chance with some friends with educated ears to have a listen to one of the Le Monstre versions mentioned by EUVL in the first post/PDF. The version available was the Le Monstre 12V, with modern semis. This was no more than a first listen, more structured listening will take place as other versions are completed, chassis prepared etc. The setup was also hardly fair to the amp - I was using my 01a DHT preamp for the task, and the speakers are my stacked quad 57s, certainly not the easiest load for an amp with higher level output impedance. But past experience tells me that the electrostatics are not as tough a drive as the specs might say. The Le Monstre was powered from a lab supply set at +/- 12V, feeding 150kuF caps and then with 1000uF decoupling on the amp boards. So I think all you can say from this is that it was a taster of initial impressions of an amp under development - the other amps heard during the evening were a J2, and later my own build of the F5Pi (https://www.diyaudio.com/community/...rfectly-integrated.406024/page-6#post-7657850)
The comments from the panel here were very complimentary towards the amp - they were very impressed with the sound. In particular, the soundstage was mentioned, and that bass seemed tighter than with the J2. They were amazed that the low output power could produce such a robust sound, and we were listening reasonably loud - I did not measure SPL, but you would have to speak up quite noticeably to talk to each other, so probably as loud as I think most people would listen. I want to note that the panel are all dedicated/seasoned listeners with excellent audio systems, and so are well used to listening to excellent sound. They are also quite brutal in their comments, no punches are pulled
.So what does this rambling post actually contribute here? I think it serves to give a brief glimpse of this adaptation, it certainly encourages the further work, testing, building and development of the versions mentioned in the PDF, and that it is highly likely that the investment of time and money will be well worth it.
2x 70Ah batteries have now been sourced to replicate the original!
Fran