Just use it because it has great potential, with adequate loudspeakers, good recording and all other things that may affect listening experience. This clone has measurements that indicate high quality amplifier, if built from quality (not necessarily expensive) components. However 'completed' Chinese board is of poor quality, a guarantee for boring sound.
For the same reason I have almost trashed this project, especially after some meassurements. Then I have realised that I have measured the board I had at that time on the bench: Chinese 'completed' board.
It suferred of premature roll-off associated with phase shift at low frequencies. At high frequencies there was an overshoot and slight ringing.
Then came Marigno..
After a couple of weeks I have realised what I was meassuring: junk soldered on otherwise fine board.
A clone with components suggested by marino offered quite different picture: a ruller-straight frequency response from practically 1 Hz up to 250 kHz. To square waves the board responded with almost ideal, rectangles within the audibe range.
I intend to invest in six monoblocks with aim to build a system based on three-way active crossover.
I couldn't agree more.
Let me refer to a JFK quote to illustrate this: "Ask not what..."
This is essential: we should never forget that the amplifier is just a part of the entire system. There are so many things that should be done by us to extract the maximum from this clone. This is our task. I am still struggling but have made progress. I have found very helpful several posts at the very beginning of this conversation by Danny_66.
Last edited:
For the same reason I have almost trashed this project, especially after some meassurements. Then I have realised that I have measured the board I had at that time on the bench: Chinese 'completed' board.
It suferred of premature roll-off associated with phase shift at low frequencies. At high frequencies there was an overshoot and slight ringing.
Then came Marigno..
After a couple of weeks I have realised what I was meassuring: junk soldered on otherwise fine board.
A clone with components suggested by marino offered quite different picture: a ruller-straight frequency response from practically 1 Hz up to 250 kHz. To square waves the board responded with almost ideal, rectangles within the audibe range.
I intend to invest in six monoblocks with aim to build a system based on three-way active crossover.
The original (the WHA-97 is very close to the NHB-108), imho, is too sentitive to the load, even using high quality components. It refuses to drive my Acoustat, the sound is horrible. The sound is not so dynamic and fatiguing even with my lab speakers, a lot easier load.
The buffered one (WHA-217=WHA-97+NSCB Buffer) drives the Acoustat like a charm. I used 11 pairs but the project can be downsized to as many pairs as you want. The Sanken MT200 pairs I used, 2SC3264/2SA1295, are multi emitter BJTs so, perhaps, 2 pairs could be enough. The mod is not difficult at all but requires some more room for a "daughter board", the temperature/bias control pair and the NSCB buffer added pairs.
I would avoid to add some other pairs to the original project (parallel).
By the way, using only one transformer does not generate any hum as some said.
For your project I would consider also the LJM L20 V9.2 (avoid V10) boards with a mod focused on adding a buffer to the VAS and some other simple mods.
Last edited:
Che piacere vederti Marigno! (Eng. How nice it is to see you!)
Thanks Marigno for valuable information
Indeed, there are several posts at the first 15-20 pages indicating the need for two transformers for attaining the goal of total silence. The most remarkable are posts by Danny_66. He had so many useful ideas.
I have already attained a goal of graveyard silent amplifier by using completely independent power supplies: for each channel one transformer, two rectifiers and four capacitors. I have made many attempts to reduce the hum when using just one transformer for two channels but couldn't attain the objective of total silence. It appears that whatever I do just reduces the hum but doesn't eliminate it because there is a large loop surrounding the entire amplifier, input, then left channel through the common transformer to the right channel back to the RCAs.
I had some ideas with hum breaking resistors (HBR) but the hum was still there, reduced but slightly audible. Unacceptable.
How do you wire to eliminate the hum with DarTZeel?
Salve!
(Eng. How nice it is to see you!)Thanks Marigno for valuable information
This is exactly what I was investigating these past few days. I have a choice: six or three transformers for my tri-amping monster: width: 500mm , depth: 350 mm and height: 750 mm, weight: probably 30-40 kg, with all these heavy-weight heat sinks and transformers.
Indeed, there are several posts at the first 15-20 pages indicating the need for two transformers for attaining the goal of total silence. The most remarkable are posts by Danny_66. He had so many useful ideas.
I have already attained a goal of graveyard silent amplifier by using completely independent power supplies: for each channel one transformer, two rectifiers and four capacitors. I have made many attempts to reduce the hum when using just one transformer for two channels but couldn't attain the objective of total silence. It appears that whatever I do just reduces the hum but doesn't eliminate it because there is a large loop surrounding the entire amplifier, input, then left channel through the common transformer to the right channel back to the RCAs.
I had some ideas with hum breaking resistors (HBR) but the hum was still there, reduced but slightly audible. Unacceptable.
How do you wire to eliminate the hum with DarTZeel?
I would like very much to use that boards, but which is the LJM L20 V9.2?
Salve!
Completely agree, same story with my Wilsons. The original output stage demands an easy load to be worth having. This is what caused the update to Model 2.
I would like very much to use that boards, but which is the LJM L20 V9.2?
Salve!
I have to run now, I'll have some time later. I want to let you know this about L20 V9.2:
LJM L20 V9.0 Amplifier
Post #53 - R34 is 68Ohm
Sorry if it doesn't translate to a link, you can copy/paste in your browser. See you later.
The NHB-108 is not an affordable DIY kit, not because of the clone price, or the PCB and component price, that, however, are not cheap, but because of the effort you have to put on it to make it work, and the required skill.
My wish was to have an amplifier with no feedback from the speaker connection so I went through lots of difficulties to achieve my wish, because I am a stubborn person. I almost surrendered and tried the LJM L20 V9.2 that remains a good no problem amplifier, that becomes a very good one with a few blows of welder. When I understood that the problem with the NHB-108 mods was the fake transistors, I left the L20 project not finalized and went back to the one that would have become the WHA-217; I had to experiment a lot about the bias and temperature control and now I am very proud of the final result, the WHA-217. Then I finalized also the L20, that is able to drive very very well the lab speakers and more than decently the Acoustat.
Back to the NHB-108.
The problems begin with the PCB: it doesn't have a correct ground path, the ground is the upper plate that collects all components to be connected to the ground. At least is seems to be progressive, following the natural signal path, but it is not the optimal ground path at all; the bottom ground plate is ok because it is not connected to any component and works as a shield but must be connected by you to the upper plate otherwise remains floating. Those, at least, are my PCBs
Then, since we are DIYers and we are not a hi-end factory, we must stay with things that the market offers. So a small asymmetry of the two branches of the transformer generates hum. You can attenuate a lot the hum by removing the trimmer but still some remains. You better use a very big section cables to connect power lines to the circuit and the cabling must have a very "natural path" interweaving the 3 lines, no loops over other wires, it doesn't matter if the path will be longer; the speaker GND must be connected to the center of the power supply. If the copper is too thick for the screw connector on the servo, you can cut out some strands to let it fit. Look at the picture of my WHA-97, a few posts ago.
When I say no hum, it properly means that, when the room is silent, you put your ear close to the woofer and you can barely hear something. It is a very good result, close to a good branded amplifier. Obviously the two transformer solution is better because you don't close any signal/ground loop, provided that you totally separate the two GNDs, to be coupled to the cabinet with the usual coupling circuit R/C/antiparalleled diode couple.
But: can you imagine a cabinet with 6 transformers inside (2 secondaries for dual power supply and 1 secondary for the protection), 6 dual power supplies, 6 speaker protection boards, 6 GND coupling circuits? Your cabinet has not standard dimensions and you will have logistic problems, Hi-Fi2000 has a 5 unit "Dissipante" 50cm deep on request. My WHA-217 is in a 5 unit "Dissipante" 40cm deep and weighs 60Kg. Imho, you better build 3 dual mono amps, a solution with 2 tri mono amps is a nonsense because you have only two sides for the needed heat sinks.
Are you sure the 3-way electronic crossover is a good solution? In Italy I still have 2 tube main amps driving in biamp a couple of bi-wired BW801, without any electronic crossover. The sound is impressive. One amp drives the woofer, the other drives the midrange and the tweeter. If you use an electronic crossover, you still have to use a passive crossover to the speakers, even if you use 1/1 amp/speaker. So, why use an electronic crossover? The less the better!
What would I do for a multiamp solution? with a 3 way speakers, I would choose a bi-amping solution with two dual mono amps and no electronic crossover; knowing by this forum that the NHB-108 has the big problem about the acceptance of the load (by now we are a lot saying this), and being sure that I cannot judge the sound before finishing the project (too late), I would choose a consolidated solution, that I have to consolidate myself by the faith in somebody or by experimenting at least one stereo solution. After knowing the potential of the WHA-217, I would choose a WHA-217 downgraded solution: 4 final pairs/140V (maximum recommended) for the woofer, 3 final pairs/120-110V for mid/tw. After that you could build a third one with 2 final pair/100V for the tw. Now you are half way to a stereo WHA-217, you could give it a try. Otherwise you could try the L20 V9.2, molto più tranquillo, even without the buffer added to the VAS (having a triple output Darlington it is not so critical) but with the other mods regarding the lower bias, the feedback components, the input capacitor, the Re's brought down to 0.1Ohm and an output coil (not in the schematic); you find it at the link I sent you in my recent post.
BTW, the L20 has the feedback loop from the speakers...
My wish was to have an amplifier with no feedback from the speaker connection so I went through lots of difficulties to achieve my wish, because I am a stubborn person. I almost surrendered and tried the LJM L20 V9.2 that remains a good no problem amplifier, that becomes a very good one with a few blows of welder. When I understood that the problem with the NHB-108 mods was the fake transistors, I left the L20 project not finalized and went back to the one that would have become the WHA-217; I had to experiment a lot about the bias and temperature control and now I am very proud of the final result, the WHA-217. Then I finalized also the L20, that is able to drive very very well the lab speakers and more than decently the Acoustat.
Back to the NHB-108.
The problems begin with the PCB: it doesn't have a correct ground path, the ground is the upper plate that collects all components to be connected to the ground. At least is seems to be progressive, following the natural signal path, but it is not the optimal ground path at all; the bottom ground plate is ok because it is not connected to any component and works as a shield but must be connected by you to the upper plate otherwise remains floating. Those, at least, are my PCBs
Then, since we are DIYers and we are not a hi-end factory, we must stay with things that the market offers. So a small asymmetry of the two branches of the transformer generates hum. You can attenuate a lot the hum by removing the trimmer but still some remains. You better use a very big section cables to connect power lines to the circuit and the cabling must have a very "natural path" interweaving the 3 lines, no loops over other wires, it doesn't matter if the path will be longer; the speaker GND must be connected to the center of the power supply. If the copper is too thick for the screw connector on the servo, you can cut out some strands to let it fit. Look at the picture of my WHA-97, a few posts ago.
When I say no hum, it properly means that, when the room is silent, you put your ear close to the woofer and you can barely hear something. It is a very good result, close to a good branded amplifier. Obviously the two transformer solution is better because you don't close any signal/ground loop, provided that you totally separate the two GNDs, to be coupled to the cabinet with the usual coupling circuit R/C/antiparalleled diode couple.
But: can you imagine a cabinet with 6 transformers inside (2 secondaries for dual power supply and 1 secondary for the protection), 6 dual power supplies, 6 speaker protection boards, 6 GND coupling circuits? Your cabinet has not standard dimensions and you will have logistic problems, Hi-Fi2000 has a 5 unit "Dissipante" 50cm deep on request. My WHA-217 is in a 5 unit "Dissipante" 40cm deep and weighs 60Kg. Imho, you better build 3 dual mono amps, a solution with 2 tri mono amps is a nonsense because you have only two sides for the needed heat sinks.
Are you sure the 3-way electronic crossover is a good solution? In Italy I still have 2 tube main amps driving in biamp a couple of bi-wired BW801, without any electronic crossover. The sound is impressive. One amp drives the woofer, the other drives the midrange and the tweeter. If you use an electronic crossover, you still have to use a passive crossover to the speakers, even if you use 1/1 amp/speaker. So, why use an electronic crossover? The less the better!
What would I do for a multiamp solution? with a 3 way speakers, I would choose a bi-amping solution with two dual mono amps and no electronic crossover; knowing by this forum that the NHB-108 has the big problem about the acceptance of the load (by now we are a lot saying this), and being sure that I cannot judge the sound before finishing the project (too late), I would choose a consolidated solution, that I have to consolidate myself by the faith in somebody or by experimenting at least one stereo solution. After knowing the potential of the WHA-217, I would choose a WHA-217 downgraded solution: 4 final pairs/140V (maximum recommended) for the woofer, 3 final pairs/120-110V for mid/tw. After that you could build a third one with 2 final pair/100V for the tw. Now you are half way to a stereo WHA-217, you could give it a try. Otherwise you could try the L20 V9.2, molto più tranquillo, even without the buffer added to the VAS (having a triple output Darlington it is not so critical) but with the other mods regarding the lower bias, the feedback components, the input capacitor, the Re's brought down to 0.1Ohm and an output coil (not in the schematic); you find it at the link I sent you in my recent post.
BTW, the L20 has the feedback loop from the speakers...
Last edited:
@marigno
Thank you very much for your exhaustive reply and your time. Now I have enough reading for the ongoing few weeks or months to go through everything you've just said. Everyone here should read this carefully.
I will build several alternatives and make choice at the end.
Take care my friend.
Thank you very much for your exhaustive reply and your time. Now I have enough reading for the ongoing few weeks or months to go through everything you've just said. Everyone here should read this carefully.
I will build several alternatives and make choice at the end.
Take care my friend.
You're very welcome.
A useful reference about crossover: engineer Bartolomeo Aloia says that, if your amplifier drives only one speaker, you cannot avoid to use a passive filter to this one speaker. I cannot remember why but I have this concept well printed in my mind. It was on a magazine a lot of years ago, before the PC, and I doubt you can find this info online.
So, if Aloia is right, an electronic crossover is more detrimental than useless. The amplifier, theoretically but, with a good approximation, also in practice, works according to the passive filter, being loaded only within the frequency interval imposed by the filter, remaining almost unloaded out of that range. A good condition.
In my experience, he is right.
Should you choose the LJM solution, do not go for the L20 V10, it is really bad! No more triple Darlington and a lot of simplifications and compensations. The best is the 9.2.
Do not hesitate to contact me, even in pm, if you need some info.
A useful reference about crossover: engineer Bartolomeo Aloia says that, if your amplifier drives only one speaker, you cannot avoid to use a passive filter to this one speaker. I cannot remember why but I have this concept well printed in my mind. It was on a magazine a lot of years ago, before the PC, and I doubt you can find this info online.
So, if Aloia is right, an electronic crossover is more detrimental than useless. The amplifier, theoretically but, with a good approximation, also in practice, works according to the passive filter, being loaded only within the frequency interval imposed by the filter, remaining almost unloaded out of that range. A good condition.
In my experience, he is right.
Should you choose the LJM solution, do not go for the L20 V10, it is really bad! No more triple Darlington and a lot of simplifications and compensations. The best is the 9.2.
Do not hesitate to contact me, even in pm, if you need some info.
Last edited:
My reasoning is as follows:
Active crossover divides frequencies and then sends separated signal into separate amplifiers. This allows us to optimize each amplifier for specific frequencies. Also, the amplifier doesn't have to cope with sudden peaks of different frequencies. Also we can, for example elliminate the input capacitor in woofer and tweeter amplifiers because DC is completely and unconditionally elliminated from these signals. And, more important, we have freedom of choice of loudspeakers and use them without passive srossovers that may severely detoriate sound. We have freedom of choice to accomodate the speakers for each frequency range. And it is the speaker that translates electronic signals into air vibrations. Like a bow and violin: each performs its own task. This is probably the most critical step in translating recordings into air vibrations. Music.
I have already built and tested threeway Linkwitz-Riley active crossover. I had much better results compared to passive crossovers. It is already in a DIY chasis made of of 3 mm aluminium sheets. I can TIG weld stainless steel and I use router with both wood and aluminium. Even though I am just a chemist.
I wish you a pleasant evening in the US. Here in Europe it is already early morning and I have to jump into the bed.
Buona mattina!
Active crossover divides frequencies and then sends separated signal into separate amplifiers. This allows us to optimize each amplifier for specific frequencies. Also, the amplifier doesn't have to cope with sudden peaks of different frequencies. Also we can, for example elliminate the input capacitor in woofer and tweeter amplifiers because DC is completely and unconditionally elliminated from these signals. And, more important, we have freedom of choice of loudspeakers and use them without passive srossovers that may severely detoriate sound. We have freedom of choice to accomodate the speakers for each frequency range. And it is the speaker that translates electronic signals into air vibrations. Like a bow and violin: each performs its own task. This is probably the most critical step in translating recordings into air vibrations. Music.
I have already built and tested threeway Linkwitz-Riley active crossover. I had much better results compared to passive crossovers. It is already in a DIY chasis made of of 3 mm aluminium sheets. I can TIG weld stainless steel and I use router with both wood and aluminium. Even though I am just a chemist.
I wish you a pleasant evening in the US. Here in Europe it is already early morning and I have to jump into the bed.
Buona mattina!
Last edited:
Very good reasons!
I have different preferences, I don't like too much active filters because they are founded on feedback loops, and I hate to use signal feedback loops as a function, in my projects. Moreover you have to cope with phase rotations and delays, even if with passive filters it is almost the same. I know there are solutions for these problems but, again, I don't like to add stuff to stuff. I always say "the less the better".
Do not be offended for what I say, it is only a different opinion from yours; sincerely I'm too old style, but I'm open minded, ready to change opinion listening at your project, once concluded. But we are too far, I'm feeling nostalgia but I cannot travel to Italy at this time and, however, I wouldn't be able to meet you because of restrictions.
If you like the sound of your active crossover equipment, that seems to be a Hi-End component, you should choose a high quality main amp (3 of them). One of the best amp I could hear is an OTL by Siaudio (Napoli) with no feedback loop at all; they also sell 3 kits with 3 different output power ratings (in 2013 they did). Then there is the old project "Superfinale" with only internal feedback loop and an external loop with low factor feedback (I don't recommend) and there is the NHB-108 modded as WHA-217 with one only internal feedback loop. I wouldn't invest money on expensive kit with feedback (BJT, IGBT, lateral MOSFET, vertical MOSFET etc.) and super expensive components, the sound is, more or less, always the same; I would rather consider solutions by Hypex/nCore with traditional PSU.
The difference between an amp using a feedback loop from the output and another amp, with no feedback, is striking. The feedback takes away all the truthfulness of the sound. But, again, this is my opinion, based on what I feel, that cannot be objectfied/measured.
From what you said I must advise you to trash:
- the straight NHB-108 because you have to find the speakers that he likes to drive;
- the L20 solution, because it is a toy that could become a Hi-End toy, but always remains a toy.
Have a wonderful day!
I have different preferences, I don't like too much active filters because they are founded on feedback loops, and I hate to use signal feedback loops as a function, in my projects. Moreover you have to cope with phase rotations and delays, even if with passive filters it is almost the same. I know there are solutions for these problems but, again, I don't like to add stuff to stuff. I always say "the less the better".
Do not be offended for what I say, it is only a different opinion from yours; sincerely I'm too old style, but I'm open minded, ready to change opinion listening at your project, once concluded. But we are too far, I'm feeling nostalgia but I cannot travel to Italy at this time and, however, I wouldn't be able to meet you because of restrictions.
If you like the sound of your active crossover equipment, that seems to be a Hi-End component, you should choose a high quality main amp (3 of them). One of the best amp I could hear is an OTL by Siaudio (Napoli) with no feedback loop at all; they also sell 3 kits with 3 different output power ratings (in 2013 they did). Then there is the old project "Superfinale" with only internal feedback loop and an external loop with low factor feedback (I don't recommend) and there is the NHB-108 modded as WHA-217 with one only internal feedback loop. I wouldn't invest money on expensive kit with feedback (BJT, IGBT, lateral MOSFET, vertical MOSFET etc.) and super expensive components, the sound is, more or less, always the same; I would rather consider solutions by Hypex/nCore with traditional PSU.
The difference between an amp using a feedback loop from the output and another amp, with no feedback, is striking. The feedback takes away all the truthfulness of the sound. But, again, this is my opinion, based on what I feel, that cannot be objectfied/measured.
From what you said I must advise you to trash:
- the straight NHB-108 because you have to find the speakers that he likes to drive;
- the L20 solution, because it is a toy that could become a Hi-End toy, but always remains a toy.
Have a wonderful day!
Last edited:
A hypothesis on the clone, an exercise to think about. TO BE VERIFIED.
The two diodes in the base of the output pairs of the diamond buffer are non-linear components that switch on when the driving signal is not close to zero, but when the load presented to the base, via the hfe of the transistor, requires more current than the one allowed by the 27Ohm resistors, during the rise of the power semi-cycle; the diodes switch off when the semi-cycle goes towards zero but is still distant from it.
You cannot simulate the fact with a function generator, you need a complex signal, a musical signal.
If you remove the diodes, you must replace the 27Ohm tiny resistors with 3W or bigger ones but expect a loss of power and driving capabilities. If you lower the value of the resistors expect a higher bias. So there are experiments to be conducted.
Different is the NSCB, where diodes are always on during the power semi-cycle and switch off a little after the opposite semi-cycle is started.
I think the WHA-217 is insensible to the load because the diodes, on the diamond buffer pairs, do not switch at all, being the current, flowing through the 27Ohm resistors, enough to drive the load via the diamond buffer and the added NSCB stage. TO BE VERIFIED.
The two diodes in the base of the output pairs of the diamond buffer are non-linear components that switch on when the driving signal is not close to zero, but when the load presented to the base, via the hfe of the transistor, requires more current than the one allowed by the 27Ohm resistors, during the rise of the power semi-cycle; the diodes switch off when the semi-cycle goes towards zero but is still distant from it.
You cannot simulate the fact with a function generator, you need a complex signal, a musical signal.
If you remove the diodes, you must replace the 27Ohm tiny resistors with 3W or bigger ones but expect a loss of power and driving capabilities. If you lower the value of the resistors expect a higher bias. So there are experiments to be conducted.
Different is the NSCB, where diodes are always on during the power semi-cycle and switch off a little after the opposite semi-cycle is started.
I think the WHA-217 is insensible to the load because the diodes, on the diamond buffer pairs, do not switch at all, being the current, flowing through the 27Ohm resistors, enough to drive the load via the diamond buffer and the added NSCB stage. TO BE VERIFIED.
Last edited:
After reading the latest posts is hard to justify the effort to build the 108 for which I had finished designing the boards.
I think I will change my focus and build a Bryston clone instead which was my main project before the Dartzeel appeared into the scene and for which I have the boards designed already.
I think I will change my focus and build a Bryston clone instead which was my main project before the Dartzeel appeared into the scene and for which I have the boards designed already.
Last edited:
By all means, Marignos' contribution is very valuable and I will abandon my plans for 3-way configuration. Instead of six channels in one box I will make six monoblocks.
I already have made: KSA-50, Accuphase A60 clones and JLH 1996. I was able to compare NHB-108 with them side-by-side and I am very happy with it. It measures exceptionally good.
Also, I will continue work on NHB-108 and will complete all three variants I have started: 1. basic clone, 2. clone with separate servo and 3. clone with the on-board servo. I also intend to use better op-amp in servo.
I am curious and adamant person who doesn't surrender half-way.
I do not have issues with speakers. In garage I use combination of 8 inch Vifa 8-ohm woofer and tweeter so I can focus exclusively on the amplifier itself.
Judging from my experiences I can conclude:
1. Sound quality is incredibly good (or I really have bad taste and impaired hearing)
2. Measurements are exceptionally good even far beyond the audible region.
I can not understand how an amplifier with so good measurements can sound bad. For me this is beyond comprehension. Over thirty years I am professionally involved in chemical measurements and I know how to appreciate measurements of otherwise "invisible" quantities, almost "abstract" values. What is an alternative? Should I use my tongue and lips to measure concentration of cyanide? Human sensations are deceptive and totally unreliable, therefore we use equipment to evaluate the subject of our observation. Only a good salesperson may say: Let's forget measurements and see what reviewers have to say. Marketing strategies...
To sum-up: this amplifier is still worth-it, modestly speaking.
I already have made: KSA-50, Accuphase A60 clones and JLH 1996. I was able to compare NHB-108 with them side-by-side and I am very happy with it. It measures exceptionally good.
Also, I will continue work on NHB-108 and will complete all three variants I have started: 1. basic clone, 2. clone with separate servo and 3. clone with the on-board servo. I also intend to use better op-amp in servo.
I am curious and adamant person who doesn't surrender half-way.
I do not have issues with speakers. In garage I use combination of 8 inch Vifa 8-ohm woofer and tweeter so I can focus exclusively on the amplifier itself.
Judging from my experiences I can conclude:
1. Sound quality is incredibly good (or I really have bad taste and impaired hearing)
2. Measurements are exceptionally good even far beyond the audible region.
I can not understand how an amplifier with so good measurements can sound bad. For me this is beyond comprehension. Over thirty years I am professionally involved in chemical measurements and I know how to appreciate measurements of otherwise "invisible" quantities, almost "abstract" values. What is an alternative? Should I use my tongue and lips to measure concentration of cyanide? Human sensations are deceptive and totally unreliable, therefore we use equipment to evaluate the subject of our observation. Only a good salesperson may say: Let's forget measurements and see what reviewers have to say. Marketing strategies...
To sum-up: this amplifier is still worth-it, modestly speaking.
Berlusconi how did you made your Accuphase? I work on it now. I have removed the rectifiers on the PCB. I will add more capacitors outside the PCB. Now I wonder what is best supplement. I bought 100 Panasonic's 1200 uf. Want to place them on 2 PCB one for each channel. I also have 4 Kendeil 180.000 uf. Maybe I will place two off them, because in the test of Accuphase the sound is talked about as " meatless" in bass. I will use a 1500 W transformer 36 vac.
We agree that the input stage/VAS with the mods suggested by 3SSS and others works very well in this amp, but the output stage is the weakest point. I have read most of this thread more than once and they have been suggestions for different output stages some including what many say is mandatory, the emitter resistors for the final pair. So it maybe worth to separate the 2 stages in 2 boards
putting the output stage in it's own board to facilitate swapping different output stages.
Just a thought.
putting the output stage in it's own board to facilitate swapping different output stages.
Just a thought.
Last edited: