Hugh,
No, it doesn't work that way in my country, it is so hard to convince people to buy your creation. People here who likes to invest in audio stuff are of two kinds, rich hobbyist buys their stuff abroad, popular brands new models kind of thing. The average folks likes to buy China made amplifiers because they are cheaper and can be easily replaced when it gets broken (when done torturing the thing). I was a member of one local forum that also discusses audio stuff and most members loves to scrutinize then ridicule the design and simply dismiss it when they get bored from questioning the obvious.
I once spent an effort to one design polishing it with helps from this forum. I shared all of my work in good faith including gerber files to a fellow citizen. He managed to fabricate the boards but since he wanted to use it for educational purposes to be introduced in classes, he needs a build documentation, problem though no one is willing to help. There are however a few hobbyist who had the interest to buy a kit but cannot promise of helping with the build documentation.
A few months after and still no one came forward to help, the boards goes to the thrash bin.
It broke my heart.
[but I'm glad I'm no engineer]
Sa muli,
Albert
This one?
http://www.diyaudio.com/forums/soli...e-feedback-me-wrong-question.html#post1950112
I'm still trying to find the one DD mentions. Not quite sure what to search for. There is a thread discussing a Putzeys paper on NFB...
Is Putzeys a frequenter of this forum?
Are you including Australia in "Far East"?
This forum is well and truly "vendor" oriented. I get the distinct impression that most members are "in the business" in one way or another.
I am a little skeptical that this forum is full of entrepreneurs looking for ideas to make serious hifi. Otherwise I'd get more messages asking for help. I reckon there may be quite a few kit vendors, perhaps Chinese, selling to beginners who just want cheap, easy to make kits that work ok. Also folks trying to sell books, etc.
Gareth, this is the other thread that might pertain: http://www.diyaudio.com/forums/lounge/306022-bruno-putzeys-paper-negative-feedback.html#post5037437
I'll comment on Putzey's paper wrt the assertion that loop NFB is the same as nested, if you wish.
His Class D system relies on using 5 poles (I think) and zeroes, closely packed, to allow very high loop gain at 20kHz whilst dramatically reducing it well before his switching amp loses phase margin. A roller-coaster of a phase response no doubt. So I suppose he has an interest in arguing that extreme, multi-pole shaping of the open loop transfer function is just as good as using nested feedback in an analogue amplifier. Perhaps the motivation for this paper?
I'll comment on Putzey's paper wrt the assertion that loop NFB is the same as nested, if you wish.
His Class D system relies on using 5 poles (I think) and zeroes, closely packed, to allow very high loop gain at 20kHz whilst dramatically reducing it well before his switching amp loses phase margin. A roller-coaster of a phase response no doubt. So I suppose he has an interest in arguing that extreme, multi-pole shaping of the open loop transfer function is just as good as using nested feedback in an analogue amplifier. Perhaps the motivation for this paper?
Bigun, the first impression of the person who made these measurements was that: these amplifiers are very similar in sound. If there are differences, they were not revealed on all compositions. In his system, Yamaha NS-10 acoustic systems were used, which are considered as a monitor level. I would personally characterize the sound of the NAIM system as: with a light touch of "haze" or "veil". The tonal balance is perceived very smoothly, without underlining any band of audio frequencies.
But, there is one "But." This can only be achieved by using the NAIM preamplifier. Without it, there is a shift in tonal balance-up. The sound becomes more "bright", "lightweight" and even, surprisingly - "fast". So, for my taste, the NAIM system - is the use of their preamplifier, which focusses and balances the overall sound and impression in an incomprehensible way. Or a simple formula: NAIM = NAC + NAP.
With respect, Nikolai.
But, there is one "But." This can only be achieved by using the NAIM preamplifier. Without it, there is a shift in tonal balance-up. The sound becomes more "bright", "lightweight" and even, surprisingly - "fast". So, for my taste, the NAIM system - is the use of their preamplifier, which focusses and balances the overall sound and impression in an incomprehensible way. Or a simple formula: NAIM = NAC + NAP.
With respect, Nikolai.
Member
Joined 2009
Paid Member
Nikolai, that's very interesting. I thought the pre+pwr amp combo was driven more by a need to control ultrasonic frequencies to prevent cross-conduction and other instabilities ( as well as drive sales of course). If there is also a tonal requirement then I'm quite intrigued. After I build my NAP 160 a pre-amp might also be required.
The parts from Digikey have shipped by the way; now waiting for the postman.
p.s. not sure what Bruno is doing these days since he left Hypex.
The parts from Digikey have shipped by the way; now waiting for the postman.
p.s. not sure what Bruno is doing these days since he left Hypex.
Bigun, the main contributor to sound changes, is the second part of the preamplifier circuit. This is part of the circuit with 5-transistors with a gain of about 10. The first part of the circuit is a high-pass filter (outside the audio range). You can delete this part of the schema.
Let me quote a link where I had a dialogue with the owner of a genuine NAIM system.
Http://soundex.ru/forum/index.php?/topic/42977-%D1%81%D0%B1%D0%BE%D1%80%D0%BA%D0%B0%%%%%%% %%% BB% D0% BE% D0% BD% D0% B0-naim-nap-200 / & page = 21
This is a subforum of our Russian "NAIM fans". Pay attention to my dialogue (№520-523). The user "sls" confirmed all my observations, in cases: with and without a preamplifier.
The only inconvenience is automatic translation. But, I think that you will understand the essence of the dialogue. Good luck in creating your clone-NAM!
Let me quote a link where I had a dialogue with the owner of a genuine NAIM system.
Http://soundex.ru/forum/index.php?/topic/42977-%D1%81%D0%B1%D0%BE%D1%80%D0%BA%D0%B0%%%%%%% %%% BB% D0% BE% D0% BD% D0% B0-naim-nap-200 / & page = 21
This is a subforum of our Russian "NAIM fans". Pay attention to my dialogue (№520-523). The user "sls" confirmed all my observations, in cases: with and without a preamplifier.
The only inconvenience is automatic translation. But, I think that you will understand the essence of the dialogue. Good luck in creating your clone-NAM!
Member
Joined 2009
Paid Member
I'd be quite happy to see my designs cloned, at least somebody would be building them !
Well, a big box arrived today. Yeah!
I think Digikey employs Santa's elves during the off season. I don't even care if this thing works - see those shiny caps, feel the weight of those Sanken power devices. Just lovely.
Well, a big box arrived today. Yeah!
I think Digikey employs Santa's elves during the off season. I don't even care if this thing works - see those shiny caps, feel the weight of those Sanken power devices. Just lovely.
Member
Joined 2009
Paid Member
the tube amp that I use now I won't post. I have no plans to sell it, but it's my baby. I don't want to read how bad it is
oh do share, I promise to laugh loudly
One of these days I'll share the secret design of a Shindo Pentode amplifier.
Whilst the NAP200 power amplifier referred to is a similar electrical design to NAP140, I think its recommened preamp is the later model NAC152X. The NAP140 matching preamp is the NAC 32.5 or alternatively, NAC 42.5. Of course, any type will work OK, but if you are looking to polish the overall response flatness and sound quality authentically, these are the recommended combinations. The source of the images, Ced. Taylor by courtesy of acoustica.org.uk, refers to the differences between the preamp models only as HF filtering with no suggestion that there is any compensating boost in the amplifiers. I can't see any evidence of it either.
Early NAC 32.5 Gainboard schematic:
and more recent version:
The buffer board:
Last edited:
Member
Joined 2009
Paid Member
that last one, I built a unity gain one of those: http://www.diyaudio.com/forums/headphone-systems/275180-grasshopper.html
Well at least it was quite similar.
Well at least it was quite similar.
Bigun,
Your TGM2.7 is actually part inspiration of my work here [CFP input, bootstrap VAS] and also by studying Hugh's AKSA55 harmonics structure.
http://www.diyaudio.com/forums/solid-state/300109-bootstrapped-amps-dx-aksa-rca-etc-23.html It was a pleasant learning experience running it in sim. Thanks to both of you.
Sa muli,
Albert
Member
Joined 2009
Paid Member
Interesting to know !
Well, truth but be told, my TGM2 was inspired by Hugh's Lifeforce amplifier. It was a huge surprise to me how simply changing the LTP to a CFP without any other significant changes brought such a profound change in the sound. Of course now I know that the change was significant in terms of using gobs of local feedback to greatly improve the linearity.
I should also tell you that I was not that happy with the sound of CFP at the LTP. I later tried CFP in the VAS of a JLH'69 and the same signature came through clearly on that too - I didn't like it there either and took it out. I did try it in the VAS position of my TGM6-version 1 and it was just OK so for version 2 I took a different approach. I also tried the CFP in the driver stage, this was my TGM3 and it worked fine there but it was used as a follower. And I tried it again in my TGM8 in the output, where it is phenomenal. I never did optimize the CFP properly when used in the front-end of an amplifier due to lack of knowledge. But something about the sound I could never warm to and as time is limited I have adopted a simple rule of thumb not to use CFP in the input stages of an amplifier. In fact I am nervous about using it in any voltage gain stage and prefer it in a follower power stage instead where it is one of my favourite topologies.
Well, truth but be told, my TGM2 was inspired by Hugh's Lifeforce amplifier. It was a huge surprise to me how simply changing the LTP to a CFP without any other significant changes brought such a profound change in the sound. Of course now I know that the change was significant in terms of using gobs of local feedback to greatly improve the linearity.
I should also tell you that I was not that happy with the sound of CFP at the LTP. I later tried CFP in the VAS of a JLH'69 and the same signature came through clearly on that too - I didn't like it there either and took it out. I did try it in the VAS position of my TGM6-version 1 and it was just OK so for version 2 I took a different approach. I also tried the CFP in the driver stage, this was my TGM3 and it worked fine there but it was used as a follower. And I tried it again in my TGM8 in the output, where it is phenomenal. I never did optimize the CFP properly when used in the front-end of an amplifier due to lack of knowledge. But something about the sound I could never warm to and as time is limited I have adopted a simple rule of thumb not to use CFP in the input stages of an amplifier. In fact I am nervous about using it in any voltage gain stage and prefer it in a follower power stage instead where it is one of my favourite topologies.
Last edited:
Member
Joined 2009
Paid Member
So you did, though i had to think about it
Actually it's a bit more subtle - I've really used it as an impedance multiplier. The CFP takes the 'load off' of R1. It worked superbly but I could never live with the ragged sound of my Grado headphones. I sold them on and currently have no headphones to use with that amp.
oh do share, I promise to laugh loudly
One of these days I'll share the secret design of a Shindo Pentode amplifier.
Oooh, don't go engineer on me now!
I know one of the Stereophile writers is comparing anything that he reviews to a Shindo that he owns. So ... more shared secrets, please!
Really not much interesting to share about my tube amp. It started as an Oddwatt with regulated HV supply. Currently I'm converting it into LTP input with transistor CCS, class A push-pull UL output and a little feedback. Looks decent on my poor-man scope FFT and make me listening to music for hours.
Also gives me the pleasures of point-to-point soldering.
Maybe one day I'll borrow some of the Wavesurfers or a a proper spectrum analyzer from work and give it a good measurement.
That brings me to questions for Nikolai:
Nice to see some comparative measurements of a Naim clone.
I can't recognise what did your friend use for the harmonics measurements of the two amps. Is that a scope FFT?
I see the signal level is low at -3db. Why is that? Is there attenuation added at the measuring equipment inputs or you just tested at low power?
What do you mean by equivalent load? Do you mean you have built a standard load for testing amps which represents an average loudspeaker? If that's the case - very neat! I'm always considering of building one, but for now other stuff take priority.
Last edited:
That's more a problem for entrepreneurs. For individual enjoyment it's a different matter. And if the market can be educated, it might not be an issue for entrepreneurs too.
People might have difficulty to judge which ones are good amps, average amps, or great amps using ears alone. But when they have a system and they experience enjoyment in using it, they will remember the system and keep it.
If a product has this characteristics, the market can be educated, the vendor will help ensuring that no mistake is done in implementation (cables are provided, setup done by vendor, etc). And as you can see it, Naim can do that successfully.
So is this only marketing and psychological and no science involved? No, this is science. There are rules how to make an amp that can give high level of longtime enjoyment.
Like they say, it's about time management. First things first. Priority.
Someone said the VAS will not have speaker feedback. My amp I show @No772 does. I suspect it is easier to drive from the LTP side also. It's a shame my son lost all my files of the test results. He built me a new computer and didn't think the old one worth keeping ( I transfered all your files dad, even the ones that are now lost ? ).
Self I seem to remember speculates that unwanted feedbacks do exist and should be thought of. My 40V to the VAS and 50V to the outputs of the NAP140 takes up Self's advice. The easy one that comes to mind is MOSFET's seem to self oscillate. I measured the gate happily giving a 5 MHz sine wave. The scope is at it's limits there so if truely sine I can not say. The 220R gate stopper makes it vanish at the VAS and nothing seen at the source regardless. The 5 MHz is always there ( grid stopper will shift the frequency a bit ), I specualte that it could become positive feedback and re-enter in many places. As Self says it's a brave man who measures this as often the FET tries to leave the planet. Thus the gate stopper may be fitted for a not too obvious reason. Audiophiles will assume a high grade stopper worth the money. In fact it is claimed the inductance of the stopper is no small factor. Glad someone told me that as carbon compositions are cheap and not too bad this far up an amp when hiss. BTW. I have no idea if the scope induced the 5 MHz, my hunch is it didn't. There was no heating of the FET. My scope floats which might have saved the day.
Time management wouldn't stop me. Making the shed tidy would. If I redid my No 772 amp I would use a Cascode VAS. I bought some 2SD756 with gain of 135. They are not 2SD756, 2N5551 is my guess.
Bootstrap VAS CCS is a very interesting project. My only doubt is the output stage is dirty, would someone like to speculate as to in phase distortion with crossover spikes good or bad. I can sort of see an arguement that it might be good. It is just possible a standard CCS less good in some ways. Distortion I measured with the Bootstraped CCS are good at > 10 kHz which makes me wonder.
Self makes a rather cryptic reference about degenerated feedback factors. He shows how we might do that and mostly I agree it has no value. However if one plays with the Bootstrap positive feedback capacitor it is remarkable how small it can be and still offer good distortion. One feels a nice 10uF 250V polyester is the right choice. If we say 3K3/2 = 1K65, looks good to me. Even 1uF is almost OK, I had to really look hard to see it was worse. 40/6K6= 6 mA
The voltage gain of MOSFET's might be worse than 80%. Actually that's not true. That is sine wave high power testing. On real music the MOSFET Bootstrap CCS is no worse than transistor CCS except perhaps a doubling of the miniscule 2nd harmonic, the very distortion we might want. I found the Bootstrap CCS will cause trouble on the Complimentary feedback pair and needs a small resistance added to tame the clipping. The Naim I suspect would need this. 100R should be about right. Gains up to 93% seem OK. The Baxendale diode might help.
Self I seem to remember speculates that unwanted feedbacks do exist and should be thought of. My 40V to the VAS and 50V to the outputs of the NAP140 takes up Self's advice. The easy one that comes to mind is MOSFET's seem to self oscillate. I measured the gate happily giving a 5 MHz sine wave. The scope is at it's limits there so if truely sine I can not say. The 220R gate stopper makes it vanish at the VAS and nothing seen at the source regardless. The 5 MHz is always there ( grid stopper will shift the frequency a bit ), I specualte that it could become positive feedback and re-enter in many places. As Self says it's a brave man who measures this as often the FET tries to leave the planet. Thus the gate stopper may be fitted for a not too obvious reason. Audiophiles will assume a high grade stopper worth the money. In fact it is claimed the inductance of the stopper is no small factor. Glad someone told me that as carbon compositions are cheap and not too bad this far up an amp when hiss. BTW. I have no idea if the scope induced the 5 MHz, my hunch is it didn't. There was no heating of the FET. My scope floats which might have saved the day.
Time management wouldn't stop me. Making the shed tidy would. If I redid my No 772 amp I would use a Cascode VAS. I bought some 2SD756 with gain of 135. They are not 2SD756, 2N5551 is my guess.
Bootstrap VAS CCS is a very interesting project. My only doubt is the output stage is dirty, would someone like to speculate as to in phase distortion with crossover spikes good or bad. I can sort of see an arguement that it might be good. It is just possible a standard CCS less good in some ways. Distortion I measured with the Bootstraped CCS are good at > 10 kHz which makes me wonder.
Self makes a rather cryptic reference about degenerated feedback factors. He shows how we might do that and mostly I agree it has no value. However if one plays with the Bootstrap positive feedback capacitor it is remarkable how small it can be and still offer good distortion. One feels a nice 10uF 250V polyester is the right choice. If we say 3K3/2 = 1K65, looks good to me. Even 1uF is almost OK, I had to really look hard to see it was worse. 40/6K6= 6 mA
The voltage gain of MOSFET's might be worse than 80%. Actually that's not true. That is sine wave high power testing. On real music the MOSFET Bootstrap CCS is no worse than transistor CCS except perhaps a doubling of the miniscule 2nd harmonic, the very distortion we might want. I found the Bootstrap CCS will cause trouble on the Complimentary feedback pair and needs a small resistance added to tame the clipping. The Naim I suspect would need this. 100R should be about right. Gains up to 93% seem OK. The Baxendale diode might help.
There is a man I should know called Alan Bond as my friend Pippa wrote up his notes. Pippa unlike me isn't dislexic, unlike me she has no capacity to remember things that would change the World. I will keep my dislexia I think, I remember things rather than words.
I mention Bond as a rocket might only get a 2% advantage of fuel weight to lift. Bond came up with basically a drinks can fuel tank where the fuel gave strength to the tank. This gave advantages. The British government in it's ultimate wisdom scapped it as too slow to get ready for use. None of the great brains saw that putting satellites up would make money.
Bond now works on to make a hybrid rocket motor. He didn't take retirement as he knew if he didn't see it through it might never happen. London to Sidney in 45 minutes.
The reason I think Bond was rejected was his inventions look too simple. Simple is not the same as easy. Bonds big problem wasn't compressing enough air to repalce liquid fuel. It has been stopping ice build up. I dare say his rocket plane could use Kerosene as we do when jets. Cheap space travel where space is incidental. Bond thought of a freighter to test the concept before passengers risk their lives. To go England to Spain would take longer as it requires an orbit. Even so an hour and a half.
I mention Bond as a rocket might only get a 2% advantage of fuel weight to lift. Bond came up with basically a drinks can fuel tank where the fuel gave strength to the tank. This gave advantages. The British government in it's ultimate wisdom scapped it as too slow to get ready for use. None of the great brains saw that putting satellites up would make money.
Bond now works on to make a hybrid rocket motor. He didn't take retirement as he knew if he didn't see it through it might never happen. London to Sidney in 45 minutes.
The reason I think Bond was rejected was his inventions look too simple. Simple is not the same as easy. Bonds big problem wasn't compressing enough air to repalce liquid fuel. It has been stopping ice build up. I dare say his rocket plane could use Kerosene as we do when jets. Cheap space travel where space is incidental. Bond thought of a freighter to test the concept before passengers risk their lives. To go England to Spain would take longer as it requires an orbit. Even so an hour and a half.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.