Wise as usual, Thorsten.
And, always surfing around for variants of Hiraga's project, what about MJAMP (http://www.homestead.com/whaan/files/page5.html)?
Or the incredible performant (too much for audio?) "Son of Godzilla" buffer described by Jim Williams in Linear Technology AN18?
And, always surfing around for variants of Hiraga's project, what about MJAMP (http://www.homestead.com/whaan/files/page5.html)?
Or the incredible performant (too much for audio?) "Son of Godzilla" buffer described by Jim Williams in Linear Technology AN18?
moer amplis to add to the list
always from the ubiquitus Hiraga. From a forum serach:
http://www.diyaudio.com/forums/showthread.php?postid=489287#post489287
Seems that the Nemesis can be a good full solid state '300B sound' amplifier.
Also this project had good review:
http://www.audiofanatic.it/Amplificatori/Monostato/Monostato_Comi.html
always from the ubiquitus Hiraga. From a forum serach:
http://www.diyaudio.com/forums/showthread.php?postid=489287#post489287
Seems that the Nemesis can be a good full solid state '300B sound' amplifier.
Also this project had good review:
http://www.audiofanatic.it/Amplificatori/Monostato/Monostato_Comi.html
little test
is it possible to have a solid state amplifier with a 300B SE comparable performances, without expensive or obsolete parts and big heat dissipator?
Starting from lot of suggestion give in this forum, I spiced the below reported circuit and found after some value tweaks that it can privide 10W at 3,3% distortion, with THD raising monotonically with power and spectral decay similar to 300B SE at every power level.
For example at 7.44V rms out over 8ohm 2nd harmonic is 131mV, 3rd 71mV 4th 1,28mV.
Quiescent current is stable from 0 to 100C and assure class A operation within the first watt.
Feedback factor is moderate, gain sufficient to have input sensitivity worst case of 350mV.
Is spice too optimistic?
is it possible to have a solid state amplifier with a 300B SE comparable performances, without expensive or obsolete parts and big heat dissipator?
Starting from lot of suggestion give in this forum, I spiced the below reported circuit and found after some value tweaks that it can privide 10W at 3,3% distortion, with THD raising monotonically with power and spectral decay similar to 300B SE at every power level.
For example at 7.44V rms out over 8ohm 2nd harmonic is 131mV, 3rd 71mV 4th 1,28mV.
Quiescent current is stable from 0 to 100C and assure class A operation within the first watt.
Feedback factor is moderate, gain sufficient to have input sensitivity worst case of 350mV.
Is spice too optimistic?
Attachments
Re: little test
No. I have al almost identical circuit, posted here a while ago that pushes the same topology but different output configuration (mine uses complimentory mosfets driven by bjts). it showed equally good performance in spice and in life.
It can be driven by either +/- rails or a single rail. and can drive very low impedance.
three draw backs for this confirguration:
a) poor output zero voltage control: it is not uncommon for it to drive +/- 50mv;
b) not particularly efficient: you need about 8v for the mostfets;
c) bad clipping behavior: especially on the negative side.
I have one built up and sing happily right now.
It also runs well with a ltp input stage.
plovati said:
No. I have al almost identical circuit, posted here a while ago that pushes the same topology but different output configuration (mine uses complimentory mosfets driven by bjts). it showed equally good performance in spice and in life.
It can be driven by either +/- rails or a single rail. and can drive very low impedance.
three draw backs for this confirguration:
a) poor output zero voltage control: it is not uncommon for it to drive +/- 50mv;
b) not particularly efficient: you need about 8v for the mostfets;
c) bad clipping behavior: especially on the negative side.
I have one built up and sing happily right now.
It also runs well with a ltp input stage.
Attachments
Kuei Yang Wang said:
Mr. Hilberinks "riped" you hard on Page 10, didn't he?
http://www.hilberink.nl/amps/lm3875schema.htm
I find his history on solid state amps quite an interesting read. It certainly brought back memories.
Konnichiwa,
I find this "fan-page" really amusing. Especially when he pans the "inverting mode". Clearly not only I, but also the NS designers (who mention it in the later generation Overture Amp datasheets as advantageous) know not what we are doing.
The reason that Mr. Hilberink takes exception to me is that I publically noted that an earlier version of his pages showed not much of interesting in terms of history, just generic designs, while praising this poorely conceived stuff to the skys as achievements.
I notice that since he has (somewhat grudgingly it seems) given space to less than conventional designs (Hiraga) even though he cannot resist showing "improved" versions, which in fact improve very little, if anything, on the originals.
He has a very easily bruised ego, this "Mr. Tannoy Gold" and he seems to think that others have the same and that stooping to such low propaganda will actually bother me (it does not in the least, it makes me chuckle).
Sayonara
tlf9999 said:
I find this "fan-page" really amusing. Especially when he pans the "inverting mode". Clearly not only I, but also the NS designers (who mention it in the later generation Overture Amp datasheets as advantageous) know not what we are doing.
tlf9999 said:
The reason that Mr. Hilberink takes exception to me is that I publically noted that an earlier version of his pages showed not much of interesting in terms of history, just generic designs, while praising this poorely conceived stuff to the skys as achievements.
I notice that since he has (somewhat grudgingly it seems) given space to less than conventional designs (Hiraga) even though he cannot resist showing "improved" versions, which in fact improve very little, if anything, on the originals.
He has a very easily bruised ego, this "Mr. Tannoy Gold" and he seems to think that others have the same and that stooping to such low propaganda will actually bother me (it does not in the least, it makes me chuckle).
Sayonara
Konnichiwa,
If I had, i would probably have set up a webpage going to great length in questioning his character, intention and abilities. As I don't, I have not and generally ignore him, unless someone bring him up....
Generally I subscribe to "What does it concern a tree if a pig scratches against it?"
Sayonara
lumanauw said:
If I had, i would probably have set up a webpage going to great length in questioning his character, intention and abilities. As I don't, I have not and generally ignore him, unless someone bring him up....
Generally I subscribe to "What does it concern a tree if a pig scratches against it?"
Sayonara
Kuei Yang Wang said:
that makes a lot of sense as most amps are used in inverting mode. But I personally don't think that was such a big deal.
I was quite amazed by the flip-flop of those experts on inverting vs. non-inverting on the chip forum. It started as all pro non-inverting and a couple years later it is all inverting. And you have countless people finding tons of theories / personal experience to support their assertions. Well, that's a story for another day.
I do think he made quite a few good points about the use of small capacitors and the output serial resistor. It simply doesn't make a lot of sense to me to go with ultra small rail capacitors and "un"dampen the output.
but I am happy to know why you disagree with him on that.
Konnichiwa,
Well, given that "damping" is limited by the DCR of the voicecoil (an appreciable proportion of the total Z), damping is a huge red herring. Thus the small value R instead of RLC, easier to build for the DIY'er. Similar output resistoirs in amplifiers by Naim, Linn, DNM and Jeff Rowland seem to have done little to hurn the sound of their Amp's, and what's good enough for them....
Small PSU capacitors, beats me, they just sound better, why, no real idea. Some wild guesses - the LM3875 has so much NFB that even a huge ripple on the rails does little. In some ways I think it is related to the transient nature of music and the fact that small capacitors have less inductance (including wiring) up where the LM3875 becomes twitchy and there seems little benefit from huge capacitance, subjectively. In other words, it works, sounds fine, why change?
Sayonara
tlf9999 said:
Well, given that "damping" is limited by the DCR of the voicecoil (an appreciable proportion of the total Z), damping is a huge red herring. Thus the small value R instead of RLC, easier to build for the DIY'er. Similar output resistoirs in amplifiers by Naim, Linn, DNM and Jeff Rowland seem to have done little to hurn the sound of their Amp's, and what's good enough for them....
Small PSU capacitors, beats me, they just sound better, why, no real idea. Some wild guesses - the LM3875 has so much NFB that even a huge ripple on the rails does little. In some ways I think it is related to the transient nature of music and the fact that small capacitors have less inductance (including wiring) up where the LM3875 becomes twitchy and there seems little benefit from huge capacitance, subjectively. In other words, it works, sounds fine, why change?
Sayonara
Thorsten,
That is interesting. Other people report that the bass suffers with the smaller caps, and of course THEY have a good explanation (for them) to go with it. 😀
Jan Didden
Konnichiwa,
Absolutely, but they usually seem to use fairly low efficiency speakers, whereas mine are rather high efficiency, this might suggest the difference in LF behaviour to be related more to overload behaviour, than to the amount of capacitance per se.
It is a complex subject and measurements do not seem to help. Even with very small value PSU capacitors the Amplifier measures essentially fine, at LF, while not near clipping.
Sayonara
janneman said:
Absolutely, but they usually seem to use fairly low efficiency speakers, whereas mine are rather high efficiency, this might suggest the difference in LF behaviour to be related more to overload behaviour, than to the amount of capacitance per se.
It is a complex subject and measurements do not seem to help. Even with very small value PSU capacitors the Amplifier measures essentially fine, at LF, while not near clipping.
Sayonara
Kuei Yang Wang said:
doesn't that suggest that small caps limit the output power of the chip? with a low efficency speaker, the chip has to output more power, thus draining the power supply cap harder -> needs more capacitance.
I think there is so much psudo science in audio that it makes con artists look like a saint, 😉
Konnichiwa,
Sure, but if you only rate the Amp at 25 Watt and it delivers 25 Watt unclipped?
Quite possibly. But what has this to do with the price of tea in china?
And enough audio absolutism to make the son of Schicklegruber look the same.
Remember, both the objectivist and the subjectivists are completely wrong on almost all counts... ;-)
Sayonara
tlf9999 said:
Sure, but if you only rate the Amp at 25 Watt and it delivers 25 Watt unclipped?
tlf9999 said:
Quite possibly. But what has this to do with the price of tea in china?
tlf9999 said:
And enough audio absolutism to make the son of Schicklegruber look the same.
Remember, both the objectivist and the subjectivists are completely wrong on almost all counts... ;-)
Sayonara
Kuei Yang Wang said:
Yes indeed. That is what often frustrates me. Someone says: this amp works fine with me, another says: its ****! There is no indication whether they use wildly different speakers, room sizes, what have you. Such statements are then essentially worthless for the rest of us.
Jan Didden
Hi, TLF9999,
I'm interested in building something like your post #85.
If I use +/0/- rail, the input voltage divider is used or not? And the -speaker taken from 0 or - ?
I'm interested in building something like your post #85.
If I use +/0/- rail, the input voltage divider is used or not? And the -speaker taken from 0 or - ?
Is this by enlarging the output capacitor? for 4 speaker? Is the non polar or 2 polars back to back is better here (I'm thinking about 2x10.000uF +--+ in series, will it give something better? //with 10uF polypropylene maybe?)
Is this after the output C?
Kuei Yang Wang said:
Sure. If you were to run your Krell at 1mw, you could probably get away with just a 0.1uf capacitor in the power supply.
lumanauw said:
you will still need the divider network: if you were to take it off, the base of Q1 will be zero (actually slightly above zero), and the speaker terminal will be at 0.7v. You don't want to put that much DC on your speaker if you are DC-coupling the speaker.
the speaker will still be terminated into the ground.
BTW, you probably have realized this: R2 (output DC adjustment) and R14 (bias adjustment) are 22k multi-turn pots. C2 can be omitted or you can use a 220u-470u cap. You can omit R5, R21, R4, C3 and C7 - they are there for testing purposes.
lumanauw said:
if you go dual rails, you can get rid of the output capacitor all together.
lumanauw said:
the DC drift is before the output capacitor.
Konnichiwa,
Yes, the opposite of this frustrates me though, when some technocrate who does not know ANY of the above dogmatically declares "the capacitors in this amplifier are too small, it cannot work!" (or similar statements about how this or that piece of design or whole design cannot work, while I have sitting here on my rack that most definitly works).
While the context info is useful, it is no absolutely essential, it is merely needed to turn data into information by providing context, the data itself still remains of value and useful EVEN without context.
The blockhead approach on the other hand is sufficiently useless to be severely counter productive.
Sayonara
janneman said:
Yes, the opposite of this frustrates me though, when some technocrate who does not know ANY of the above dogmatically declares "the capacitors in this amplifier are too small, it cannot work!" (or similar statements about how this or that piece of design or whole design cannot work, while I have sitting here on my rack that most definitly works).
While the context info is useful, it is no absolutely essential, it is merely needed to turn data into information by providing context, the data itself still remains of value and useful EVEN without context.
The blockhead approach on the other hand is sufficiently useless to be severely counter productive.
Sayonara
Kuei Yang Wang said:
Both of you are correct. A good thing, when taken to extreme, is going to be bad.
I guess what needs to be done is for people to have an open mind, understanding that our own perspectives aren't the last word, and be ready to accept other people's points of view.
otherwise, we are no different from the morons we so love to hate.
- Status
- Not open for further replies.