c37 source
Hi Luke,
you are right - thats the guy who manufactures C37 varnish. You can buy it and also read more info about his theories and products at http://www.ennemoser.com/
Hi Luke,
you are right - thats the guy who manufactures C37 varnish. You can buy it and also read more info about his theories and products at http://www.ennemoser.com/
Vuki,
thanks for the reply, sorry bout the thread jacking here all, but I was quite interested in this stuff hence I filed the article for future reference. Its quite pricy stuff I see but would like to try it sometime in the future. Can any one here comment on applications and effect? I was thinking it may be better used on speaker drivers as opposed to pcbs.
What I find most fascinating about this is how could anyone possibly conceive such a line of research and discover something so clever. Interesting guy.
thanks for the reply, sorry bout the thread jacking here all, but I was quite interested in this stuff hence I filed the article for future reference. Its quite pricy stuff I see but would like to try it sometime in the future. Can any one here comment on applications and effect? I was thinking it may be better used on speaker drivers as opposed to pcbs.
What I find most fascinating about this is how could anyone possibly conceive such a line of research and discover something so clever. Interesting guy.
I'm responseble for that post, and it very simple to see that the gain of the amp is less than 10x when the pot is set to the middle and the feedback resistor value is chossen to be 220K. For high frequencies the DC-blocking cap has very low impedance, so the inverting input of the amp sees a 220K impedance to the output of the amp and a 10K in series with 50K||50K (=25K) equals 35K. So gain is less than 10x. The datasheet says that the thing should be operated at gain > 10x.
I ommited the 100k in my first attempt to build a GainClone and the thing oscilated, got very hot, etc. I realized than that the thing was operating at unity gain.. then I installed a 10K resistor to ground instead of the volume pot. Everything runs smoothly now...
gr,
Thijs
I ommited the 100k in my first attempt to build a GainClone and the thing oscilated, got very hot, etc. I realized than that the thing was operating at unity gain.. then I installed a 10K resistor to ground instead of the volume pot. Everything runs smoothly now...
gr,
Thijs
Hi,
Actually I don't use a vulome pot. From the output of my pre-amp/headphone amp (Rout is only a few Ohms) the signal goes directly (DC-coupled) to the 10K resistor to ground and from the same node 10K to the inverting input, which is connected to the amp output node through a 220K.
I hope this explains it more clearly,
gr,
Thijs
Actually I don't use a vulome pot. From the output of my pre-amp/headphone amp (Rout is only a few Ohms) the signal goes directly (DC-coupled) to the 10K resistor to ground and from the same node 10K to the inverting input, which is connected to the amp output node through a 220K.
I hope this explains it more clearly,
gr,
Thijs
In my metal body gainclone I was using 50K pot. It tended to oscillate at low volumes until I shielded the case and then it stopped.
Yesterday, I was running first time my acrylic gainclone. I'm using 100k Alps Black Beauty pot at the input. I had the Hovland caps after the pot. The whole amp was oscillating like crazy, except at the maximum volume setting. Initially I didn't know what was going on, there was not much different between this version and the previous one, yet this one was giving me a big trouble. I started to play with grounds and it came out that the pot, the heatsinks, the whole circuit has to be shielded and then the oscillations disappear (maybe a bit at very low volume settings, but this doesn't bother me). So I shielded whatever I could and the amp is fine now.
I also removed coupling caps and there is only 20mV of offset.
Yesterday, I was running first time my acrylic gainclone. I'm using 100k Alps Black Beauty pot at the input. I had the Hovland caps after the pot. The whole amp was oscillating like crazy, except at the maximum volume setting. Initially I didn't know what was going on, there was not much different between this version and the previous one, yet this one was giving me a big trouble. I started to play with grounds and it came out that the pot, the heatsinks, the whole circuit has to be shielded and then the oscillations disappear (maybe a bit at very low volume settings, but this doesn't bother me). So I shielded whatever I could and the amp is fine now.
I also removed coupling caps and there is only 20mV of offset.
Have you checked out this site?
There's another guy who "knows" that all of your gainclones sound bad - he can see it from the schematic
http://www.hilberink.nl/amps/lm3875schema.htm
(BTW I haven't yet compared integrated amps to, say, an Aleph 5 clone, but I tend to trust people that build things and listen)
There's another guy who "knows" that all of your gainclones sound bad - he can see it from the schematic
http://www.hilberink.nl/amps/lm3875schema.htm
(BTW I haven't yet compared integrated amps to, say, an Aleph 5 clone, but I tend to trust people that build things and listen)
Hi AMT-freak,
Thanks for the link. I've only briefly looked at the page but I agree with atleast some of his objections, although I do not agree with everything that is published on that page and I do NOT qualify as :'another guy who "knows" that all of your gainclones sound bad' ..
I like my Gainclone, but it's not that special. it IS great value for mnoy and effort! There is however alot of hocus spocus being published around the www about this amp that's not very educated argumented and I feel that the best, most highly regarded DIY audio forum of the www is the rigth place to post those concerns.
Serious discussions about the IC topology is therefore still very welcome!
Maybe this helps,
yesterday I compared my GainClone with a number of commercial amps, among was a Technics SU-500 'New Class A' integrated amp. It sounded very simmilar, and I was intriged to find out it is a IC based amp too! Makes me think.....
gr,
Thijs
Thanks for the link. I've only briefly looked at the page but I agree with atleast some of his objections, although I do not agree with everything that is published on that page and I do NOT qualify as :'another guy who "knows" that all of your gainclones sound bad' ..
I like my Gainclone, but it's not that special. it IS great value for mnoy and effort! There is however alot of hocus spocus being published around the www about this amp that's not very educated argumented and I feel that the best, most highly regarded DIY audio forum of the www is the rigth place to post those concerns.
Serious discussions about the IC topology is therefore still very welcome!
Maybe this helps,
yesterday I compared my GainClone with a number of commercial amps, among was a Technics SU-500 'New Class A' integrated amp. It sounded very simmilar, and I was intriged to find out it is a IC based amp too! Makes me think.....
gr,
Thijs
Recently I played one of my very tiny gainclones ("orginal" non-inverting) to a friend:
First we listened to his amp (I forgot, a technics or Onkyo something), than I hooked the GC. The guy was completely baffled and sank into his seat with an open mouth, speechless.
I mean even if the GC is not the world's best amplifier, it is pretty dammed good, at least for the ridiculous amount of money it costs.
Klaus
ps: Oh, there are more "hate pages" maybe, but the mentioned one is pretty stupid - the guy probably never did actually build any amp himself.
First we listened to his amp (I forgot, a technics or Onkyo something), than I hooked the GC. The guy was completely baffled and sank into his seat with an open mouth, speechless.
I mean even if the GC is not the world's best amplifier, it is pretty dammed good, at least for the ridiculous amount of money it costs.
Klaus
ps: Oh, there are more "hate pages" maybe, but the mentioned one is pretty stupid - the guy probably never did actually build any amp himself.
tschrama,
agreed, I have some objections, too. But this page just fails to convince me. This guy is being too ironic for my liking - and too much into theory.
I've seen STK 30xx voltage amplifiers even in Accuphase gear... Integrated/hybride or discrete" is a similar topic as "solid state or tubes"... You won't convince the fanatics... Yet good results can be achieved with both.
Peter Daniel once mentioned that at least two audiophiles visiting him preferred his gainclone to his Aleph. There are four possible conclusions:
1) Something is wrong with Peter Daniel's Aleph 5 - improbable considering his experience
2) His gainclone is sounding exceptionally good compared to other clones - then I'd like to know why
3) A gainclone sounds indeed better than an Aleph 5 - I don't like the idea, but noone is asking me here...
4) The impression that his gainclone sounds better than his Aleph 5 is false, because the difference they heard was caused by the different room / speakers / DACs / whatever - for the moment, this is what I assume...
I recently finished my first Aleph 5 channel and did some (mono) listening test... Hell, I'd be surprized if you get better sound by a circuit even simpler!
(Ok, the LM3875 itself is not that simple, but building an amplifier with it is)
agreed, I have some objections, too. But this page just fails to convince me. This guy is being too ironic for my liking - and too much into theory.
I've seen STK 30xx voltage amplifiers even in Accuphase gear... Integrated/hybride or discrete" is a similar topic as "solid state or tubes"... You won't convince the fanatics... Yet good results can be achieved with both.
Peter Daniel once mentioned that at least two audiophiles visiting him preferred his gainclone to his Aleph. There are four possible conclusions:
1) Something is wrong with Peter Daniel's Aleph 5 - improbable considering his experience
2) His gainclone is sounding exceptionally good compared to other clones - then I'd like to know why
3) A gainclone sounds indeed better than an Aleph 5 - I don't like the idea, but noone is asking me here...
4) The impression that his gainclone sounds better than his Aleph 5 is false, because the difference they heard was caused by the different room / speakers / DACs / whatever - for the moment, this is what I assume...
I recently finished my first Aleph 5 channel and did some (mono) listening test... Hell, I'd be surprized if you get better sound by a circuit even simpler!
(Ok, the LM3875 itself is not that simple, but building an amplifier with it is)
Hi,
I think a lot of the Gainclone "buzz" comes from the fact that it is an accessible project, that sounds rather good and does not require an EE Degree to complete.
I am not sure what you mean. The IC has an ABSOLUTELY BOG STANDARD quasicomplementary output stage which makes sure that the LIMITING factors for the "speed" of both amplifier halves are equal, unlike with most complementary designs.
The rest again is BOG STANDARD, with the exception of buffering each and every base of an amplification stage with an emitter follower and extensive current source/current mirror loading.
There is a PNP Differentail Amp (inputs buffered and with plenty of local degenration), a NPN Voltage Amplifier (input buffered current source collector load, again local degneration) and an Output stage that reminds of any Naim Amplifier since the 1970's. Read D. Selfs Amplifier book and check out the "blameless" Class B Amplifier, well the LM3875 is just such one, on a chip and with a quasi complemenatary output stage.
NOTHING UNUSUAL AT ALL. Discrete amplifiers have been build on the same basic scheme for decades, just the consequent minimisation of current loops (monolithic IC), linearisation of the circuit (degeneration, current sources, current mirror load) and making it an IC that tolerates +/-40V Rails are RELATIVELY new.
Anything else you want to know?
L8er T
I think a lot of the Gainclone "buzz" comes from the fact that it is an accessible project, that sounds rather good and does not require an EE Degree to complete.
I am not sure what you mean. The IC has an ABSOLUTELY BOG STANDARD quasicomplementary output stage which makes sure that the LIMITING factors for the "speed" of both amplifier halves are equal, unlike with most complementary designs.
The rest again is BOG STANDARD, with the exception of buffering each and every base of an amplification stage with an emitter follower and extensive current source/current mirror loading.
There is a PNP Differentail Amp (inputs buffered and with plenty of local degenration), a NPN Voltage Amplifier (input buffered current source collector load, again local degneration) and an Output stage that reminds of any Naim Amplifier since the 1970's. Read D. Selfs Amplifier book and check out the "blameless" Class B Amplifier, well the LM3875 is just such one, on a chip and with a quasi complemenatary output stage.
NOTHING UNUSUAL AT ALL. Discrete amplifiers have been build on the same basic scheme for decades, just the consequent minimisation of current loops (monolithic IC), linearisation of the circuit (degeneration, current sources, current mirror load) and making it an IC that tolerates +/-40V Rails are RELATIVELY new.
Anything else you want to know?
L8er T
Hi all,
AMT-freak:
I've taken Peter very serious and I trust his opinion. This does however raises some intriging questions about the perceived quality of the amp. Does our situation mean that 'small signal paths' are much more importand? or that top-quality components and construction make all the differnce? My guess is a person can get spoiled with top-quality audio and some refressing differnce of menu can be perceived as top quality. Much like my experience with wine
Kuei Yang Wang: I have never seen an amplifier project that requires an EE diploma to complete, but I understand the point your making. However am glad I build mine, but would like the WWW.diyaudio.com community to be the place for some serious discussion about this amp.
Your imply that the circuit is actually very standard, that no discussion is needed. Maybe it's my fault of lacking the ability to inspire others to take on such discussion. Well, not everybody can be a Pass or Grey etc etc.. but I would like some educated remakrs about the influence of the emitor folowers. To my knowlegde they introduce more distortion than that they increase gain and thus feedback, so their influence is one of increasing distortion, or isn't it? What;s the function of the 1.1K resistor in the LTP? What's your opinion of the amount of Quensient current? Infact, I've never seen such a design for discrete components, why not? Whay does everybody use Vbeamplifiers instead of diodes for biasing? Why does nobody use those emittor folowers, and why does NS use them? Would would you change to the design if it was posted as a discrete design?
well, I hope some of you can save this thread from becoming meaningless and discuss the LM3875, but seriously....
gr,
Thijs
AMT-freak:
I've taken Peter very serious and I trust his opinion. This does however raises some intriging questions about the perceived quality of the amp. Does our situation mean that 'small signal paths' are much more importand? or that top-quality components and construction make all the differnce? My guess is a person can get spoiled with top-quality audio and some refressing differnce of menu can be perceived as top quality. Much like my experience with wine
Kuei Yang Wang: I have never seen an amplifier project that requires an EE diploma to complete, but I understand the point your making. However am glad I build mine, but would like the WWW.diyaudio.com community to be the place for some serious discussion about this amp.
Your imply that the circuit is actually very standard, that no discussion is needed. Maybe it's my fault of lacking the ability to inspire others to take on such discussion. Well, not everybody can be a Pass or Grey etc etc.. but I would like some educated remakrs about the influence of the emitor folowers. To my knowlegde they introduce more distortion than that they increase gain and thus feedback, so their influence is one of increasing distortion, or isn't it? What;s the function of the 1.1K resistor in the LTP? What's your opinion of the amount of Quensient current? Infact, I've never seen such a design for discrete components, why not? Whay does everybody use Vbeamplifiers instead of diodes for biasing? Why does nobody use those emittor folowers, and why does NS use them? Would would you change to the design if it was posted as a discrete design?
well, I hope some of you can save this thread from becoming meaningless and discuss the LM3875, but seriously....
gr,
Thijs
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.