I think there's been a lot of misunderstanding, and undue frustration, surrounding what I'm looking for, or asking for... and I apologize to any who have been confused, that's not my intent.fedde said:
All I've been looking for, is to be pointed to good sources of information, so that I can read up and research this myself.
Right on!
I'm not asking any specific questions... there's no solicitation for some "answer".
Yet it seems that some people are frustrated that they can't answer my questions... i've posed none.
I'm just asking for directions.
Metaphorically speaking, I just want to be pointed to "the repository of chip-amp information", and fill out the paperwork to get a library card at the place.
If you look back, I think I've gotten that, to a large degree.
There are lots of great links that have been given in this thread, and even references to books, that serve that purpose.
Hopefully, in retrospect, you can appreciate the links and references that were provided along the way, and their diversity.
I certainly thank all who contributed for that!
I'm not the first one to come to this forum, looking to learn a little more about chip amps...
And I'm sure I won't be the last.
I'm sure this thread will do other chip-amp newbies (and maybe beyond?) service as well in the future.
So, if you have any more references personally bookmarked, or bookshelved... please do post them!
It would be nice if such a thing did exist as "a directory of chip amp information"... even if in hopelessly incomplete form.
power supplies
http://sound.westhost.com/psu-wiring.htm
http://sound.westhost.com/power-supplies.htm
heatsinking
http://sound.westhost.com/heatsinks.htm
basic amplifier circuits
http://sound.westhost.com/dwopa.htm
for the most part, the design notes on www.national.com are also pretty good.
my suggestion -- review the above links, especially the last one, and the lm3886 or lm3875 datasheet and the highpower apps design notes, then formulate a list of questions.
for the most part cloning is an art -- there are some people who love the inverting config over the noninverting. some who use RF input protection, some without. some use the output zobel, some don't. some use 1000uF of supply capacitance, some use more. some use monoblocks, some use 6 channels in one case. some use nice cases, some improvise with copper pipe or other stuff. on all of these subjective things you will either need to test for yourself, or run a poll.
on the technical side, i'd have to say the circuits are pretty easy to understand, if you want an explaination of any of them, any part, then you should ask.
http://sound.westhost.com/psu-wiring.htm
http://sound.westhost.com/power-supplies.htm
heatsinking
http://sound.westhost.com/heatsinks.htm
basic amplifier circuits
http://sound.westhost.com/dwopa.htm
for the most part, the design notes on www.national.com are also pretty good.
my suggestion -- review the above links, especially the last one, and the lm3886 or lm3875 datasheet and the highpower apps design notes, then formulate a list of questions.
for the most part cloning is an art -- there are some people who love the inverting config over the noninverting. some who use RF input protection, some without. some use the output zobel, some don't. some use 1000uF of supply capacitance, some use more. some use monoblocks, some use 6 channels in one case. some use nice cases, some improvise with copper pipe or other stuff. on all of these subjective things you will either need to test for yourself, or run a poll.
on the technical side, i'd have to say the circuits are pretty easy to understand, if you want an explaination of any of them, any part, then you should ask.
How much power can they be built for?
-- 200watts @ 4ohms using 2 lm4870 would be my best guess.
How much output power can I build a gainclone to produce?
-- the best start may be 50 watts per channel, 2 channels. otherwise i'd say the dual lm3886 or single lm4780 based Bridged 120W @ 8ohm amplifer would be another start.
How can I manipulate this in my design, if it's a design variable?
-- purchase a different power transformer, this will change the voltage rails.
Down to what sort of impedances can I build a gainclone to be stable down to?
-- 4ohms.
How can I manipulate this in my design, if it's a design variable?
-- paralleling ICs. this requires a higher degree of matching components though. not commercially feasible. with 2 lm3886 in parallel, you should be able to run a 2ohm loudspeaker, but no one has done so yet.
Is the gainclone best suited for full-range duty?
-- the gainclone's bandwidth extends well beyond the audio spectrum, it will work fine for full range.
Can it be built for subwoofer duty (ie. higher output power requirement applications)?
-- kinda. if you want 1kW, then no. for 100 watts or 200 watts, then maybe.
What is the cost of the componentry?
-- depends. there is a wide range of components as normal and exotic components will work (not explode spontaneously). the cost can be made low or high,
Where can I best obtain the transformers and chips?
-- national has a sample program. since you are a partner in a buisness, i'm sure they would send you samples. in general, you shouldn't abuse this, though i wonder if they just don't write off the samples for taxes... the best learner IC is the lm3875TF -- a 50W IC that has only 5 used pins and an isolated package. www.apexjr.com and www.bgmicro.com have some useful amp parts.
What are the differences between chips?
-- lm3875 is made for 50W and has no mute. the lm3886 has a mute and is made for 60W. the lm4780 and lm4732 are dual amp chips made for 60x2 and 50x2. the lm1875 is a smaller sized IC - 5 pins, and is now obselete IIRC. it was made for 20W useage. some of the ICs are made more for 8ohm use. not sure which ones specifically.
Why would I select one over another?
-- power, size, the 3875TF is my suggested learning IC.
Could a gainclone be built to run off of a 12 volt DC supply?
-- no, the IC will not turn on with less then 24V, or +-12V.
Would this impart any limitations on the level of output power that could be produced by the gainclone?
-- yes, even at +-12V the power is much less then at +-30V.
-- 200watts @ 4ohms using 2 lm4870 would be my best guess.
How much output power can I build a gainclone to produce?
-- the best start may be 50 watts per channel, 2 channels. otherwise i'd say the dual lm3886 or single lm4780 based Bridged 120W @ 8ohm amplifer would be another start.
How can I manipulate this in my design, if it's a design variable?
-- purchase a different power transformer, this will change the voltage rails.
Down to what sort of impedances can I build a gainclone to be stable down to?
-- 4ohms.
How can I manipulate this in my design, if it's a design variable?
-- paralleling ICs. this requires a higher degree of matching components though. not commercially feasible. with 2 lm3886 in parallel, you should be able to run a 2ohm loudspeaker, but no one has done so yet.
Is the gainclone best suited for full-range duty?
-- the gainclone's bandwidth extends well beyond the audio spectrum, it will work fine for full range.
Can it be built for subwoofer duty (ie. higher output power requirement applications)?
-- kinda. if you want 1kW, then no. for 100 watts or 200 watts, then maybe.
What is the cost of the componentry?
-- depends. there is a wide range of components as normal and exotic components will work (not explode spontaneously). the cost can be made low or high,
Where can I best obtain the transformers and chips?
-- national has a sample program. since you are a partner in a buisness, i'm sure they would send you samples. in general, you shouldn't abuse this, though i wonder if they just don't write off the samples for taxes... the best learner IC is the lm3875TF -- a 50W IC that has only 5 used pins and an isolated package. www.apexjr.com and www.bgmicro.com have some useful amp parts.
What are the differences between chips?
-- lm3875 is made for 50W and has no mute. the lm3886 has a mute and is made for 60W. the lm4780 and lm4732 are dual amp chips made for 60x2 and 50x2. the lm1875 is a smaller sized IC - 5 pins, and is now obselete IIRC. it was made for 20W useage. some of the ICs are made more for 8ohm use. not sure which ones specifically.
Why would I select one over another?
-- power, size, the 3875TF is my suggested learning IC.
Could a gainclone be built to run off of a 12 volt DC supply?
-- no, the IC will not turn on with less then 24V, or +-12V.
Would this impart any limitations on the level of output power that could be produced by the gainclone?
-- yes, even at +-12V the power is much less then at +-30V.
Peter, you seem intent on being nothing but condescending and giving attitude, rather than providing any helpful info.
Let me word this clearly for you (you've shown you need that). I'm looking for helpful info. Thanks.
If you have an issue with new people coming on board trying to get an introduction to gainclones, then state that publicly.
If that's not your bent, then you've simply misinterpreted the intent of this thread since your first post, and since realizing that, all you've done since is discuss that your misunderstanding my intents must be my fault... outside of that, there's been precious little of value in your posts.
You misunderstood my intent, and absolutely that could be my wording.
Does it matter?
I did change my wording sevaral times trying to drive it to you.
Now, to placate you, let's address your concern.
Yes, I presented a list of questions that I've developed in the pursuit of these fundamentals.
I think I put clear wording to that end, in fact... in that post in fact (that you neatly edited out ).
The breadth of the questions themselves were meant to attract posts pointing to sources of broader information, as I didn't expect specific one-off individual responses to each question (although I thank those who did!).
My later wording throughout this thread was clearer, obviously.
The questions were (as stated in the part you edited out) questions that I hadn't found answers to in the application-specific threads that I had been poring through.
I think my intent has been clarified through the course of this thread (but even in that post) that I'm looking for any sources of fundamental information.
My apologies for any confusion along the way.
My intent isn't to gather specific answers to specific questions, that only helps here-and-now.
I've been looking to locate sources of answers, which helps both now and in the future.
I think I summed it up best in saying:
The spirit of this thread is to do both.
Let's keep it that way, OK?
It shouldn't require a post like this.
Let me word this clearly for you (you've shown you need that). I'm looking for helpful info. Thanks.
If you have an issue with new people coming on board trying to get an introduction to gainclones, then state that publicly.
If that's not your bent, then you've simply misinterpreted the intent of this thread since your first post, and since realizing that, all you've done since is discuss that your misunderstanding my intents must be my fault... outside of that, there's been precious little of value in your posts.
You misunderstood my intent, and absolutely that could be my wording.
Does it matter?
I did change my wording sevaral times trying to drive it to you.
Now, to placate you, let's address your concern.
Yes, I presented a list of questions that I've developed in the pursuit of these fundamentals.
I think I put clear wording to that end, in fact... in that post in fact (that you neatly edited out
).The breadth of the questions themselves were meant to attract posts pointing to sources of broader information, as I didn't expect specific one-off individual responses to each question (although I thank those who did!).
My later wording throughout this thread was clearer, obviously.
The questions were (as stated in the part you edited out) questions that I hadn't found answers to in the application-specific threads that I had been poring through.
I think my intent has been clarified through the course of this thread (but even in that post) that I'm looking for any sources of fundamental information.
My apologies for any confusion along the way.
My intent isn't to gather specific answers to specific questions, that only helps here-and-now.
I've been looking to locate sources of answers, which helps both now and in the future.
I think I summed it up best in saying:
Only two good reasons to participate on a forum... to learn, and to help others learn.geolemon said:All I've been looking for, is to be pointed to good sources of information, so that I can read up and research this myself.
Metaphorically speaking, I just want to be pointed to "the repository of chip-amp information", and fill out the paperwork to get a library card at the place.
The spirit of this thread is to do both.
Let's keep it that way, OK?
It shouldn't require a post like this.
theChris - thanks for the incredible help!
I definitely understand that there's a large part "art"...
...as there is with any aspect of audio.
But my problem is, I am not knowledgable enough yet to get to the "art" part.
It's like saying that mogul skiing is an art...
But before you can even get to the "art" part, you have to know how to ski, fundamentally...
...and even before you can ski, you have to know what equipment to rent or buy... as the wrong equipment would impede the learning that is standing in your way to getting to the "art" of mogul skiing.
I'm just looking to learn what common issues, applications, limitations, etc. are... and you've both given me great answers to that end, and great links.
I really do appreciate it... I'll definitely shoot you a PM one of these days, because I do have some other questions for you (as I mentioned before), regarding the custom processing componentry you were asking about on CAP.
Thanks!
theChris said:
I definitely understand that there's a large part "art"...
...as there is with any aspect of audio.
But my problem is, I am not knowledgable enough yet to get to the "art" part.
It's like saying that mogul skiing is an art...
But before you can even get to the "art" part, you have to know how to ski, fundamentally...
...and even before you can ski, you have to know what equipment to rent or buy... as the wrong equipment would impede the learning that is standing in your way to getting to the "art" of mogul skiing.
I'm just looking to learn what common issues, applications, limitations, etc. are... and you've both given me great answers to that end, and great links.
I really do appreciate it... I'll definitely shoot you a PM one of these days, because I do have some other questions for you (as I mentioned before), regarding the custom processing componentry you were asking about on CAP.
Thanks!
geolemon said:
It shouldn't require a post like this.
Actually it requires. I just can't help it, but I won't change my attitude towards your posts as they represent everything that I personally dislike (in others). I'm sorry, but that's just me.
So unless you start asking more sensible questions, you won't be getting sensible answers from me. But I don't see a problem here, as others are pretty helpful.
You might say that I give attitude, but it's actually your attitude that triggers that. If you check all my previous responses to other members, you'll notice that I was always trying to be more than helpful.
I think it makes sense if I'll permanently check out from this thread.
geolemon said:
It's like saying that mogul skiing is an art...
But before you can even get to the "art" part, you have to know how to ski, fundamentally...
...and even before you can ski, you have to know what equipment to rent or buy... as the wrong equipment would impede the learning that is standing in your way to getting to the "art" of mogul skiing.
I'm just looking to learn what common issues, applications, limitations, etc. are... and you've both given me great answers to that end, and great links.
But before I'll check out, let me comment on the above. You seem to have a problem with a grasp on reality. You see a big picture, but for some reson you don't notice what this picture is made off. You stand in front of the ocean, but are afraid to make a first step and enjoy the waves. You are not a practical person.
I never cared what equipment to use for mogul skiing. I just used whatever I had and went for it. I saved a lot of time this way.
Hello all,
My impression from his posts is that geolemon wants to understand how this thing works rather than just copy the work of another to get an amp that works. I must admit that the reason for building my amp was the quest for knowledge rather than trying to acheive audio nirvana. I just typed "how does an op amp work" into google and came up with these:
http://www.williamson-labs.com/480_opam.htm
http://www.m.case.btinternet.co.uk/html/basic_knowledge.html
Is this the kind of thing you were looking for geolemon?
Chris
My impression from his posts is that geolemon wants to understand how this thing works rather than just copy the work of another to get an amp that works. I must admit that the reason for building my amp was the quest for knowledge rather than trying to acheive audio nirvana. I just typed "how does an op amp work" into google and came up with these:
http://www.williamson-labs.com/480_opam.htm
http://www.m.case.btinternet.co.uk/html/basic_knowledge.html
Is this the kind of thing you were looking for geolemon?
Chris
Why am I not suprised to see another presumptiuous post by you.Peter Daniel said:
Such strong wording regarding my attitude? I've been nothing but grateful for all helpful posts.
Yours remain ludicrous, insulting and(inherently) quite childish.
If only you had a clue, a hint, or any familiarity with my past work (some of it is up on the web
), you'd discover to your embarassment how wrong you are. Experimenting is a big part of my life... but you can only learn from it, if you are scientific about it...
And if you don't learn from it, what's to be gained in the end?
Pretty little artifacts?
Saving time, yes.Peter Daniel said:
Shortcuts will certainly always save time.
But you won't get ideal results... if you even get satisfactory - or working results.
Worse, you won't have any idea what's been compromised.
I'd use the old the old, trusty "look before you leap", but in light of your statements, it seems that "don't be careless" might be more appropriate.
Carelessness is simply not my way.
To be honest, I can't even comprehend for the life of my how you could get worked up and insulted, at the fundamentals of my inquiring about how to learn about gainclones, from the ground floor... the basics.
I've never heard such a ludicrous thing as to criticize someone for trying to learn... much less state that the very concept is insulting to you.
Maybe once my first project is done, you can offer some valuable "hindsight is 20/20" insight for me.
Yes, particularly the first link... I think!chrish said:
It looks like there's a lot there.. I'm going to have to invest some time there, I've hardly skimmed it even.
Thanks for that!
This is not the thread for it... but one thing I'm trying to get a handle on are the limitations of these chip amps...
And I can already feel myself getting pulled down the road of "how different are these gainclone chip amps from traditional solid state amps?"
...and definitely "could the internal schematic of an op amp be built using traditional components, and modified to reach some goal not obtainable with chip amps... while yielding the same superior sound quality?"
Again, this isn't the thread for that... but the more that's learned, the more questions.
Unfortunate, in that way... the fundamentals tend to get the imagination going in dangerous ways...
Particularly when I suspect that many of those resulting ideas surely must have some yet-unlearned reason why they may not work...
At any rate... here's a very fundamental question...
What is the difference between the textbook "op amp", and the amplifier chips recommended for use in these forums?
Geolemon,
I think what Peter might be trying to say is that if you seek all of the theoretical answers before applying yourself in practice, you will:
A - Unnecessarily delay your beginning and
B - Learn less from the practical experience
I tend to agree with him. You wont build the best amp on your first attempt no matter how much study and research you put in upfront.
Knowing all the proper theory may keep you from trying some things that you will prefer sonically.
You may as well parallel your practical education with the theoretical. Like Peter said, it will save time. I know, I was researching my first valve amp for five years. I would have been much happier if I'd just built something - anything to start with.
I think what Peter might be trying to say is that if you seek all of the theoretical answers before applying yourself in practice, you will:
A - Unnecessarily delay your beginning and
B - Learn less from the practical experience
I tend to agree with him. You wont build the best amp on your first attempt no matter how much study and research you put in upfront.
Knowing all the proper theory may keep you from trying some things that you will prefer sonically.
You may as well parallel your practical education with the theoretical. Like Peter said, it will save time. I know, I was researching my first valve amp for five years. I would have been much happier if I'd just built something - anything to start with.
geolemon said:
The ideal op amp usually described in textbooks have three
fundamental (and unrealistic) assumptions
1) Infinite gain
2) Infinite input impedance
3) Zero output impedance
In reality the gain is finite and rolls off with frequency. If you check
some op amp datasheets and look at the open loop gain and phase
responce vs. frequency, you will see that they differ a lot and they
all differ very much from the infinite gain assumption.
The input impedance may be high, but not infinite. Besides, it will
be frequency dependent. Often it has a capacitance that varies
non-linearly with the voltage.
The output impedance may be low, but not zero, for low frequencies
but it will rise with frequency. You can often find a Zout vs. freq.
diagram in datasheets.
One could go on and say there are even more assumptions, but
most of them are anyway unimportant because of the inifinities.
PSRR is another issue usually not even considered for ideal
op amps, most certainly because the ideal power supply doesn't
have any noise anyway.
only to illustrate the above (theory vs. practice) i had a friend of mine who is an electronics engineer for philips review the schematics of an Aleph 5 by Nelson Pass.
He laughed his *** off and boldly stated "this just CAN't sound good" and started 'improving' the schematics.
i am a sailing instructor and have found that there are two major things you can do in teaching somebody how to sail.
A: learn all theory and then start sailing
B: just start sailing (with an instructor) and gradually, while sailing, learn the theory.
In my experience the last option works best. So:
geolemon, if you'd ask me (which you are not but i'll give my answer anyways), i'd just start building a simple, proven-concept, amp. Read up on your theory, think about how you can improve the amp, try it out and voila, you're there.
He laughed his *** off and boldly stated "this just CAN't sound good" and started 'improving' the schematics.
i am a sailing instructor and have found that there are two major things you can do in teaching somebody how to sail.
A: learn all theory and then start sailing
B: just start sailing (with an instructor) and gradually, while sailing, learn the theory.
In my experience the last option works best. So:
geolemon, if you'd ask me (which you are not but i'll give my answer anyways
), i'd just start building a simple, proven-concept, amp. Read up on your theory, think about how you can improve the amp, try it out and voila, you're there."And I can already feel myself getting pulled down the road of "how different are these gainclone chip amps from traditional solid state amps?"
...and definitely "could the internal schematic of an op amp be built using traditional components, and modified to reach some goal not obtainable with chip amps... while yielding the same superior sound quality?"
Again, this isn't the thread for that... but the more that's learned, the more questions."
-you could, but it would no longer be a gainCLONE. the gaincard had a few goal - short signal and shorter feedback paths physically, low PSU capacitance, low parts count, and vibration minimization. for the most part these are ALL subjective or counterintuitive. short feedback path shouldn't really matter much, but it is an ideal, so why not? low capacitance (still) seems like a bad idea, but worked, vibration shouldn't matter, but they adressed it too.
further, the matching of on-IC parts is very good. detection of thermals are likewise taken from inside the transistor case.
as for differences, an opamp will be smaller and run off lower voltage. the high current output also makes ossilation easier for the power IC.
...and definitely "could the internal schematic of an op amp be built using traditional components, and modified to reach some goal not obtainable with chip amps... while yielding the same superior sound quality?"
Again, this isn't the thread for that... but the more that's learned, the more questions."
-you could, but it would no longer be a gainCLONE. the gaincard had a few goal - short signal and shorter feedback paths physically, low PSU capacitance, low parts count, and vibration minimization. for the most part these are ALL subjective or counterintuitive. short feedback path shouldn't really matter much, but it is an ideal, so why not? low capacitance (still) seems like a bad idea, but worked, vibration shouldn't matter, but they adressed it too.
further, the matching of on-IC parts is very good. detection of thermals are likewise taken from inside the transistor case.
as for differences, an opamp will be smaller and run off lower voltage. the high current output also makes ossilation easier for the power IC.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.