Houston,we have problem...

[Upupa Epops]

Hi guys,I am here new one,but I read this forum over one year and I have fundamental question(like boys on Apollo 13):is here somebody,who love "notsimplicity" circuity?What I mean:circuity,especial power amplifiers,where are every details made in mode "state of art" and I say again every details.When I was younger,I was happy when I saw circuity of Tom Collangelo,Erno Borbelly,Robert Cordell,Malcolm Hawksford and many others-it was for me victory of brain against matter.Now I am studing datasheets of OA's,where only I see "similar victory".In my opinion " apostle of simplicity"Nelson Pass lead people to the wrong way and many people mean,that with one transistor (!) will be Heaven on the Earth.Fortunately is here people like Bruce Candy of Halcro,who have still brain on the head(look at his US patent-enjoy!).Many people here quote Douglas Self or prof. Leach,but it is circuity,which was modern before approximately fifteen years.Some people here makes pretty nice machines(Peranders,Jan Dunlop etc.) and I have question:have somebody here the same "love" like I have?I construct amplifier over twonyfive years and I know,that simplicity is not(in most of times) the righ way and with this statement I will stand alone.In the end of my "overture" guys,beg your pardon for my english-most of my life I spend on east ,but I hope that will be better.My name is Pavel Dudek,I am living in Prague in Czech Republic (for american guys:it is in middle of Europe) and my job is developement of electronic.P.S.:I am computer idiot.

 

[Koy]

Maybe give a try in different topic.

 

[john curl]

I am not clear what you want. Do you want simple? Do you want obsolete, you know, from the old days? This is it: Simple is good. Sophisticated and simple is better. Too complex, is interesting, but problematic. Choose for yourself.

 

[PMA]

I am not clear what you want. Do you want simple? Do you want obsolete, you know, from the old days? This is it: Simple is good. Sophisticated and simple is better. Too complex, is interesting, but problematic. Choose for yourself.

John,
most probably he wants sophisticated with the lowest distortion. Like Bruce Candy's Halcro dm58/dm68. You probably now the Stereophile review : http://www.stereophile.com/showarchives.cgi?683


What he tries to say is that similar results are not achievable with simple topology. These results are achievable only in case that you reduce distortion in every amp's stage, which is perfectly impossible in simple topology. On the other hand - difficult way for DIYers, I agree.

Pavel

 

[jcarr]

Upupa Epops:

>In my opinion "apostle of simplicity" Nelson Pass lead people to the wrong way.<

You may not agree with Nelson's way, but the ease of building simple designs like the Zen family and the Gainclone variants has undoubtedly been a major factor in enticing many people into trying their hand at building amplifiers, whereas more complex designs would in all likelihood be too intimidating for most beginners. All of us were beginners at one time, and no matter how much progress we may make, we should never forget what it feels like to be a beginner. You cannot be a good parent if you have forgotten what it is like to be a child, isn't this so?

My opinion is that the most important things are whatever goals that _you_ decide that you want to achieve, with success measured in terms of how closely the end-results align with your defined goals. Complexity, simplicity, tubes, discrete semiconductors, ICs, switching-mode operation; these are all tools for us to use in trying to achieve our goals, and should not be construed as being holy grails unto themselves. My advice is to use whatever approaches, tools or methods you think offer the best chance for you to attain your goals. And if the end-results fall short of your initial goals, analyze what went wrong and try again. Making mistakes is an essential part of learning.

In general, I enjoy complex, sophisticated electronics designs far more than simple ones, and I would certainly like to see other designs that demonstrate a lot of thought and innovativeness.

But that is a reflection of my personality and my preferences. I have no illusions that this is the best approach for anyone except myself.

regards, jonathan carr

 

[Steve Eddy]

In general, I enjoy complex, sophisticated electronics designs far more than simple ones, and I would certainly like to see other designs that demonstrate a lot of thought and innovativeness.

But that is a reflection of my personality and my preferences. I have no illusions that this is the best approach for anyone except myself.

Bravo, Jonathan. Such words are a breath of fresh air compared to the stifling absolutism and intolerance of individuality and diversity that is too often preached by others.

se

 

[john curl]

Upupa, I understand you now. Well, it is difficult to give difficult circuits to amateurs.
Halcro is my greatest and most direct competitor. I liked the sound that they produced at the CES with the Dave Wilson speakers.
They have AMAZING specs. Better than anything that I will ever do. However, they would be almost impossible to reproduce easily. This is real engineering! It requires proper layout and individual compensation at each stage. This is almost impossible without a real company behind you.
Go ahead if you must, but you might as well try to make a Porsche on your own, as to make a Halcro. It is cheaper and easier just to buy one. Become a dealer in your area, get it at wholesale or better.

 

[PMA]

Upupa, I understand you now. Well, it is difficult to give difficult circuits to amateurs.
Halcro is my greatest and most direct competitor. I liked the sound that they produced at the CES with the Dave Wilson speakers.
They have AMAZING specs. Better than anything that I will ever do. However, they would be almost impossible to reproduce easily. This is real engineering! It requires proper layout and individual compensation at each stage. This is almost impossible without a real company behind you.
Go ahead if you must, but you might as well try to make a Porsche on your own, as to make a Halcro. It is cheaper and easier just to buy one. Become a dealer in your area, get it at wholesale or better.

John,

nice post, thanks for it. Upupa is a remarkable designer and he reached similar parameters as Halcro (a little bit worse, but not much) some 10 years ago. But in the past our "two worlds" were separated and bi-directional exchange of information has started few years ago. Halcro is a great challenge for him. We are doing comparative listening tests of well known famous world amplifiers in high quality audio chains (with speakers like Wilson Audio Maxx), but never had a chance to hear Halcro ...

Pavel

 

[john curl]

Actually, each amplifier design has a number of tradeoffs. It is fairly easy to make an amplifier that has almost no distortion. This is done by negative feedback, both local and loop. It can also be done with feed-forward, or the Quad 'current dumping' type circuit.
Each designer must decide what sounds best: ultra-low distortion, single ended, open loop operation, class A, etc. No one approach is necessarily better in every way, from the others. This is about all that I have to say on the subject.

 

[Diode]

Hey John,

WOW! I wish I could say stuff like that in a sticky situation!

Pavel,

John is right, It depends on what you want to do and how you will use it. I like things to be simple and work well but not too many people can hear the things that a computer and test gear can, so where is the balance of the two? You can build the most linear, most efficient, highest parts count, expensive, complicated, scientific marvel amp, but if my human ears can't tell the difference, what's it all for anyway. OK, you have the best specs ever achieved in amplifier technology, but for what? Bragging rights? Good for you but that won't change the fact that 98% plus, of the population physically won't be able to hear the difference and your $30,000.00 amp will be yours and yours alone, while the rest of us are happy with a $500.00 amp we built ourselves or bought from the electronics store, with even lower specs than most of us DIY people have.

I'm not trying to be a jerk. I'm just trying to be realistic and practical here. I do agree with you. It sometimes takes some pretty fancy circuitry to get the best of results.

Good luck,
Chris

 

[PMA]

Chris,

but you can hear the difference . It just depends on quality of the whole audio chain, especially loudspeakers. Amps and speakers distortion are very different in its nature and they superimpose, not mask. And the amp made for me is much less expensive than you declared . It is only a question of the goal you have - the DIYer's luck from working amp or one of the best sounding amp even in a competition with the tops ?

Pavel

 

[MBK]

Pavel,

as a sidenote I like what you published on your www page - real measurements! Many apostles of low distortion here discuss mostly simulations, something where I don't follow so easily because reality usually bites differently.

I have an OPA2134/BUF134 combo as headphone amp that is painfully transparent, your measurements show me why . I just wonder, did you use a pot when you measured distortion? I suspect often contacts and pots considerably limit the distortion performance of the circuit before and after. I also wonder if one could just parallel the OPA2134/BUF134 's to get them up to power amp level (say 12 per leg, balanced to get the voltage needed and additional distortion cancellation). For a not-too-expensive Halcro competition accessible to advanced beginners like me.

All my DIY gear so far was based on op amp implementations. Still I wanted to try single ended class A so i now I am building a Mini-Aleph to hear for myself. The sound of the prototype appeals to me but I have to wait for the "proper" implementation to judge. Also, my chip amps are not so good on distortion: rising distortion with falling output power, not so good. This thought tempted me to go Class A, to see what an amp with the inverse relationship can do for detail.

I agree with you that the hardest thing to reproduce is symphony or complex music. I am afraid my speakers (Diatone based) resolve too little though to show amp differences here. They are well esteemed but I have my doubts - I hear soooo much more on headphones with OPA2134/BUF134.

 

[PMA]

MBK,

regarding THD spectral measurements, they are made for the circuit according to the image shown (with pot, connectors and power supply). The preamp circuit is mounted inside Rittal VarioCase instrument case with good screening effect. Separate measurement for Audio Buffer (OPA627+BUF634T) is also shown. The measurements were updated on August 14. The preamp based on mentioned circuits has a very very clear pure sound with very good space and localization. The cable termination is a must to achieve this. This was tested on Wilson Audio Maxx Speakers, for example. You are right - headphones can disclose very much.

Pavel

 

[MBK]

Pavel,

since you mentioned it, I was wondering about that termination. I get a bit confused about the whole termination issue - on one hand impedance matching isn't recommended anymore in audio standards, on the other hand it seems to me that a) higher signal current means less influence of picked up RFI and b) reflected RFI in unmatched lines can intermodulate the signal... (and seep into driver outputs too).

Then again if you do terminate as indicated , at say 20 V p-p you use up half the current capability of the BUF134. So, what is uour rationale ... is my analysis correct?

 

[PMA]

Pavel,

since you mentioned it, I was wondering about that termination. I get a bit confused about the whole termination issue - on one hand impedance matching isn't recommended anymore in audio standards, on the other hand it seems to me that a) higher signal current means less influence of picked up RFI and b) reflected RFI in unmatched lines can intermodulate the signal... (and seep into driver outputs too).

Then again if you do terminate as indicated , at say 20 V p-p you use up half the current capability of the BUF134. So, what is uour rationale ... is my analysis correct?

First - I do not know BUF134. I use BUF634 - maybe only writing mistake.

Termination - I know that this is a tricky issue. Probably it does not have much to do with wave impedance of the cable. But results of many many listening tests with many listeners (not being informed what it is about) are unambiguous. I think that you pointed at the important item - resistance to EMI/RFI and probably intermodulation. See measurements of generator with Audio Buffer (50 Ohms terminated cable) - the result is better than generator alone (unmatched cable with no termination). I think that there is another issue - complex permitivity of the cable and its frequency characteristics. The dielectric phenomena and their time constants. This seems to be minimized by termination - also from the reason that high output current is used.
I have more information and especially images in the czech version of the web pages. I suppose that images would tell you enough, so click on it.

There seems to be no problem with buffer output current - you can see buffer THD measurements done with termination. The voltage halves, but buffer amplifies 2x before termination.

Pavel

 

[MBK]

Pavel,

interesting. Yes, I meant BUF634, sorry for careless typing, I used a lot of OPA 2134 and INA134 and DRV134 and the "134" must have stuck in my head!

The termination, as I said I don't fully understand the issues. I hear that scopes and RF equipment etc., all get terminated at line impedance, usually 50 to 300 Ohms. For audio the recommendations now say, use low Z out and high Z in. But in my little head I figure this: Any error by RFI on the cable will induce a *current*. On the amp out side we have low Z - so in th ideal case the amp sees the RFI error and corrects it. In the un-ideal case we get RFI slew rate limitation errors on the amp side etc. On the receiver side, the induced current will see a high Z. Ohm's law U=R*I means then, to me, that this tiny current gets transformed into a large voltage if using a high impedance input, or, depending on frequenzy, it gets reflected right into the amp output (which has the perfect output impedance to match the cable!!!). Or am I completely wrong. But, with medium Z out and medium Z in, matched to line, terminated to line at both ends, we would get 1) a better signal - to - RFI ratio (because higher signal current) and 2) a -6dB drop in noise at the receiver input rather than reflections or ideal voltage gain of the RFI error current.

Then again I don't have any EE qualification.

 

[PMA]

MBK,

the only think I can say that we are group of people (EE) trying to find the reason WHY it works as it works. An we do not have it . Much theory, better or worse presumptions ...

But it works and it is repeatable, which I consider to be the most important issue now.

Pavel

 

[MBK]

Lucky I am not the only one! In fact I always try to understand from first principles - often the details elude me though. So you basically say that your amp inputs are set to a 50R series+50 R shunt input impedance? Or do you use the cable termination at 50R shunt, and then again say 1k series and 47k shunt to ground? And, have you tried with 300 Ohm cable (ex.: TV coax) - that would lower the current requirements...

 

[PMA]

MBK,

the chain is like this:
CD player with Audio Buffer mounted inside, 51R series resistance at the buffer output (R10), coaxial cable, 51R shunt resistor in RCA connector, preamp with 100k input impedance, buffered output of preamp with 51R series resistance, coaxial cable, 51R shunt resistor in RCA connector, power amp with input impedance of 100k. As a signal cable RG-58 and RG-62 were tested (and many well-known expensive audio cables). RG-62 seems to be the best choice, though the wave impedance of it is 93 Ohm. Audio Buffer as a replacement of very expensive signal cable was always better than any of them. It seems to be very important to isolate the OpAmp from cable interactions by means of buffer. OpAmp makes precision (and is practically unloaded), buffer delivers power and isolates from cable capacitance.

As the cable is terminated by 51R shunt resistor, it does not make any change whether the amp has 10k, 47k or 100k input impedance, also input capacitance is out of any interest.

 

[Diode]

Pavel,

Yes, I agree with you to an extent and there are many examples of these amps. I think you know what I mean about the point at which most people can't tell the difference.... THAT, was my point. I absolutely respect engineers for their knowledge of circuits and what-not. I have worked with many engineers as a technician. The thing I can't stand about them is their personalities. They behave like 5 year old children when they don't get their way or someone else disagrees with them. I once told a man I worked with.... "I respect your knowledge...You are a brilliant engineer.... BUT you absolutely SUCK as a person" The 5 year old then surfaced again and I then threatened to take him to day care." Needless to say, he left the lab and didn't speak to me for days. Such is the life as a technician. I'm not disagreeing with you about quality sound. I love quality sound but as long as the cheaper amp and speakers satisfy my basic needs, the more expensive stuff will remain on the sellers shelf. BUT, now that I understand what you are doing and where you are going with your knowledge..... I absolutely agree with that!!!!!

By the way.... Do you speak only Czeck or do you speak Russian too? I'm just asking.... My wife and daughter are Russian......

Chris