Zero Knowledge but eager to learn

LOL - Chris/anatech! Well, you can always print it out (assuming your printer won't run out of ink or burst into flames. 🤣

Don't think most of us on this forum are young'uns anymore. (Probably) Most require reading glasses (due to the aging process or staring at LCD/LED monitors to long). (Still think when we had CRTs, our eyes had less fatigue/blurriness/headaches.) 🤓

Yes, physical books were a comfort in our hands. Not this thumb/mouse scrolling and can't/hard to live without gadgets so many are enamored with and are constantly absorbed in. You see it wherever you go - in their hands/mounted on their vehicle dashboards, having a bite to eat, even amongst family time. Of course with the resultant head bowed (creating a commonwealth of arch necks/backs in the foreseeable future). Times were simpler when we were young'uns. 🤔

Thanks.

Sincerely,
Kingsley.
 
  • Like
Reactions: Z.K
Hi Kingsley,
An easier way to solve the problem. Buy the book like we used to do (remember the wet copiers? - not an option). Papercuts are a very real hazard. I wonder when legal will have warning labels on books?

I have both editions and several other books. I like to read during power failures.
 
Good morning/afternoon/evening, Z.K:

Pretty much in the same boat as you, Mulburg, Read-cap, and probably others. Can follow schematics a bit but understanding how it all works/what does what/etc. is a headbanger at times (aside from something maybe simple).

Found this online...The Art of Electronics - 2nd ed pdf. So, had at it. (A very) Long read for sure.

Thanks. (and thanks to the University of Wisconsin-Madison for providing the PDF.) 😎

Sincerely,
Kingsley.
THANKS! Well appreciated BUT the most important chapter for folks like me, they left out.

Chapter 1 - FOUNDATIONS is missing

Unless I need to get reading glasses and I missed it......

I think you are way ahead of me. I'd have trouble following most if not all schematics.

With a single exception, where I identified which board was responsible for the picture getting squeezed vertically, every old fashioned fishbowl tv I opened up... well lets say it did not end well.

I did manage to fix a router/modem that had trouble connection to the ISP by spotting a leaking capacitor on the board but that would be the end of my "success" stories.

On the other hand as far as I remember 99%+ of all empty circuitry boards I filled with components ended up working after following instructions
 
Last edited:
...RPI, Linear power supply, DAC HAT, I/V stage...
That approach can be difficult and expensive to get working well. RPI is an EMI/RFI noise source, and its GPIO bus is very jittery. If you really want to use RPi, then its probably better to use it as a USB host for a USB dac. That way you can put some distance between the radiated-and-conducted noise source and the dac itself.
 
  • Like
Reactions: anatech
Hi ZK,
The link is addressing concerns you don't have to an extent they will affect anything. Some folks worry about things that either are not a problem, or you may have more significant problems.

Balanced is not an intelligent signal format to follow in any small system. The negatives outweigh the imagined benefits. I worked in the test and measurements field, the telecom field in addition to audio. In audio I worked in several recording studios, and one major (still existing) one. The reasons to go balanced are:
  • Long runs (leading to high cable capacitance)
  • Electrically noisy environments, cable shields often have some % of effectiveness. Balanced cables allow you to use the CMRR of the input circuit to cancel most
  • Lower impedance also forms a voltage divider with radiated noise to reduce the induced noise level
  • Original balanced lines used signal transformers which theoretically have perfect CMRR performance and no added noise. Transformers add distortion and they are not perfect.
  • Electronic balanced converters have lower CMRR, better frequency and phase performance and a touch of noise. Better than transformers sometimes.
Every circuit has some amount of noise and distortion. Balanced circuits each have the same noise, but since it isn't correlated noise it isn't double, just 1.414 x. Going to and from balanced means you are adding a circuit or transformer on the ends of your circuit, degrading performance. That and just try to keep the gain and phase perfectly matched. If you have bigger problems to deal with, balanced signal lines are the lesser evil.

Understand?
Well I am trying to. I get that you are saying it has disadvantages that are worse than imagined benefits.

I prefer running xlr cables to other rooms rather than speaker cables. The amplifier modules are expecting balanced input according to the manual. And last but certainly not least: I have a whole pile of xlr chassis connectors, plugs and cables that I wanted to finally put to good use.

Main issue is still the getting it to work from within the software that runs on the RPI. That thread starter seems to set the volume with the help of Arduino maybe he knows how to do it with a RPI

IF that works we have the upside of empty PCBs being available and having the software issue solved. The downside is 1.414x more noise.

I understand that a balanced circuit would be needlessly complicated but it is FAR less complicated than for me having to design a circuit, figure out how PCB software works, make prototype PCBs, figure out why they dont work or why they sound horrible.

Filling an empty PCB, soldering it and making it work I would guess take me 2 days whereas starting from scratch with my lack of knowledge the first option would take 2 (?) years or more?

Regarding the extra noise, yes, not good but what makes you think that I could get to a design myself that would be less than double the noise and distortion than something designed by someone who REALLY knows what he is doing?

I am however clueless how to spot if a design is excellent from a sonic point of view rather than terrible
 
I get the drift, Chris/anatech. 🤔

And Z.K. apologies that Chapter One (the fundamentals) is missing (which probably explains why it's free for download). 😳 (Was only trying to help/assist.)

Thanks.

Sincerely,
Kingsley.
It was well appreciated 🙂 but being as clueless as me, it would have been better if ALL the other chapters were missing

HOWEVER a quick qoogle search will let you find the 3rd Edition as well and there the first chapter IS available. As long as the first chapter in both editions is similar, I can always start in the third edition with chapter 1 and then move to the 2nd Edition for chapter 2 and the rest
 
You young guys are lucky.

I don't retain stuff off a monitor well. I like a book. Plus it doesn't take batteries and I can relax without emails or ads. lol!
Point taken. For me it is even worse I don't retain what i read until AFTER I use it a few times. More than a few days between reading/instructions and application, I can start over. With 1000+ pages I see a problem coming because I suspect this will be FAR from a quick read

Firefox + uBlock Origin (should be free) and your internet experience will quiet down A LOT. No ads from Google and ad free youtube as well. Silence is BLISS
 
Last edited:
That approach can be difficult and expensive to get working well. RPI is an EMI/RFI noise source, and its GPIO bus is very jittery. If you really want to use RPi, then its probably better to use it as a USB host for a USB dac. That way you can put some distance between the radiated-and-conducted noise source and the dac itself.
There will be an isolator / shielding HAT in there and the PI is not powered from the same PS as the other components.

I am not going to stuff this in a small box and I can put extra shielding around critical components, if need be, no problem
 
Reclockers.

I have a $30,000 instrument I measure clock performance with (HP 5372A). It is locked to the GPS giving me a maximum error of approx 5 exp -12, so way more accurate than the clocks used in creating the music. However, when I need to characterize a clock I have to disconnect the GPS reference clock. That's because the internal double oven crystal oscillator is more stable in the short term.

So what is the most important in music decoding? Short term stability. Long term doesn't matter at all. The most accurate and stable reference is a crystal oscillator, period. The normal crystals can be down to 10 -6 error, and SC cut more accurate. Ovenized oscillators get down to 10 -9 for example. So what do you need? A normal crystal oscillator, that's all and it is far better than you'd think. Now consider your encoded musics will have embedded all jitter from the source, that's in there as sample interval errors. Reclocking will not change anything. So all this noise about the best clock totally fails to consider the source. On top of that, even normal crystal variations cannot be detected by a human being. I don't care who you are, the human brain isn't capable.

So when considering how to make your system better, you have to consider the entire system and source and that includes what your own body limits are.

Why not stay locked to the GPS signal? Because the disciplining methods makes occasional corrections to the crystal frequency, causing tiny jumps in frequency. You would never hear this, but for testing it shows up like a sore thumb.
Do you mean source as in the AD conversion when the recording is made?

If clocks and reclocking makes no difference and don't reduce jitter, then why do people report a sonic difference when they swap clocks?

In fact if stuff like that would make no difference or could not make a difference then why do people hear differences between $150 chinese DACs and $1500 and $5000 DACs?

There seem to be quite a bit of activity in this forum of people playing with this sort of stuff.

Is this like with cables where some people claim that stupid expensive cables are sonically far superior?

Regarding the cables I am rather skeptical. I can see why a crap cable could be a bottle neck or act like an antenna and as such introduces noise. On the other hand I have yet to hear much, if any difference between say $50 cables and $500 cables etc
 
Hi ZK,
The major point is this ... then entire reason.

Expectation bias. If you spend a lot of money or go to a lot of trouble, it must be better - right? Not always as it turns out. This is how the audio industry has run since the mid 1980's. A good story, something expensive and a discovery. Well, discoveries are not made in basements and garages anymore. They are made in labs with engineers and very good test equipment.

Everything we do comes from industry. Parts, circuits, everything. Audio represents a tiny fraction of the electronics industry income no matter how great individuals think they are. The circuit hasn't a clue whether it is a servo control signal, an audio signal or any other signal. It hasn't a clue as to what came before or what is likely to come. It simply does what it was designed to do as long as you keep signals within the parameters (frequency, voltage) it was designed for. Test instrumentation is far more accurate than audio circuits are. I update audio circuits using basics and what I learned from test and measurement. The roots in audio? The telephone. Yup, balanced circuits and everything else. Even sampling audio signals. Everything we do came from the telecommunications industry, and they are far ahead of us. Lives are not threatened by some audio distortion in a sound system, but an industrial control process may well kill. Weapons systems absolutely must be accurate.

These are the cold hard truths.

Now we have many of us, me included, who make a living in the audio industry. Some can't produce really good stuff, so they spin a story (snake oil), and people like stories. They like to think they have something exclusive, and they love to think they can make something great with their own hands. Don't we all? Give them the promise they can improve things and wishful thinking takes over.

So with clocks for example, these are factual, quantifiable things. Humans have limits they cannot perceive beyond, but we always have people claiming they are special and super human. Our brains can determine phase differences only to a certain degree depending on frequency, and the same goes for timing and jitter. "Yeah but ... I can hear it even if you can't measure it". Nope, this has not been true for a couple decades anyway. In the 1970's when I started - yes. It carried over.

So what makes a difference? How you treat the audio after D/A conversion. You can have several DACs with the same chip set, some sound different due to PCB layout, power supply and other factors. Errors can be made with signals (clock or clock recovery) that get into the mix, but they only report such and such a clock. You have to eliminate all variables except the one under test to make any statement you can rely on. None of these tests are truly controlled, nor are they backed up by responsible, repeatable testing.

I came in the bottom floor of CDs and D/A conversion when it was introduced. We were taught far more than we had to know, so I know more than I want to. Many others did as well. I work in the field and have the equipment to back things up, as do others. Be very suspicious of any claim when there is commercial gain or reputation at stake. Check mine out - please! Not on the informal internat, but where people do know without being invested. I'm just trying to show you the way, I don't sell anything.
 
  • Like
Reactions: Z.K
Now we have many of us, me included, who make a living in the audio industry. Some can't produce really good stuff, so they spin a story (snake oil), and people like stories. They like to think they have something exclusive, and they love to think they can make something great with their own hands. Don't we all? Give them the promise they can improve things and wishful thinking takes over.
Chris, you are about the most self-confident sounding person I can think of in this forum.

Unfortunately, that's also the case when you might be a little off-base. The story you tell in the above quote reads like a mix of fact and fiction, and IME part of it is fiction. That's because it paints many very different people and many true advances with a overly broad brush of condemnation.
 
Hi Mark,
I try to be as factual as I can, that's all. If I have measured and proved something, I'll say so. I also know what industry norms are. What works and what doesn't.

I'm also always open to correction and will admit to being wrong. But, you have to be factual and agree with how physics works.

So if I am off base, let me know. You are well aware I don't get mad or upset - unless I see someone ripping others off.
 
Sure it is. But most folks prefer better, meaning more accurate, less noise.

I constantly see people waste anywhere from $100 to a great deal more. These days, not many can afford to waste that money. Even if they can, is it right to sell product on the basis of something that is not true? I don't think so.

Honestly, if you like the sound of an early 1920's technology thing - go for it. However, you have made it clear you are looking for something as good as it can reasonably be. Want to get a better clock? Okay, then make sure it actually solves a problem and is better than what you have. Otherwise you have wasted money and maybe even end up worse off while thinking performance is better.

The best clocks for digital are crystal (quartz) and will perform at least as good as the recording equipment. A bad power supply or digital circuit can degrade that performance. Execution is at least as important as the basic design. Does it matter if it slowly drifts a few Hz over an hour? Nope, you will never hear that. Will short term shifts in frequency be audible? Maybe, depending on how bad. So you worry about short term shifts in frequency. It's really simple.

No matter how good a clock you do have, it will never correct for sampling jitter when the music was encoded. Anyone who tries to tell you differently doesn't understand how this stuff works. Please fact check, come to your own conclusions.