We will have to disagree. I am well versed in digital logic. I made a very good living at it. Rise time is not a red herring, it is one of the many real factors we must contend with in the digital world along with propagation delay, transmission lines and dynamic current supply issues. ( incorrect bypass caps, quite common). One more time: Digital is a mathematical concept. It is executed in the real world as analog bi-stable circuits. Circuits that want to be fully on and fully off. It takes time for any transistor to switch. During that time it is in the linear region. It is generating more heat than when in the on or off. Is the drive current is slow, or of insufficient drive for the capacitance of the gate, then you will increase the power dissipation of the gate. It can be done to the point of failure. As DAC's are single die, having one end of the chip heat up could quite possible heat up the other end of the chip where the analog output stage is. How mush it does so in our real DAC chips would have to be tested. I am not saying it does, I am saying it is reasonable physics.
It is far more likely the poor design of earlier or low end chips did not buffer and clock the inputs well so waveform distortion of the input stream would cause jitter that the DAC design did not compensate for. Still, a quite valid cause and effect for the symptom they are talking about. How much? Again, go measure.
Physics don't change for anyone. Audiophiles just don't read the book. Actually I take a bit of offence. The term Audiophile has been degraded to the same level of mythology as taro cards. In the old days, we computer hackers were proud. We tore into our computers, built them, and figured out how they worked. We made them better. We were not computer CRACKERS. Now hacker means the scum who ruin the world for the rest of us. I know many audiophiles who are well versed in physics. Many are graduate level engineers. ( a few hang out on this very thread.) I was proud to design and build on a wire-wrap card my first 6502 computer.
I do make an acceptable living at it.......Again this insistence on analogue, we know it is at a lower level, but we are designing digital so we follow digital guidelines.
http://www.diyaudio.com/forums/digi...-wave-spdif-cable-possible-6.html#post3817956
I was on about rise time and temperature increase, it is not really going to affect a DAC or I should imagine any logic unless you start using lonnnnngggg rise times of say a second, but the rise time is chosen to be commensurate with the logic and clock speeds used, and having too fast a rise time is detrimental.
Look at some of the fantastic claims and products and wonder why the term Audiophile has such a derogatory ring to it these days.
Now we are getting somewhere. OK, for good practice in the digital domain, we take our input and sent it through a wave shaping as we pass it into a buffer out of which we clock the output. So, if the system was designed correctly, we should not have any issues with the distortion caused by the cable. Look up the external signal processors as used in HDMI to see how to address this problem. ( and notice how many bad designs do not use these recommended parts, then claim the HDMI spec is bad)
Too many designers just look at the digital problem and forget the real world IS ANALOG. I spent a lot of time in the failure analysis lab teaching our wiz-kid digital designers that lesson. Those who listened I did not see again. Those who thought digital was ones and zeros would be back in the lab as their boards were failing. Please try and stay out of the latter group.
I have SEEN IN THE LAB serial inputs with degraded rise time actually cause thermal problems in the receiver. It was a high speed digital PPL data recovery system for dasd disk back in the 80's. It was a bad design, but it happened.
I quote from WIKI definition:
"An audiophile is a person enthusiastic about high-fidelity sound reproduction."
If you say all audiophiles are lunatic fringe magic brick or foil triangle believers, you have a very narrow view of the world.
Too many designers just look at the digital problem and forget the real world IS ANALOG. I spent a lot of time in the failure analysis lab teaching our wiz-kid digital designers that lesson. Those who listened I did not see again. Those who thought digital was ones and zeros would be back in the lab as their boards were failing. Please try and stay out of the latter group.
I have SEEN IN THE LAB serial inputs with degraded rise time actually cause thermal problems in the receiver. It was a high speed digital PPL data recovery system for dasd disk back in the 80's. It was a bad design, but it happened.
I quote from WIKI definition:
"An audiophile is a person enthusiastic about high-fidelity sound reproduction."
If you say all audiophiles are lunatic fringe magic brick or foil triangle believers, you have a very narrow view of the world.
So if your source is 50 ohms, the test set is 50 ohms, and the cable is 75 ohms, is that a reasonable test? What does the test tell you then?
jan
No I don't say all audiophiles are loonies, and I and others who are into audio, do use the term more to refer to the more Avant garde beliefs🙂
And I am aware that deep down (below the surface) digital is analogue based, the sim tools work in the analogue domain doing multi passes to cover the full spectrum of harmonics, but here saying digital is analogue leads to mistaken views on how digital data is transferred.
🙂
And I am aware that deep down (below the surface) digital is analogue based, the sim tools work in the analogue domain doing multi passes to cover the full spectrum of harmonics, but here saying digital is analogue leads to mistaken views on how digital data is transferred.
🙂
It means you are testing what you will actually be using which makes a valid test. Although who knows what most cables are. Anything to about 90 I guess. If you really want to evaluate the waveform, you have to have the exact cable, source, and sink. I would not bet the real source or sink was 50 Ohms, and I would not bet vendor to vendor either. I also wonder who picked 50 Ohms, but that was probably set many years ago for lab equipment.
Of course, a well designed circuit should not care. "Digital" is supposed to make it not care. A good transient clipping circuit, Schmitt trigger with sufficient bypass caps being clocked out of a buffer should make it all work just fine.
Data is transferred digitally only in software simulation. In the physical world, analog is used to transfer digital information as physics is analog. We both understand this. I think it is far more misleading to perpetuate the digital simplicity myth to those who do not understand what is under the hood than to inform folks it is more complicated than that.
Digital information may be transmitted by many analog methods. We think of the simple two level representation, but we have many other ways; PE, NRZ, and modulated carriers.
50 ohms and 75ohms were chosen as standard coaxial cable impedances because one gives the maximum power handling capability for a given size and the other gives the minimum attenuation. I can't remember which is which, but I'm sure Google can help you.tvrgeek said:
Power handling and impedance are not directly linked. To cable spec yes, as you can have tiny micro power 50 Ohm and big honking coax for a TV transmitter. Why these values were picked is even before my time. It probably had more to do with transformers, or what was built happened to measure. No ITU in those days.
Can't find the spec, but it seems to me consumer PCM is 75 Ohms. 50 to 100 is probably fine as we have a lot of power. Likewise, without a spec on rise time, the BW is unknown, but I doubt it is more than a meg or so. You could probably use bell wire. The AES/BAU spec (diff) is 110 Ohms.
Can't find the spec, but it seems to me consumer PCM is 75 Ohms. 50 to 100 is probably fine as we have a lot of power. Likewise, without a spec on rise time, the BW is unknown, but I doubt it is more than a meg or so. You could probably use bell wire. The AES/BAU spec (diff) is 110 Ohms.
Yes they are. Coaxial cable power handling is limited by either copper heating or flashover voltage. The impedance affects both. There is an optimum impedance, which may assume a copper conductor (I can't remember). As I said, Google will tell you.tvrgeek said:
It was before my time too, but that doesn't mean I haven't read about it.
Nothing whatsoever to do with transformers, or what they happened to build. It was carefully calculated, as all engineering was in those days.
For local interconnect power is not really an issue, but it should be 75R. This was probablt chosen because it was already a video standard so technology (e.g. cables, connectors, driver chips) was available. Cable bandwidth will be hundreds of MHz, and at least many tens of MHz are needed even for 44.1kHz SPDIF. Bell wire might do for a very short run - it has a characteristic impedance somewhere in the 100 ohm region - but I would not recommend it unless you like extra 'detail' with your music.
Audiophiles are people that want great fidelity. "Cable" people have issues that go far beyond wanting fidelity. Loopback, Square wave and null tests are often banned from places that sell gear to people with issues. Simple truth. Those tests tell all, spoil the game and make people face facts about themselves and their choices they would rather not face. The truth does not sell gear, or worse it sells older gear. We reached speaker driver zenith un the 40s amp Zenith in the late 60s and dsp zenith in around mid 2000s.
Some people with very expensive equipment feel that the cables should be proportionately expensive. I'm not sure that theory holds water technically; but typically the same people have an embarrassing amount of disposable income, and if it makes them feel better I'm OK with that. The ones that puzzle me are those with very moderately priced (and performing) equipment who are willing to pay vast amounts for cables and power cords.
DF, a quick browse in the Belden catalog will clarify a few things. There is a bit of difference in how much power you can send down RG-174 vs RG-8.
http://www.belden.com/products/catalogs/mastercatalog/brilliance/upload/06Coaxial_Cables.pdf
As I know a lot of old engineers; don't fall to the temptation of looking backwards through rose colored glasses. A lot of "good enough" engineering has been SOP for thousands of years. Many standards are only because that is what we always did.
At audio frequencies, for analog cables the impedance is pretty much irrelevant. 75 works fine ( I use stranded Belden 75 Ohm as it is well shielded and flexable.) but the cheap twin cables which are about 100 Ohms are fine too. This thread was about wave shape in a PCM cable, where as I mentioned, there could be some issue with impedance as a square wave by definition contains high frequencies, but I doubt it. Conclusion remains, the OP concept is not against the laws of physics, but for a well designed system should not be relevant.
fpitas,
What is amazing is how a salesman can sell a customer a $200 pair of speakers and then try to sell them $100 in exotic cables. To me, this is way into the unethical range.
A lot of inexperienced mid-fi users read the audio magazines where magical attributes are perpetuated. I think the classic term is "sucker".
Pete,
I think you need to hear some better modern speakers. I have been building them for about 40 years and I am amazed how much better drivers have gotten in the last decade. We still have a long way to go. When were the best amps made? I think sometime in the late 70's early 80's amps got pretty good, but some new ones are still making progress. DSP has a long, long way to go. It has opportunities we have not even dreamed of yet.
http://www.belden.com/products/catalogs/mastercatalog/brilliance/upload/06Coaxial_Cables.pdf
As I know a lot of old engineers; don't fall to the temptation of looking backwards through rose colored glasses. A lot of "good enough" engineering has been SOP for thousands of years. Many standards are only because that is what we always did.
At audio frequencies, for analog cables the impedance is pretty much irrelevant. 75 works fine ( I use stranded Belden 75 Ohm as it is well shielded and flexable.) but the cheap twin cables which are about 100 Ohms are fine too. This thread was about wave shape in a PCM cable, where as I mentioned, there could be some issue with impedance as a square wave by definition contains high frequencies, but I doubt it. Conclusion remains, the OP concept is not against the laws of physics, but for a well designed system should not be relevant.
fpitas,
What is amazing is how a salesman can sell a customer a $200 pair of speakers and then try to sell them $100 in exotic cables. To me, this is way into the unethical range.
A lot of inexperienced mid-fi users read the audio magazines where magical attributes are perpetuated. I think the classic term is "sucker".
Pete,
I think you need to hear some better modern speakers. I have been building them for about 40 years and I am amazed how much better drivers have gotten in the last decade. We still have a long way to go. When were the best amps made? I think sometime in the late 70's early 80's amps got pretty good, but some new ones are still making progress. DSP has a long, long way to go. It has opportunities we have not even dreamed of yet.
I have a silly personal issue with people that substitute income for intellect, via cables etc.. but sometimes the hifi gods have a good laugh.......to witt......This guy had sold me a top shelf speaker set on Craigslist. I iater discovered he had switched a parts express crossover into one..... same day before he snaked away and still believing he was a fellow audiophile ...., he was bragging about some $2000 rca cables he had pratically stolen from a widow. I asked "what are you going to do with those silly thing?" He proudly said " I'm going to use them on my own system" So you see, Karma does happen, he has both minimal ethics and micro mentality (-:
@ tvrgeek. Well heck I have studios full of the most modern speakers to compare to, comparing in all truth is what I do for a living. Even the best of todays direct radiator can not touch older horn loaded speakers when professed with the dsp filters of today. You cant beat the laws of physics. Cone and compression drivers beyond what JBL Altec EV etc were producing in the 40s onward is truly academic and incremental.
We sure do agree on the fact that when you challenge mother nature, she always wins. But, I know of the speakers you mention. I am very glad of modern drivers. I agree we knew WHAT to build in the 50's by reading the AES papers. Material science is what is getting far better. We have better paper, better magnets, better understanding of motor flux fields etc. Maybe you are so comfortable with them and know their details so well it allows you to do your studio work well. Knowing a tool well is usually better than a better tool unknown. I would like to hear the GedLee modern application of a compression driver. What they have done seems to make a lot of sense to me, but they are too big physically for any of my rooms.
DSP gets better all the time, but we still have a long way to go. I am no analog purist, and from my work in communications I am somewhat versed on DSP. Heck, just a few months ago someone came out with a totally new algorithm for doing FFT's. About 80% more efficient. Right now for the DIY, about the best we can do is Bodzio UE running on a PC. I find that not suitable. ( the PC part) The miniDSP is a possibility, but I have not heard it. Kind of expensive to go into blind. I have played with various consumer eq (Audessy and several others) finding them seriously flawed. I use a cheap Behringer DCX on my workbench for quick crossover verification. I could not tolerate it in my system. Not sure I would tolerate it in a club PA.
I think its ABX time for a few of us here. Where is Floyde Toole when you need him? FYI neither the Minidsp or the Behringer units survive using them ADDA, ya must must use digital in to even have a chance of clarity. I am now using no external crossover or processing. I have a 3 way crossover running completely in software running J.River with LR24 and 48 filters via Motu HD 192. Also working up linear phase dsp (a work in progress) My Behringer DCX units, (which btw I feel sounded great doing DA only chores with the direct out mod via caps) sits unused now. The mini DSP was IMO pretty ragged as ADDA the unit I tried was 48k yuck.
I tweak plenty you see, but I dont fool around with bullcrap inaudible tweaks like cables. Client impressions mean everything, these guys actually listen.
I tweak plenty you see, but I dont fool around with bullcrap inaudible tweaks like cables. Client impressions mean everything, these guys actually listen.
Last edited:
- Status
- Not open for further replies.