That is hardly any change for 30 years. Imagine how much computing has advanced. You also exaggerate and I assume you know. LISP? I programmed in it but it was a tiny blip in computing world. Smalltalk? You have to be kidding. Major programming languages were Fortran, COBOL, Basic, and then C came about and over a decade or so dominated the computing industry.
And C++ for financial services? Where on earth did that come from? My son as a hobby works an open source game and it is all written in C++. I worked for the world's largest software company and when I say C/C++ and its lineage dominate, that is the case, not some corner case.
But let's put all that aside because we are talking about a narrow field in this topic. Namely, embedded programming where a computer is used as part of an end-part that is not a computer as far as the customer is concerned. In this world by far C language dominates. For custom OS development it is king because it is close to hardware and you can't afford to have garbage collection. Linux operating system dominates in audio products and its kernel is entirely written in C language. And ask any embedded engineer what language they like to use and by far the will tell you C language -- the same language that dates back to early 1970s or more than 40 years ago. Windows kernel (core OS) is also written in C. As is MacOS.
Your experience must not have extended to embedded and OS development to hold the views you are expressing.
Good Morning Amir,...I thought about a very lengthy reply here publically to put some facts about the software industry from other than a totally MSFT-centric point of view, my background and experience, out there in full view BUT it's frankly not right and respectful to other forums members and even though I have it drafted and ready to send, I will not. I will say that my background includes alot more than you think and we also have a bit more common ground than you would believe. Instead, I am sending you a PM with all that with hopes that we can get this topic back on track from the point before I respectfully disagreed with SOME of your points and agreed 100% with others. Have a great day!
I don't care what kind of equipment it is if it has current going through it,that creates EMI or "noise" and that degrades the signal. Even a better cable will not cure the problem totally as there will be a residual amount left.....
I don't care what kind of equipment it is if it has current going through it,that creates EMI or "noise" and that degrades the signal. Even a better cable will not cure the problem totally as there will be a residual amount left.....
I don't care what kind of equipment it is if it has current going through it,that creates EMI or "noise" and that degrades the signal. Even a better cable will not cure the problem totally as there will be a residual amount left.....
It's rather unfortunate: I've come to the conclusion working on my latest iteration of my USB cable that the whole of the connection (that is extending to the connectors) is inherently flawed, especially if you're doing real-time audiophile streaming listening. Specifically for USB, we have power + noise riding alongside the data lines, but apparently the power ground computer-side is also connected to the Jack shell and to the GND pin there. DAC-side, the jack shell and the signal GND are connected (?) and the GND pin is also connected to that (?).
Result: utterly noisy environment for the USB receiver to work in, itself frantically creating a lot of noise, thereby affecting the DAC chip and clock.
Now, you can also split a cable and make it a Y-cable and inject clean power to the DAC, but within the jack, the connectors run close and in parallel again - they really should be at right angles. It's messed up beyond recognition.
At least this is what I find with my DAC - its USB Receiver is powered by USB and the computer USB messes up the SQ a lot! (YMMV). Totally self-powered DACs may not be as susceptible or may be immune to a large extent.
Now, at least with Ethernet we have the benefits of some isolation through each tiny transformer withing the jacks, but, I'll know for sure when I get to doing some experiments with these as well.
It's rather unfortunate: I've come to the conclusion working on my latest iteration of my USB cable that the whole of the connection (that is extending to the connectors) is inherently flawed, especially if you're doing real-time audiophile streaming listening. Specifically for USB, we have power + noise riding alongside the data lines, but apparently the power ground computer-side is also connected to the Jack shell and to the GND pin there. DAC-side, the jack shell and the signal GND are connected (?) and the GND pin is also connected to that (?).
Result: utterly noisy environment for the USB receiver to work in, itself frantically creating a lot of noise, thereby affecting the DAC chip and clock.
Now, you can also split a cable and make it a Y-cable and inject clean power to the DAC, but within the jack, the connectors run close and in parallel again - they really should be at right angles. It's messed up beyond recognition.
At least this is what I find with my DAC - its USB Receiver is powered by USB and the computer USB messes up the SQ a lot! (YMMV). Totally self-powered DACs may not be as susceptible or may be immune to a large extent.
Now, at least with Ethernet we have the benefits of some isolation through each tiny transformer withing the jacks, but, I'll know for sure when I get to doing some experiments with these as well.
If the unit is in a enclosure...what you might try is to run a strap from the enclosure to the duplex ground (the one that the unit is powered from). I use 6 awg wire and terminate with 10 awg connectors. All the EMI migrates to the underside of the metal enclosure. You could also improve the grounding scheme of the enclosure itself,but I have found this is easer. If you have a Tripoint which is actually a "sink" you could tie it to that also,but might not be as good,but should make a difference.
Here's something that'll help you out, from the basis of electronics.
Voltage only divides; so either you have some (which may be infinitesimally small) or it's not turned on! You're always trying make the division in your favor.
Here's something that'll help you out, from the basis of electronics.
Voltage only divides; so either you have some (which may be infinitesimally small) or it's not turned on! You're always trying make the division in your favor.
It's rather unfortunate: I've come to the conclusion working on my latest iteration of my USB cable that the whole of the connection (that is extending to the connectors) is inherently flawed, especially if you're doing real-time audiophile streaming listening. Specifically for USB, we have power + noise riding alongside the data lines, but apparently the power ground computer-side is also connected to the Jack shell and to the GND pin there. DAC-side, the jack shell and the signal GND are connected (?) and the GND pin is also connected to that (?).
Currents being carried via Vbus wire & returned via ground wire in a USB cable in order to power a USB device, is not a good design from an analogue perspective but remember the design is mainly concerned with digital transmission & digital thinking believes that's all that is needed. Some USB cables use foil shielding around the USB power lines & separately around the USB signal lines which gives some measure of isolation. BTW, the whole shields issue can be complicated even when the shields & grounds are connected properly at both ends. In the USB receiver end, the shield should not be connected directly to ground. Why? Because of something called SCIN: Shield-current-induced noise - read about it here. As can be seen this is related to a well known problem called the "pin 1 problem" which was delivered in a 1994 AES paper & audio designers & users should know about - essentially it's about how RF noise can enter audio equipment & cause audible issues. Different cable construction (foil/braid, etc) & different connection configurations lead to different audible results
Result: utterly noisy environment for the USB receiver to work in, itself frantically creating a lot of noise, thereby affecting the DAC chip and clock.
Now, you can also split a cable and make it a Y-cable and inject clean power to the DAC, but within the jack, the connectors run close and in parallel again - they really should be at right angles. It's messed up beyond recognition.
I don't believe the last inch or so of USB power running in parallel to USB signal has much crosstalk?
At least this is what I find with my DAC - its USB Receiver is powered by USB and the computer USB messes up the SQ a lot! (YMMV). Totally self-powered DACs may not be as susceptible or may be immune to a large extent.
There is a certain level of noise which can be reduced by self-powered USB receivers but even with this one finds that noise is actually riding on the USB signal lines which still audibly effects the sound.
Now, at least with Ethernet we have the benefits of some isolation through each tiny transformer withing the jacks, but, I'll know for sure when I get to doing some experiments with these as well.
Yes, the ethernet transformers, both differential & common mode are both needed, should provide some level of isolation but probably not complete isolation?
If the unit is in a enclosure...what you might try is to run a strap from the enclosure to the duplex ground (the one that the unit is powered from). I use 6 awg wire and terminate with 10 awg connectors. All the EMI migrates to the underside of the metal enclosure. You could also improve the grounding scheme of the enclosure itself,but I have found this is easer. If you have a Tripoint which is actually a "sink" you could tie it to that also,but might not be as good,but should make a difference.
If you're speaking of chassis grounding, I have experimented with that but it was with the iMac (3-prong AC), DAC (USB-powered) and SS amp (2-prong AC).
With my recent experiments with a new DIY USB cable and with the DAC instead connected to my Tube Amp (3-prong AC), I've had to disconnect the chassis-grounding as perhaps a ground loop was bugging the setup.
Thus, I have been able to focus specifically of re-architecting the USB connection.
Results are stellar. I will most probably re-test chassis grounding at a later point.
Currents being carried via Vbus wire & returned via ground wire in a USB cable in order to power a USB device, is not a good design from an analogue perspective but remember the design is mainly concerned with digital transmission & digital thinking believes that's all that is needed.
I know! Around 2014 I had built my own USB cable with separately double-shielded and physically separated power lines and data lines. Here's a 2015 update about it. This was my favourite cable although quite long. From time to time it was de-commissioned (e.g. when I was testing a USB 3.0 cable from my girlfriend's Verbatim external HDD iwht my DAC which also has a USB 3.0 connector/input). But every time I put that cable in, we were floored.
The very latest iteration of my DIY cable (also contains a handful of components on the power lines) drove me to that conclusion. I must be on a v.12 of that USB connection. I finally got something utterly satisfying after several implementations.
In the USB receiver end, the shield should not be connected directly to ground. Why? Because of something called SCIN: Shield-current-induced noise - read about it here. As can be seen this is related to a well known problem called the "pin 1 problem" which was delivered in a 1994 AES paper & audio designers & users should know about - essentially it's about how RF noise can enter audio equipment & cause audible issues. Different cable construction (foil/braid, etc) & different connection configurations lead to different audible results.
I recommend you test it to find out for yourself. We've seen time and time again how assumptions hide a different reality for the simplest of things we've been taught or can read about. I have seen reports of people using a piece of cardboard around the disconnected power lines computer-side and still get a bump in SQ because of supposed hot-line induction...
Now, if the shield, which is further than the power lines in a USB cable can affect the data lines through a magnetic field, the power lines can't? The latter are closer...
There is a certain level of noise which can be reduced by self-powered USB receivers but even with this one finds that noise is actually riding on the USB signal lines which still audibly effects the sound.
Indeed, and that's despite the data lines being differential (whereby people assume all noise issues are solved by using such an interface and coding)...
Yes, the ethernet transformers, both differential & common mode are both needed, should provide some level of isolation but probably not complete isolation?
For now, I am on USB, but when I implement Ethernet with say, HQ Player and an NAA, I'll make sure to explore this taking into account my findings with USB.
Here's a pic of my latest DIY USB cable experiment (Glorious sound although I'm powering the DAC currently with a simple iPhone charger, to be replaced later by batteries or my DIY Linear Regulated PSU):
If you're speaking of chassis grounding, I have experimented with that but it was with the iMac (3-prong AC), DAC (USB-powered) and SS amp (2-prong AC).
With my recent experiments with a new DIY USB cable and with the DAC instead connected to my Tube Amp (3-prong AC), I've had to disconnect the chassis-grounding as perhaps a ground loop was bugging the setup.
Thus, I have been able to focus specifically of re-architecting the USB connection.
Results are stellar. I will most probably re-test chassis grounding at a later point.
My experience is some tube amps have excellent grounding schemes,which could actually in your case make further chassis grounding degrade performance. Glad it worked out...experimentation is a wonderful thing.
I recommend you test it to find out for yourself. We've seen time and time again how assumptions hide a different reality for the simplest of things we've been taught or can read about.
