Changing ground reference of general electronic devices.

Thread Starter

coinmaster

Joined Dec 24, 2015
502
LOTS of them. It is a trivial matter to throw together a circuit that is not going to like what you are doing. It requires much more care and effort to make one that doesn't care whether you do what you are trying to do or not. In order for what you are trying to do to work safely all-but-requires that the person that designed that DAC specifically set out as one of the primary design objectives to allow someone to do what you are trying to do. Did they do that? I have absolutely no idea whether they did or didn't. It all depends on what is inside that box! Our crystal balls simply are not good enough to allow us to guess what is in there.
I don't get how the circuit will function differently by shifting the ground reference. Let's just say that everything in the DAC ran on +15v. It would still run on +15v if I shifted the ground to -220v.

He is going to anyway because he is convinced he is right.
You're being arrogant. I'm not convinced of anything, I'm trying to understand.
 

WBahn

Joined Mar 31, 2012
29,979
Well he was warned. Seriously this must be only thread I have ever read where everyone, without exception, said don't do it (after the first few exploratory questions).

And the problem is that he is wrong and has no clue about what he is wrong.
To give the devil his due, it could possibly be more a case of this just being the way he learns. It is not too different from my learning style, particularly about 30 years ago. If I didn't understand something and couldn't wrap my head around the explanations I'd been given, what I tended to do was to adopt the position that my present, flawed, understanding was correct and then set out to defend it. In doing so, I forced one of two things to happen -- either someone else found a way to knock some sense into me or I knocked some sense into myself as I tried to defend my position. It wasn't that I was truly claiming that I was right and the rest of the world was wrong, it was just a means by which to focus on the discrepancy between the way I was looking at things and they way they really were. The people (including professors) that I interacted with using this approach did not always respond well after a while and, in retrospect, I understand why. Eventually I discovered that while this method is indeed valuable, it often requires that you put all your cards on the table to get others to participate.
 

Thread Starter

coinmaster

Joined Dec 24, 2015
502
to give the devil his due, it could possibly be more a case of this just being the way he learns.
At this point it's less about the dac and more about me trying to understand why a circuit would function differently at a different ground reference. This is fundamental stuff and I can't go forward if I have some fundamental flaw in my understanding. It was my understanding that ground is the zero point of a circuit no matter what it is relative to anything else in the world.
The circuit should function the same at any ground reference because it is still working on the same relative voltages.

Where is my flaw?
 

grahamed

Joined Jul 23, 2012
100
Try harder. Some of your arguments are correct but all conditions must be correct for it work.

Also consider this. If -220V is connected to ground (with whatever other adjustments are required) then the previous 0V and near 0V sections are now at +220V. This includes the input to the DAC. Is this a problem? Not if it's optical input only, or has fully opto-isolated inputs which are designed to cope with this somewhat eccentric connection both functionally and wrt to safety, otherwise probably and the entire kit, amplifier, speakers, all connections including antennas would need to be insulated to 220v.
 

WBahn

Joined Mar 31, 2012
29,979
There's ground and then there's ground. It all depends on how that terms is being used and how consistently it is being applied. In the global sense, you can throw a dart at something and whatever node it hits you can call that ground and measure all other voltages relative to it. But you have to do so consistently which means applying a shift to every voltage everywhere in the universe. If that universe includes, at some point, a metal rod shoved into the earth, then you have to apply that same shift to the earth and to everything that is connected to it (including YOU). That's were the details of what is in that box come into play -- you have to take into account the shift of the potential of everything in that box that has any way of interacting with that earth connection because now the Earth is a couple hundred volts different than it used to be.
 

Thread Starter

coinmaster

Joined Dec 24, 2015
502
because now the Earth is a couple hundred volts different than it used to be.
It's a couple hundred volts below what it used to be if it still had any connection to earth ground, but the 1:1 transformer would isolate the DAC from earth ground and if I apply -220v (below earth ground) to the AC voltage then the new reference would be -220v below ground. If I connect the neutral pin to a stable -220v then as far as the DAC is concerned, it is in its own little world where -220v is now 0v. Right? Or am I not getting it?
 

grahamed

Joined Jul 23, 2012
100
OK, points taken. We all had to learn and believe me I was a swine to teach.

Coinmaster -

You are completely correct that a circuit designed to work around zero will work if its zero is floated to any other potential. The problem is that you must maintain the float and that may not be possible.

Imagine you had a circuit with no connections other than a reference (0V often called ground, though it is obviously only ground if is in fact grounded) fed by a fully isolated supply. You could now connect the "grounds" of your circuit and the PSU to the -220V line and the circuit will still work. All voltages within the circuit would be shifted by 220V.

Now you must ensure there are no other connections - unfortunately this would include any input to the circuit.

So you break the ground line to the kit and float the circuit - this may be possible with no other changes, the kit may even be floating already and not be grounded at all so the -220V line is -220V only wrt to the internal 0V (not ground at all). You might need an isolation transformer - I would assume it.

If you can be sure the -220V line is truly floating (try measuring ALL POINTS IN KIT to ground with a meter on ohms - kit unpowered - and then on volts with kit powered) then you may chose to ground the -220V rail and the kit will not blow up and the input to your DAC will be at ground. HOWEVER IF YOU NOW TOUCH THE KIT ANYWHERE EXCEPT THE (so called) -220V LINE YOU WILL GET A SHOCK.. IF YOU CONNECT ANY OTHER KIT THERE WILL EITHER BE SMOKE OR MORE THINGS NOT TO TOUCH.

Do you begin to see the problem?

I AM NOT ADVISING YOU TO DO ANY OF THIS.
 

grahamed

Joined Jul 23, 2012
100
Oh and you must not connect -220V straight to neutral - that's a whole other story and in my house at least would cause the whole house to disconnect form the line.
 

Thread Starter

coinmaster

Joined Dec 24, 2015
502
You bring a good point about the input and output.
The output is not an issue because the entire point of doing this is to shift the output to -220v.

Although technically the output is 3 pin XLR so one of the pins is "ground". Would leaving this pin disconnected from the amp cause any issue?

The input is usb from the computer, digital is basically PWM right? So a capacitor should work here?
Oh and you must not connect -220V straight to neutral - that's a whole other story and in my house at least would cause the whole house to disconnect form the line.
Well by connecting -220v to neutral I meant the neutral to the DAC not neutral to the wall.
"If" I did go about doing this, I would make sure it was in a case protected from everything around it. I would also realistically probably use a different "cheapo" dac as an initial trial as it scares me to use my expensive dac as a test subject.
 
Last edited:

WBahn

Joined Mar 31, 2012
29,979
You bring a good point about the input and output.
The output is not an issue because the entire point of doing this is to shift the output to -220v.

Although technically the output is 3 pin XLR so one of the pins is "ground". Would leaving this pin disconnected from the amp cause any issue?
That depends on circuitry that you know nothing about.

"If" I did go about doing this, I would make sure it was in a case protected from everything around it. I would also realistically probably use a different "cheapo" dac as an initial trial as it scares me to use my expensive dac as a test subject.
Remember, what happens depends on the details of circuitry that you know nothing about. You could very well see that it works just fine with the cheapo DAC and then destroy your expensive DAC immediately thereafter. Why? Because the two DACs have different circuitry inside them and it all depends on the details of the circuitry -- that you know nothing about.

So you must work at a tighter level of proof -- namely you have to prove that it will work REGARDLESS of ANY of the details of not only what is inside the DAC, but also what is inside EVERY piece of equipment that connects to the DAC. That will probably be impossible to do.
 

Thread Starter

coinmaster

Joined Dec 24, 2015
502
That depends on circuitry that you know nothing about.
Well I know about the amp, I just don't know what the purpose of the XLR ground is. Normally it would simply be connected to the amp ground that is at the same reference anyway. Makes me wonder why bother with the extra pin.

Remember, what happens depends on the details of circuitry that you know nothing about. You could very well see that it works just fine with the cheapo DAC and then destroy your expensive DAC immediately thereafter. Why? Because the two DACs have different circuitry inside them and it all depends on the details of the circuitry -- that you know nothing about.

So you must work at a tighter level of proof -- namely you have to prove that it will work REGARDLESS of ANY of the details of not only what is inside the DAC, but also what is inside EVERY piece of equipment that connects to the DAC. That will probably be impossible to do.
Maybe what I need to do is to build my own dac then since I've already gone into R&D of building a pretty advanced tube amplifier.

Quality dacs take years to develop though don't they?
 
Last edited:

Papabravo

Joined Feb 24, 2006
21,159
Well I know about the amp, I just don't know what the purpose of the XLR ground is. Normally it would simply be connected to the amp ground that is at the same reference anyway. Makes me wonder why bother with the extra pin.


Maybe what I need to do is to build my own dac then since I've already gone into R&D of building a pretty advanced tube amplifier.

Quality dacs take years to develop though don't they?
The DAC project I was part of took 3 years from 2008-2011. I was one of four engineers.

http://www.tonepublications.com/review/wadia-121-decoding-computer/
 

Papabravo

Joined Feb 24, 2006
21,159
What exactly takes so long to develop them?
The control processor was an Atmel ATMega 2651, and the programming was about 32,000 lines of C code. There was a Cyclone II FPGA from Altera that implemented a DSP to do the digital filtering, and stuff like that does not just spring to life. There is a certain amount of trial and error involved in handling all the options and squeezing more functionality out of a 16 button remote than you would imagine. The PC board required four turns for reasons I don't feel the need to explain. Probably the biggest impediment to quick release was that we kept changing the requirements and allowed feature creep, and in the last revision some things were eliminated.

The power to the unit is from a purchased "power brick" with a custom DC power cord and connector. At least we didn't have to cram a transformer inside the box.
 

Thread Starter

coinmaster

Joined Dec 24, 2015
502
Hmmmmmm. I've only gone into the basics of digital and C programming. I intend to go deeper soon, but I don't see how a basic DAC would require much or any C programming. I'm assuming all that code was for the additional features?
How bout making a bare bones functional, but high quality DAC? Does the DAC itself take that much time to develop?
 

Papabravo

Joined Feb 24, 2006
21,159
The actual DAC was a chip with an I2C interface. It takes the two channel digital output from the digital receiver (CS8416) at 192 kHz. and converts that to an analog output. The SABRE-DAC (ESS9018) chip has over 120 bytes of parameters that must be programmed over the I2C interface to access all of its functions. Sure, your mileage may differ especially if you're not making a product to sell. Knock yourself out.

Oh...I forgot to mention the maximally flat FIR filter implemented in the FPGA. Digital in and digital out with 32x32 multipliers and 64 bit accumulators.

https://myl8test.files.wordpress.com/2013/05/download-es9018-datasheet-here.pdf

https://www.cirrus.com/en/pubs/proDatasheet/CS8416_F3.pdf
 
Last edited:

WBahn

Joined Mar 31, 2012
29,979
Hmmmmmm. I've only gone into the basics of digital and C programming. I intend to go deeper soon, but I don't see how a basic DAC would require much or any C programming. I'm assuming all that code was for the additional features?
How bout making a bare bones functional, but high quality DAC? Does the DAC itself take that much time to develop?
There is almost always much more complexity than what is apparent from the outside. A huge amount of it comes about just by saying that you want "high quality". As the saying goes, you can have it quick, cheap, or good -- any two of the three. You also have to define what "high quality" even means.
 
Top