[nopietns]

audioengr,, You've worked with the North star models quite a bit, any ideas if the I2S signal from the Juli@ would be ok for North star M192 DAC?

 

[audioengr]

Quote:

Originally Posted by dahlhaus /img/forum/go_quote.gif I've found out something strange. My multimeter says the voltage of the signal is half the normal. The three clock pins are 1.65V while the data pin is only 0.85V. The multimeter reads 5V normally from USB. What can I do?

If the clock is running, your multimeter will only show 1/2 the voltage. Same with a data line. You need an oscilloscope to do this right, not a multimeter. You need to be able to see the signals and verify correct signal connections and good signal integrity at the terminators on the P-3A board.

Steve N.

 

[audioengr]

Quote:

Originally Posted by nopietns /img/forum/go_quote.gif audioengr,, You've worked with the Northstar models quite a bit, any ideas if the I2S signal from the Juli@ would be ok for Northstar M192 DAC?

If it's 5V level I2S, it should work with the 192. BTW, the Northstar DAC's I2S input is a bit compromised soundwise. It has a lot of strange terminations to eliminate RFI I think. I redo all of this in my mods.

If you are technical enough to do this, I can give you the pinout of the Northstar RJ-45.

Steve N.
Empirical Audio

 

[nopietns]

Huh, when I emailed the North star design support on the layout of the RJ-45, the answer was:
Quote:

On the RJ45 connector there are 8 connection used in this way: 4 for the signals and 4 for digital ground. To connect the card to the DAC you need to have 4 signals SDATA Serial Data SCK Serial Clock LRCK Left/Right clock MCLK Master Clock

Are You sure the other 4 pins are not just plain ground?

 

[audioengr]

Quote:

Originally Posted by nopietns /img/forum/go_quote.gif Huh, when I emailed the North star design support on the layout of the RJ-45, the answer was: Are You sure the other 4 pins are not just plain ground?

There is no such thing as "plain ground".

There are four signals as Guiseppe describes and then there are four returns, which connect to ground, eight wires in all, four pairs.

Steve N.
Empirical Audio

 

[nopietns]

I'm not sure I'm technical enough to know where should I attach the 4 different grounds.

And the Stello DA100 Sig uses only one common ground for all the 4 signals:
pin1 : word clock
pin2 : bit clock
pin3 : audio data
pin4 : master clock
pin5 : not connected
shield(shell) : circuit ground

So are the 4 different grounds just a fancy design of the North star Transport/DAC?

So, - If someone doesn't just tell me where to attach the 4 grounds, I simply won't be using the Juli@ => NS M192 trough I2S.

 

[FallenAngel]

Quote:

Originally Posted by nopietns /img/forum/go_quote.gif So are the 4 different grounds just a fancy design of the North star Transport/DAC?

That's not particularly nice to say, you do need the ground.

Quote:

Originally Posted by nopietns /img/forum/go_quote.gif So, - If someone doesn't just tell me where to attach the 4 grounds, I simply won't be using the Juli@ => NS M192 trough I2S.

No big loss there for the rest of us.

As your email told you, Connect the damn 4 wires to DIGITAL GROUND!

 

[nopietns]

Hey, I didn't mean to be rude or tell that You should feel sorry that I won't be using the I2S. I didn't know the word "fancy" is that abusive.

It's just that if some designs use a common ground for all the signals, but audioengr says that each ground is for it's own signal (and I beliehe he knows what he's saying (at least when talking about North Star Design)), it makes me wonder how important the differentiation of the grounds are for the NS M192 DAC.

 

[audioengr]

Quote:

Originally Posted by nopietns /img/forum/go_quote.gif Hey, I didn't mean to be rude or tell that You should feel sorry that I won't be using the I2S. I didn' know the word "fancy" is that abusive. It's just that if some designs use a common ground for all the signals, but audioengr says that each ground is for it's own signal (and I beliehe he knows what he's saying (at least when talking about North Star Design)), it makes me wonder how important the differentiation of the grounds are for the NS M192 DAC.

The I2S cabling should have 4 grounds, even if they are tied to the same ground-plane at the source and at the destination. The wires should be twisted-pairs or coaxial cables with correct impedance to match the terminations.

This is simply batteries and bulbs, not rocket science. The "ground" wire for each signal provides a separate "return" path for each signal. Remember, in order to light a light bulb, you need both a forward current wire and a return current wire. Same with digital, you need two wires. If you make all of the signals share the same return wire, then the signals will crosstalk onto each other. This is the case for the Perpetual and the Stello products. Not optimum.

Steve N.

 

[FallenAngel]

Quote:

Originally Posted by audioengr /img/forum/go_quote.gif The I2S cabling should have 4 grounds, even if they are tied to the same ground-plane at the soruce and at the destination. The wires should be twisted-pairs or coaxial cables with correct impedance to match the terminations. This is simply batteries and bulbs, not rocket science. The "ground" wire for each signal provides a separate "return" path for each signal. Remember, in order to light a light bulib, you need both a forward current wire and a return current wire. Same with digital, you need two wires. If you make all of the signals share the same return wire, then the signals will crosstalk onto each other. This is the case for the Perpetual and the Stello products. Not optimum. Steve N.

Interesting, I've never seen it implemented any other way. Are you suggesting that the "optimal solution" would be to have 4 digital grounds? I don't see how that could actually work, could you please elaborate.

Thanks

 

[audioengr]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif Interesting, I've never seen it implemented any other way. Are you suggesting that the "optimal solution" would be to have 4 digital grounds? I don't see how that could actually work, could you please elaborate. Thanks

This is not what I said at all. I said that the four return paths in the cable should be separated and isolated from each other for best performance.

Even in the common ground-plane in the PC board, the four signal return currents take different paths. This is because RF and digital signalling is not like DC. High-frequency return currents follow the nearest return path. This is not about voltages, it is about currents. AC currents are created by EM fields, not the other way around. This is basic first year EE school fields 101.

Steve N.

 

[dahlhaus]

Quote:

Originally Posted by audioengr /img/forum/go_quote.gif Having modded a LOT of P-3A DAC's in the past (not anymore), and having several I2S interfaces that can drive this, I can give you some tips on this: 1) the P-3A needs 5V signal levels, not 3.3V 2) The inputs are terminated with 100 ohms to ground 3) MCLK needs to be 128Fs or 96kHz will not work (this is usually 256Fs (Fs is the sample-rate frequency)) You will probably have to divide MCLK by 2. 4) the input receiver on the P-3A for the I2S is a rather slow RS-422 receiver - (I replaced it in my mods with a faster device and changed the voltage to 3.3V)

I'm using DA100 Signature, not P-3A. But yesterday I've contacted an April Music guy and he said DA100 Signature needs 5V signal for I2S.

And he also said the clock is 256Fs. I don't know the clock frequency of Juli@ - ESI has not answered my question yet.

Quote:

Originally Posted by audioengr /img/forum/go_quote.gif If the clock is running, your multimeter will only show 1/2 the voltage. Same with a data line. You need an oscilloscope to do this right, not a multimeter. You need to be able to see the signals and verify correct signal connections and good signal integrity at the terminators on the P-3A board. Steve N.

By the way, is it normal that the voltage of the data signal is half of that of the clock signals?

 

[audioengr]

Quote:

Originally Posted by dahlhaus /img/forum/go_quote.gif is it normal that the voltage of the data signal is half of that of the clock signals?

The data signal is changing all the time. The clock signal is repetitive. The average DC on the clock signal is probably 2.5V if the signal level is 5V. The data signal will have a lower average DC level. These average DC levels are not important. What is important is how it looks on an oscilloscope using a good probe with low-inductance ground. I think this is probably too technical for you to attempt. If you dont have a scope, you are flying blind.

Steve N.
Empirical Audio

 

[nopietns]

However - looks like the I2S signal is a 3.3V:
VIA Vinyl Envy24HT PCI Audio Controller - VIA Technologies, Inc.

Wonder how it corresponds to the OP, if the P-3A needs 5V signal levels, not 3.3V, but still got it working.

 

[dahlhaus]

Quote:

Originally Posted by audioengr /img/forum/go_quote.gif The average DC on the clock signal is probably 2.5V if the signal level is 5V.

Now that nopietns has verified that Envy chip has 3.3V I2S signal, I wonder which converter should I use? 3.3V-to-5V one or 1.65V-to-2.5V one?