You weren't supposed to leave it in!!!!!!!
The pad was only there to isolate the cable to make a nice transition for your "TDR".
But since you did leave it on , you can now tell what the input of your receiver looks like.
I have no idea why your RX will not work with a 5m cable. I routinely use one longer for my tests. If it looksl ike 75R on your "TDR", then it should work.
Anyway.....sounds like Guido's clock is not 75R...at least not yours.
You do need to have 75 ohms at both ends for SPDIF. Reflection are reflections, no matter where they occur. On a TDR, you can have a mismatch at the transmit side, but only if you know how to ignore them.
The distortions that he is seeing indicate that there is an impedance mismatch.
Jocko
The pad was only there to isolate the cable to make a nice transition for your "TDR".
But since you did leave it on , you can now tell what the input of your receiver looks like.
I have no idea why your RX will not work with a 5m cable. I routinely use one longer for my tests. If it looksl ike 75R on your "TDR", then it should work.
Anyway.....sounds like Guido's clock is not 75R...at least not yours.
You do need to have 75 ohms at both ends for SPDIF. Reflection are reflections, no matter where they occur. On a TDR, you can have a mismatch at the transmit side, but only if you know how to ignore them.
The distortions that he is seeing indicate that there is an impedance mismatch.
Jocko
Re: You weren't supposed to leave it in!!!!!!!
Well you say it and for S/PDIF you are right. Termination at the transmitting side is only necessary if there is a mismatch at the receiving side. It then helps dampen the reflected signal. But ideally when a line is properly terminated at the receiving end there are no reflections.
Nevertheless my diy passive 1:20 probes work perfectly well for high speed digital work. These consist of a 950 ohm smd resistor soldered to the core of RG316U 50 ohm coax. Other side terminated by the scoop’s internal 50 ohms.
Back to the CS8414. It is my experience that this one is quite sensitive to the proper input signal level. I myself have put a SN75157 differential line receiver in front of it. It has 2 receivers, one is used directly for S/PDIF and the other transformer coupled for AES/EBU.
Jocko Homo said:
Well you say it and for S/PDIF you are right. Termination at the transmitting side is only necessary if there is a mismatch at the receiving side. It then helps dampen the reflected signal. But ideally when a line is properly terminated at the receiving end there are no reflections.
Nevertheless my diy passive 1:20 probes work perfectly well for high speed digital work. These consist of a 950 ohm smd resistor soldered to the core of RG316U 50 ohm coax. Other side terminated by the scoop’s internal 50 ohms.
Back to the CS8414. It is my experience that this one is quite sensitive to the proper input signal level. I myself have put a SN75157 differential line receiver in front of it. It has 2 receivers, one is used directly for S/PDIF and the other transformer coupled for AES/EBU.
Re: You weren't supposed to leave it in!!!!!!!
I have no idea why your RX will not work with a 5m cable. I routinely use one longer for my tests. If it looksl ike 75R on your "TDR", then it should work.
---------------------------------------------------
I am know why; it's to do with these bare centre pin 75R sockets.
The heavy cables (5m) sag somewhat on the socket and there is not a good connection. Pinched the centre pin and all is well.
Curiously, with everthing fixed, my 1m propriety cables still sound better than the Belden 1505A, with better resulution of instruments and space and image. They also give a better square wave with virtually no tilt in the top hat. Lower bandwidth of 1505 despite the 3 GHz label??
I have no idea why your RX will not work with a 5m cable. I routinely use one longer for my tests. If it looksl ike 75R on your "TDR", then it should work.
---------------------------------------------------
I am know why; it's to do with these bare centre pin 75R sockets.
The heavy cables (5m) sag somewhat on the socket and there is not a good connection. Pinched the centre pin and all is well.
Curiously, with everthing fixed, my 1m propriety cables still sound better than the Belden 1505A, with better resulution of instruments and space and image. They also give a better square wave with virtually no tilt in the top hat. Lower bandwidth of 1505 despite the 3 GHz label??
About the tilt...........
Is it going up or down at the tail end of the square wave??
I see I still can't type worth a flip. Glad you deciphered it.
Jocko
Is it going up or down at the tail end of the square wave??
I see I still can't type worth a flip. Glad you deciphered it.
Jocko
It affects the timing. Square waves yield the most accurate timing (less jitter). If it is not square, it is that much harder to tell where a bit begins and ends. (At least that's my take on it)
Stu
Re: About the tilt...........
[Is it going up or down at the tail end of the square wave??
Jocko [/QUOTE]
-----------------------------------------------
No, what happens with the Belden is one to two cyles of distorted ringing at the leading edge of the top hat, then a very gentle tilt topwardes the tail compared with the others.
Fred
[Is it going up or down at the tail end of the square wave??
Jocko [/QUOTE]
-----------------------------------------------
No, what happens with the Belden is one to two cyles of distorted ringing at the leading edge of the top hat, then a very gentle tilt topwardes the tail compared with the others.
Fred
Re: About the tilt...........
I broadly agree, timing better, resolution of instruments improves, less 1D layout, etc. Is it jitter though as these are also a fuction of the cable itself?
--------------------------------------------maczrool said:
I broadly agree, timing better, resolution of instruments improves, less 1D layout, etc. Is it jitter though as these are also a fuction of the cable itself?
Hi,
Is that bean 37 you just spilled there?
Come'on folks transmissionlines aren't all that hard now, or are they?
Cheers,😉
Is that bean 37 you just spilled there?
Come'on folks transmissionlines aren't all that hard now, or are they?
Cheers,😉
maczrool said:
Thanks all!
Trying honestly not to be difficult, but surely the slope or error or whatever of the circuit will be constant. The DAC or whatever will still trigger on the same rise point, yes, it may be delayed a little, but the waveshape is the same from bit to bit. Ringing and overshoot is irrelevent surely. I will try and draw a diagram of what I mean later, but at the moment I need to sleep😉
Jocko Homo said:
I don't know enough about the subject to pick a fight, ( not that that stops some people😉 ), just honestly interested!
Trying honestly not to be difficult, but surely the slope or error or whatever of the circuit will be constant. The DAC or whatever will still trigger on the same rise point, yes, it may be delayed a little, but the waveshape is the same from bit to bit. Ringing and overshoot is irrelevent surely.
------------------------------
You can hear the differences between a nicely matched line and a poorly matched one and good jitter busters do that as well. This is for the same cable. On top, you have the differences amongst cables as well.
On a home built upsampler/dac, I can hear the differences and good digital then begins to sound like good analogue. None of that 1D aggressive upfront digititice.
------------------------------
You can hear the differences between a nicely matched line and a poorly matched one and good jitter busters do that as well. This is for the same cable. On top, you have the differences amongst cables as well.
On a home built upsampler/dac, I can hear the differences and good digital then begins to sound like good analogue. None of that 1D aggressive upfront digititice.
On a lossy line, the back wave is attenuated, so the slope of the display on a TDR will have an upward tilt. Works the same way for a very long line also. (By very long, I am talking far too long for SPDIF use.)
Jocko
Jocko
pinkmouse said:
Sorry, but no. Check out the AES app notes on data-dependent errors. We're also concerned about ringing about the transition point...this is a violation of the setup-and-hold requirements and the behavior of the receiver may not be defined. It isn't simply a constant error.
- Status
- Not open for further replies.