I have a question regarding using the digital input on the DCX. Per a few messages from a couple pages back in this thread, I built a cable with an XLR plug on one end and an RCA on the other end. The purpose is to run the digital output from my receiver to the DCX digital input (input C.) From my reading of the manuals it appears both use SPDIF. Also, by way of background, my receiver (Denon DN-A7100 http://www.d-mpro.com/users/folder.asp?FolderID=4214&Tab=Data+Sheet ) is in a rack with only A/V sources and processing. There is a separate rack about 12' away that has all of the amplifiers and the DCX2496.
The DCX received a ton of noise and a very faint signal in the background. I tried a few different settings in the DCX and nothing seems to make a difference in the noise level. There isn't much detail in the manual about connecting the digital input.
The cable is a twisted pair with a ground wire that is running in a snake alongside the balanced audio lines from the receiver to the amplifier rack. This snake is about 20' long.
First, I thought maybe the line was too long so I placed the two racks side by side and ran a 6' cable from the receiver's digital out to the DCX and it does the same thing.
Specifically, I have the ground wire on pin 1, the hot on pin2 and the cold on pin3 of the XLR plug. On the other end, I have the ground and the cold wire soldered to the ground connection of the RCA plug and the hot wire on the pin of the RCA plug.
Any ideas why I am getting so much noise?
Thanks very much.
The DCX received a ton of noise and a very faint signal in the background. I tried a few different settings in the DCX and nothing seems to make a difference in the noise level. There isn't much detail in the manual about connecting the digital input.
The cable is a twisted pair with a ground wire that is running in a snake alongside the balanced audio lines from the receiver to the amplifier rack. This snake is about 20' long.
First, I thought maybe the line was too long so I placed the two racks side by side and ran a 6' cable from the receiver's digital out to the DCX and it does the same thing.
Specifically, I have the ground wire on pin 1, the hot on pin2 and the cold on pin3 of the XLR plug. On the other end, I have the ground and the cold wire soldered to the ground connection of the RCA plug and the hot wire on the pin of the RCA plug.
Any ideas why I am getting so much noise?
Thanks very much.
Graydon said:
http://www.behringerdownload.de/DCX2496/DCX2496_ENG_Rev_B.pdf page 4.
Input A is switchable between Analog or Digital...
Long Digital leads
I'm running an S/PDIF signal on a 10m (30') cable with minimal troubles into my DCX.
I started with a 10m video rca-rca lead (75 ohm characteristic impedance). I chopped the RCA off one end and wired an XLR connector in its place, the core and shield across pins 2 and 3 respectively. I assembled two 470 ohm resistors in parallel (resulting resistance 235 ohms) and placed these across pins 2 and 3 also in order to match the 110 ohm input of the DCX to the 75 ohm SPDIF cable and output.
Using this arrangement alone worked reasonably well although there would be the odd click or pop when plugging other things into the same power point as the cd player (since the cd player was double insulated there was probably a significant amount of common mode noise being unearthed). I connected a 62 ohm resistor between the shield of the 75 ohm coax (pin 3) and pin 1 on the XLR to earth the system and this eliminated 99% of the clicks when plugging things in and entirely eliminated any random clicks and pops.
I have on occasions noticed that the system must lock onto the signal wrong and it sounds very odd, no top end and lots of background noise. Unplugging and plugging the lead back in usually has the DCX lock on properly, no idea of the cause of the problem though.
Cheers,
Duncan
Using the above lead I've run the system off generator and off mains and had no major troubles either way.
I'm running an S/PDIF signal on a 10m (30') cable with minimal troubles into my DCX.
I started with a 10m video rca-rca lead (75 ohm characteristic impedance). I chopped the RCA off one end and wired an XLR connector in its place, the core and shield across pins 2 and 3 respectively. I assembled two 470 ohm resistors in parallel (resulting resistance 235 ohms) and placed these across pins 2 and 3 also in order to match the 110 ohm input of the DCX to the 75 ohm SPDIF cable and output.
Using this arrangement alone worked reasonably well although there would be the odd click or pop when plugging other things into the same power point as the cd player (since the cd player was double insulated there was probably a significant amount of common mode noise being unearthed). I connected a 62 ohm resistor between the shield of the 75 ohm coax (pin 3) and pin 1 on the XLR to earth the system and this eliminated 99% of the clicks when plugging things in and entirely eliminated any random clicks and pops.
I have on occasions noticed that the system must lock onto the signal wrong and it sounds very odd, no top end and lots of background noise. Unplugging and plugging the lead back in usually has the DCX lock on properly, no idea of the cause of the problem though.
Cheers,
Duncan
Using the above lead I've run the system off generator and off mains and had no major troubles either way.
Boye said:
Hi Boye,
Yes, my (short) vacation was nice, thanks!
On the balanced issue, there are two issues really. Firstly, why balanced lines? Because if we run two signal wires closely twisted, AND we make sure that they have identical impedances to ground, any noise, interference, etc will be impressed identically on each wire. At the receiver end we will have a differential stage that amplifies the signal only and not the noise etc because that is common mode.
Note that there is NO requiremment for the signals to be equal in amplitude! If we do drive the signals in antiphase, we get 6dB more output level from a given supply voltage which may get you 6dB more S/N ratio, depending on other factors. But it has NO effect on the advantages of balanced interconnects to reject noise, hum, buzz, etc. In fact, you can get impedance balance with one wire driven from an opamp through a defined output impedance, and the other wire to ground through the same impedance. In this scenario, only one wire carries the signal but you still have the full advantages of fully balanced interconnects.
More details and a link to a very good writeup
here
Jan Didden
Re: Long Digital leads
Thanks for your input jakobsladderz. I got my digital input going last night using a short cable into input A. The long twisted pair didn't work but a 6' coax cable worked like a charm. I will try a longer coax cable as you describe but only if I can solve another problem that surfaced once I got the 6' coax installed. My receiver outputs only full volume through the S/PDIF digital output. The receiver's volume control has no effect on the digital output volume. If I can't change that, then I won't have any need to run the digital cable to the DCX anyway. If I can change the receiver's volume control, then I will try your suggestions on the coax cable.
jakobsladderz said:
Thanks for your input jakobsladderz. I got my digital input going last night using a short cable into input A. The long twisted pair didn't work but a 6' coax cable worked like a charm. I will try a longer coax cable as you describe but only if I can solve another problem that surfaced once I got the 6' coax installed. My receiver outputs only full volume through the S/PDIF digital output. The receiver's volume control has no effect on the digital output volume. If I can't change that, then I won't have any need to run the digital cable to the DCX anyway. If I can change the receiver's volume control, then I will try your suggestions on the coax cable.
The nature of most digital outputs is to have no volume control - to make use of the full dynamic range of the digital signal.
many of us use some form of volume control after the DCX - like a 6-channel unit, either as an external device with 6-in/6-out, or by modifying the DCX internally to attenuate, by adjustment, the analog outputs.
many of us use some form of volume control after the DCX - like a 6-channel unit, either as an external device with 6-in/6-out, or by modifying the DCX internally to attenuate, by adjustment, the analog outputs.
Hi all,
I'm using DCX with digital input. When switching on/off this input signal 10-20 times (may vary) with a mechanical switch or relay there occor a situation where frequencies above 10 kHz seem to be cut off.
The PLL of the CS8420 seems to work properly. So this might be a problem of the sample rate converter in the CS8420 or some other error in the DSP.
Has anybody made same experiance?
Best Regards,
Frank
I'm using DCX with digital input. When switching on/off this input signal 10-20 times (may vary) with a mechanical switch or relay there occor a situation where frequencies above 10 kHz seem to be cut off.
The PLL of the CS8420 seems to work properly. So this might be a problem of the sample rate converter in the CS8420 or some other error in the DSP.
Has anybody made same experiance?
Best Regards,
Frank
Hello oettle,
In my system I use 2 DCX one is working perfectly the other not.
On the bad one I had same problem as you only by switch on/off my cd player.
The medium and high frequency response drop.
Turn off / on dcx cancel the problem.
Also I'can not get high digital level in 96 kHz without "clic" noise.
I make this big modification and now it is working perfectly.
http://freerider.dyndns.org/anlage/Behringer-Input-Stage-E.htm
J-C B
In my system I use 2 DCX one is working perfectly the other not.
On the bad one I had same problem as you only by switch on/off my cd player.
The medium and high frequency response drop.
Turn off / on dcx cancel the problem.
Also I'can not get high digital level in 96 kHz without "clic" noise.
I make this big modification and now it is working perfectly.
http://freerider.dyndns.org/anlage/Behringer-Input-Stage-E.htm
J-C B
Hi,
I'm aware of a lot of interfacing problems with digital input. This 10kHz cut-off problem seems to be not a problem of interfacing, because normally it works properly. It also seems to be not a problem of the PLL because RMCK of the CS8420 is OK.
I'm using a 44.1 KHz input. Sometimes the cut-off problem occurs during power-up or by simply plugging in and out the digital input connector. The DCX locks to this 10kHz cut-off error. That means if it occurs it is always there until you switch off/on the DCX or plug off/on the connector.
If there is no 10kHz cut-off problem with the Oehlbach mod most probably the reason is the SRC in the CS8420. Unfortunately I can't double-check this.
Best Regards,
Frank
I'm aware of a lot of interfacing problems with digital input. This 10kHz cut-off problem seems to be not a problem of interfacing, because normally it works properly. It also seems to be not a problem of the PLL because RMCK of the CS8420 is OK.
I'm using a 44.1 KHz input. Sometimes the cut-off problem occurs during power-up or by simply plugging in and out the digital input connector. The DCX locks to this 10kHz cut-off error. That means if it occurs it is always there until you switch off/on the DCX or plug off/on the connector.
If there is no 10kHz cut-off problem with the Oehlbach mod most probably the reason is the SRC in the CS8420. Unfortunately I can't double-check this.
Best Regards,
Frank
BTW,
I double-checked the problem with an 48-kHz input. The error is still there.
I'm using 3 DCX. They all have the same problem and they all have a CS8420 Rev. D. There is a newer version D1. See on page 88: http://www.cirrus.com/en/pubs/proBulletin/cs8420pi.pdf
I double-checked the problem with an 48-kHz input. The error is still there.
I'm using 3 DCX. They all have the same problem and they all have a CS8420 Rev. D. There is a newer version D1. See on page 88: http://www.cirrus.com/en/pubs/proBulletin/cs8420pi.pdf
Sorry wrong link. Here the correct one: http://www.cirrus.com/en/pubs/proDatasheet/CS8420_F4.pdf
I had a similar problem with one of my DCX'es. I traced it to the level (in volts, not digital value) of the S/PDIF signal going into the 8420.
The 8420 is supposed to automatically lock on both an S/PDIF and an AES/EBU signal. The difference between the two, apart from some status bits, is the voltage level: AES/EBU is at least twice the level of S/PDIF.
I solved it by building a small S/PDIF-to-AES/EBU converter that basically ups the signal level. It was an old Elektor circuit for which I made a pcb. The circuit and pcb layout is in the files section of the yahoo DCX2496 user group.
It completely solved my problem.
Jan didden
The 8420 is supposed to automatically lock on both an S/PDIF and an AES/EBU signal. The difference between the two, apart from some status bits, is the voltage level: AES/EBU is at least twice the level of S/PDIF.
I solved it by building a small S/PDIF-to-AES/EBU converter that basically ups the signal level. It was an old Elektor circuit for which I made a pcb. The circuit and pcb layout is in the files section of the yahoo DCX2496 user group.
It completely solved my problem.
Jan didden
Hi Jan,
I'm aware of those interfacing problems. I'm using standard SPDIF and a CMOS gate with 0-5 V output as well. The 10 kHz cut-off problem occurs with both input levels. I'm not talking about clicks or cracks. My audio output is sonically absolute OK even in this error mode. So far I can't imagine how digital input level should cause an analogue 10 kHz cut-off. I measured at all pins of the CS8420 with a scope and couldn't see any difference in this error mode (although I can't analyze data stream).
I assume switching on/off the digital input causes some digital undefined start sequence. Might be that a certain sequence locks the SRC or the DSP to this cut-off mode. I have this problem with all of my 3 DCXs so it seems to be a general problem.
Best Regards,
Frank
I'm aware of those interfacing problems. I'm using standard SPDIF and a CMOS gate with 0-5 V output as well. The 10 kHz cut-off problem occurs with both input levels. I'm not talking about clicks or cracks. My audio output is sonically absolute OK even in this error mode. So far I can't imagine how digital input level should cause an analogue 10 kHz cut-off. I measured at all pins of the CS8420 with a scope and couldn't see any difference in this error mode (although I can't analyze data stream).
I assume switching on/off the digital input causes some digital undefined start sequence. Might be that a certain sequence locks the SRC or the DSP to this cut-off mode. I have this problem with all of my 3 DCXs so it seems to be a general problem.
Best Regards,
Frank
Hi Jan,
This 'dull sound mode' is exactly what I'm talking about. Might be that not input level causes this problem, but digital start sequence? It also might be that your level shifter suppresses certain start sequences and so it looks like level is the reason?
I'm using a relay at my input to select different digital inputs.
Would be interesting when you implement a switch between your DCX and your level shifter whether the problem occurs again.
Regards,
Frank
This 'dull sound mode' is exactly what I'm talking about. Might be that not input level causes this problem, but digital start sequence? It also might be that your level shifter suppresses certain start sequences and so it looks like level is the reason?
I'm using a relay at my input to select different digital inputs.
Would be interesting when you implement a switch between your DCX and your level shifter whether the problem occurs again.
Regards,
Frank
Hi Jan,
Possibly the 10kHz roll-off error is that what is desribed on page 49 of the CS8420 SRC datasheet http://www.cirrus.com/en/pubs/proDatasheet/CS8420_F4.pdf with 'notches occuring in the frequency response'. Behringer uses CS8420 in hardware mode 1. So I monitored RERR but couldn't find any abnormalities. RERR goes low after 20-40 ms after detecting digital input data.
Has anybody else the same 10kHz roll-off (dull sound) problem or any idea how to go on?
Best Regards,
Frank
Possibly the 10kHz roll-off error is that what is desribed on page 49 of the CS8420 SRC datasheet http://www.cirrus.com/en/pubs/proDatasheet/CS8420_F4.pdf with 'notches occuring in the frequency response'. Behringer uses CS8420 in hardware mode 1. So I monitored RERR but couldn't find any abnormalities. RERR goes low after 20-40 ms after detecting digital input data.
Has anybody else the same 10kHz roll-off (dull sound) problem or any idea how to go on?
Best Regards,
Frank