tnargs said:
p32 is closest to my application -- one DCX per 4-way speaker. But the manual only shows analog stereo feeds into the DCX, not digital, and hence separate feed line per DCX. Nowhere does it show identical feed into dual DCX's.
It seems a bit wierd that for something as simple as my application (dual DCX from a stereo digital feed in the exact format the DCX requires) I need to hand-build an adaptor cable. I mean, where on earth can I get a 110-ohm balanced diplexer?
I don't think you can do that. When you have a 110 ohms source trying to feed two 110 ohms loads you either loose signal amplitude (by resistively padding to get the 110 ohms source load) or you get mismatch. That's probably why Behringer doesn't show it. However, if you have an AES/EBU source you can probably survive losing half the level because the DCX does accept S/PDIF (although sometimes not without problems) which is essentially AES/EBU with a much lower level.
So, you could make up a passive pad: feed each DCX input via 55 ohms series resistors from your 110 ohms load. That makes the source- and load impedances not quite right, but close. You lose a couple dB in level, but that should be OK. The proof is in the eating..
Jan Didden
Hi,
This is not a solution, but tries to cover the problem and go part way to a solution.
I'm not up on this aerial stuff, but that ref to 55r should be 110r.
I think each of the outputs from DCX should be fed into a parallel pair of 110R and these feed into the pair of paralleled 110ohm terminated cables. That way each half becomes 110+110 =220 and when paralleled show 110r back to source.
But I cannot see whether this correctly terminates the source ends of the paralleled cables. I think this is 110r + 110r//110r = 165r, no good. There must be some padding down to ground in there so that source termination comes out at 110r.
Help.
BTW, a single wire or narrow trace spaced away from a ground plane exhibits a characteristic impedance of 110ohms. This spacing could be the thickness of a PCB or the insulation of the wire. I wonder if this could be the origin or 110ohm impedance? It was easy to achieve with conventional wiring/traces with the correct wire size/trace width and insulation/PCB thickness.
This is not a solution, but tries to cover the problem and go part way to a solution.
I'm not up on this aerial stuff, but that ref to 55r should be 110r.
I think each of the outputs from DCX should be fed into a parallel pair of 110R and these feed into the pair of paralleled 110ohm terminated cables. That way each half becomes 110+110 =220 and when paralleled show 110r back to source.
But I cannot see whether this correctly terminates the source ends of the paralleled cables. I think this is 110r + 110r//110r = 165r, no good. There must be some padding down to ground in there so that source termination comes out at 110r.
Help.
BTW, a single wire or narrow trace spaced away from a ground plane exhibits a characteristic impedance of 110ohms. This spacing could be the thickness of a PCB or the insulation of the wire. I wonder if this could be the origin or 110ohm impedance? It was easy to achieve with conventional wiring/traces with the correct wire size/trace width and insulation/PCB thickness.
AndrewT said:
Andy,
Indeed the 55 ohms is not optimal. Each DCX wants to 'see' 110 ohms, right? So lets say we have one resistor Rpad in series between each input of two DCX'es to a common source with 110 ohms source R. Then, each DCX 'sees' that Rpad, then the parallel value of the 110 source and, in parallel to that, the Rpad and 110 ohms of the other DCX. So, Rpad + (110 // (Rpad + 110)) should be 110. Then, the source 'sees' two parallel values of each Rpad + 110 so 0.5(Rpad + 110) should also be 110. This last equation gives Rpad = 110. I don't know what the first equation gives, but probably also 110 ohms, is that what you are saying?
Edit: If I plug Rpad = 110 in the first equation, that seems to jive...
Jan Didden
110ohm two output diplexer
Hi
36r7 =110/3
make up a three legged star of 36r7 resistors.
connect the outpoard end of each resistor to the respective hot pin of the XLR sockets.
Do likewise for the cold pins of the three XLR sockets.
connect the three ground pins together.
Mount the three sockets in a little metal box.
The female socket is your input from the source.
The two male sockets are the supplies to the DCX.
The attenuation is about -3.8db at each output.
Please check my numbers.
The experts out there are all chuckling up their sleeves, but I'm learning while we entertain the crowd.
Experts,
is there an easier way?
How does one maintain the 110ohm characteristic impedance across the diplexer?
Is it trial and error spacing, or calculation, or does it not matter in the short distances involved?
Hi
36r7 =110/3
make up a three legged star of 36r7 resistors.
connect the outpoard end of each resistor to the respective hot pin of the XLR sockets.
Do likewise for the cold pins of the three XLR sockets.
connect the three ground pins together.
Mount the three sockets in a little metal box.
The female socket is your input from the source.
The two male sockets are the supplies to the DCX.
The attenuation is about -3.8db at each output.
Please check my numbers.
The experts out there are all chuckling up their sleeves, but I'm learning while we entertain the crowd.
Experts,
is there an easier way?
How does one maintain the 110ohm characteristic impedance across the diplexer?
Is it trial and error spacing, or calculation, or does it not matter in the short distances involved?
resolution question
I am using the DCX to correct the frequency response of a dipole woofer. To compensate for the 6db/oct loss of the woofer I am increasing the level in the digital domain towards lower frequencies. The woofer operates below 100 Hz. In order to prevent clipping I need to reduced the level of the entire woofer channel prior to the frequency dependent increase. Am I losing resolution by doing this ? (Since the woofer covers two + x ovtaves we are talking about 12-15 db attenuation here)
Thanks for your help
Carsten
I am using the DCX to correct the frequency response of a dipole woofer. To compensate for the 6db/oct loss of the woofer I am increasing the level in the digital domain towards lower frequencies. The woofer operates below 100 Hz. In order to prevent clipping I need to reduced the level of the entire woofer channel prior to the frequency dependent increase. Am I losing resolution by doing this ? (Since the woofer covers two + x ovtaves we are talking about 12-15 db attenuation here)
Thanks for your help
Carsten
Hi,
the sub channel will need reducing in the digital domain to remove the clipping. Does it show on the LED indicators?
You have added a high gain or high Q to achieve your desired EQ.
If you are clipping then you have used all the bits and a little more so resolution is not lost by reducing gain. But, you will have to reduce the gain BEFORE the stage that starts the clipping problem.
the sub channel will need reducing in the digital domain to remove the clipping. Does it show on the LED indicators?
You have added a high gain or high Q to achieve your desired EQ.
If you are clipping then you have used all the bits and a little more so resolution is not lost by reducing gain. But, you will have to reduce the gain BEFORE the stage that starts the clipping problem.
AndrewT said:
You are right, if I do not reduce the gain for the woofer channel the LED indicators show clipping.
I can run this setup without any clipping problems by reducing the gain frequency independent first and raising it by the same amount frequency dependent afterwards.
But if I assume that gain reduction in the digital domain is done e.g. using a bit shifting operation then increasing the gain at a later stage will not recuperate the original information. The least significant bits are gone and cannot be restored properly.
Does anybody know if that is the case and how I can check this?
Just a link I got at yahoo :http://www.tcelectronic.com/TechLibrary#Mastering
about recording levels and distortion,will read it..
about recording levels and distortion,will read it..
conclusion
I don't want to rain on anyone's parade here, but I am giving up on using the DCX2496 in my system and want to share my experience so that it may help the next person... Here is the summary:
1. I use Monitor Audio GR60 speakers with B&K 200.7 amp and Reference 30 preamp. My source is digital. When using the DCX I feed it digital. When using the passive crossover I feed digital to the Ref 30 (which incidentally also uses AK4393 DACs) whose analog out goes to the amp.
2. I was convinced that removing the passive crossover from the speakers and implementing an active crossover with the DCX2496 + output mods should improve the musical quality of the system.
3. I used the French mod kit and made some additional changes to it such as using smaller caps for the mids/highs in order to attempt to further improve sound. I can definitely say that the mods do make sonic improvements over the stock version. The cimprovements are not great but noticeable.
4. I experiemented with different crossover settings and settled on what sounds the best.
5. After all this I decided I should do a sanity check and use the old configuration witht the passive crossover. I thought it would sound much worse because my system did start to sound better gradually. It turns out that I prefer the sound with passive crossovers! It has a little more depth and presence compared to the DCX setup. GR60 speakers are known for needing long break-in times so that may have explained why things started to sound better with time when I was fiddling around with the DCX.
I have now restored my speakers to passive crossovers and am going to sell the DCX on EBay because I no longer have a need for it. In conclusion, I believe that active crossovers are great for DIY speakers because passive crossover design is difficult and there is a lot of driver interaction. If a system is well designed, however, then just replacing the passive crossover with an active one will not necessarily improve the system sonically (even if everything is done "right"). My theory is that the passive crossover and drivers are designed to work together so well that it is difficult to improve the sound by simply popping in an active crossover, no matter how good that active crossover is...
Anyway, the DCX is a nice piece of equipment and is very flexible for DIY projects, but I marginally prefer the sound of the passive crossovers of the GR60 speakers and the system is certainly simpler and more flexible (surround sound etc) without the active crossover.
Best of luck to you all...
Philip
I don't want to rain on anyone's parade here, but I am giving up on using the DCX2496 in my system and want to share my experience so that it may help the next person... Here is the summary:
1. I use Monitor Audio GR60 speakers with B&K 200.7 amp and Reference 30 preamp. My source is digital. When using the DCX I feed it digital. When using the passive crossover I feed digital to the Ref 30 (which incidentally also uses AK4393 DACs) whose analog out goes to the amp.
2. I was convinced that removing the passive crossover from the speakers and implementing an active crossover with the DCX2496 + output mods should improve the musical quality of the system.
3. I used the French mod kit and made some additional changes to it such as using smaller caps for the mids/highs in order to attempt to further improve sound. I can definitely say that the mods do make sonic improvements over the stock version. The cimprovements are not great but noticeable.
4. I experiemented with different crossover settings and settled on what sounds the best.
5. After all this I decided I should do a sanity check and use the old configuration witht the passive crossover. I thought it would sound much worse because my system did start to sound better gradually. It turns out that I prefer the sound with passive crossovers! It has a little more depth and presence compared to the DCX setup. GR60 speakers are known for needing long break-in times so that may have explained why things started to sound better with time when I was fiddling around with the DCX.
I have now restored my speakers to passive crossovers and am going to sell the DCX on EBay because I no longer have a need for it. In conclusion, I believe that active crossovers are great for DIY speakers because passive crossover design is difficult and there is a lot of driver interaction. If a system is well designed, however, then just replacing the passive crossover with an active one will not necessarily improve the system sonically (even if everything is done "right"). My theory is that the passive crossover and drivers are designed to work together so well that it is difficult to improve the sound by simply popping in an active crossover, no matter how good that active crossover is...
Anyway, the DCX is a nice piece of equipment and is very flexible for DIY projects, but I marginally prefer the sound of the passive crossovers of the GR60 speakers and the system is certainly simpler and more flexible (surround sound etc) without the active crossover.
Best of luck to you all...
Philip
Philip,
One question. Did you do any measuring of your speakers when using the DCX? How did you decide the settings for the DCX? I am asking this because the difference (and your preference) could be because the DCX configuration (xover freq, filter types, equalisers etc) may have been suboptimal, while the passives are of course tweaked for the speakers.
Jan Didden
One question. Did you do any measuring of your speakers when using the DCX? How did you decide the settings for the DCX? I am asking this because the difference (and your preference) could be because the DCX configuration (xover freq, filter types, equalisers etc) may have been suboptimal, while the passives are of course tweaked for the speakers.
Jan Didden
Jan,
I used RoomEQWizzard to analyze the frequency response (nice and free tool, BTW). The response with the DCX was just as good as with the passives. In fact, I could make it even better by applying EQ settings etc. I would say that from a frequency response perspective the DCX setup was as good or better than the passive setup.
I also played quite a bit with the crossover settings. I first started with the settings from the speaker specs and then played around by changing filter types and frequencies. I was able to do this because I kept the DCX at my listening position so I could quickly A/B changes... I am pretty confident that my crossover setup was quite good. BTW, I found that I liked the sound of the LR24's slightly more than that of the LR48's. It sounded a bit "cleaner", so my final preferred setting was LR24's. It was, however remarkable how little difference the filter type made overall...
I also played around with different caps etc. Scott Endler from http://mysite.verizon.net/vze4c5pt/id14.html (the one who started all these mods) suggested different caps etc so I went ahead and tried them. They did seem to make a slight difference for the better.
Overall, through, I would say that the DCX sound is very close to the passive sound. I prefer the passive sound because it sounds slightly more lively and engaging. Some people might prefer the DCX sound. The DCX sound unfortunately did not improve the sound stage in any way. I even increased trebble on the DCX slightly in order to match the frequency response of the passives. Anyway, in the end I decided that all this is just not worth it because I certainly cannot say that the DCX setup sounds better. Since the DCX complicates the system dramatically, I don't think it's worth it. For DIY speakers, on the other hand, it would be the only way to go, I think... For mid to high-end speakers designed with passive crossovers, I think the DCX reconfiguration is certainly not worth the effort. I wish I knew this ahead of time. I spent endless hours on this project...
Regards
Philip
I used RoomEQWizzard to analyze the frequency response (nice and free tool, BTW). The response with the DCX was just as good as with the passives. In fact, I could make it even better by applying EQ settings etc. I would say that from a frequency response perspective the DCX setup was as good or better than the passive setup.
I also played quite a bit with the crossover settings. I first started with the settings from the speaker specs and then played around by changing filter types and frequencies. I was able to do this because I kept the DCX at my listening position so I could quickly A/B changes... I am pretty confident that my crossover setup was quite good. BTW, I found that I liked the sound of the LR24's slightly more than that of the LR48's. It sounded a bit "cleaner", so my final preferred setting was LR24's. It was, however remarkable how little difference the filter type made overall...
I also played around with different caps etc. Scott Endler from http://mysite.verizon.net/vze4c5pt/id14.html (the one who started all these mods) suggested different caps etc so I went ahead and tried them. They did seem to make a slight difference for the better.
Overall, through, I would say that the DCX sound is very close to the passive sound. I prefer the passive sound because it sounds slightly more lively and engaging. Some people might prefer the DCX sound. The DCX sound unfortunately did not improve the sound stage in any way. I even increased trebble on the DCX slightly in order to match the frequency response of the passives. Anyway, in the end I decided that all this is just not worth it because I certainly cannot say that the DCX setup sounds better. Since the DCX complicates the system dramatically, I don't think it's worth it. For DIY speakers, on the other hand, it would be the only way to go, I think... For mid to high-end speakers designed with passive crossovers, I think the DCX reconfiguration is certainly not worth the effort. I wish I knew this ahead of time. I spent endless hours on this project...
Regards
Philip
Another thing to note is that the DCX is quite noisy. The power supply is very bad at filterting out AC line noise. If I plug my laptop in next to the DCX and play no music, I can hear quite a bit of digital noise artifacts from the tweeter. My B&K pre-amp does not have this problem. No matter what I plug in next to the pre-amp I get the same minimal hiss.
This was not a big issue for me since the noise it barely audible from the listening position and not at all audible when music is playing...
Philip
This was not a big issue for me since the noise it barely audible from the listening position and not at all audible when music is playing...
Philip
audio-kraut said:
Philip was saying that this happens when he *PLUGS* his laptop into the DCX. I don't think it is the fault of the DCX, but the process of plugging the laptop into it, creating an entry point for the noise generated in the laptop to enter the DCX.
My DCX is also silent and I have it connected to some pretty sensitive pro compression drivers...