MJK said:
When this Behringer hit the forums it cost $350 US, plus
shipping and people liked it.
Today you can find it for $250 shipped! If $350 + shipping
was good 2 years ago, then $250 shipped should be sweeter right?
The gamble is low risk. If you don't like it, sell it and recoup
most of your money back.
You need to look at the features and see if you can take
advantage of them for your system. Does the analog crossover
do what you need? Do you need more stuff that the digital one
offers?
If you DIY many loudspeaker projects the Behringer also makes
for a good development tool even if you don't want to use it
on your permanent installation.
Has anyone received improved Excel spread sheet for LeCleach callculation? If you guys did could you email me. I am just planning to try that set up.
Also to add what Thylantir started. No question Behringer is one of the best buys I made. Versatility and quality outperforms investment. I used Merchand crossover and nothing is wrong with that crossover. But Behringer gives you much more options, and I wouldn't think twice about it. As long as you have means of attenuating analog signal at the output of Behringer it is better and similarly priced. Just think of it as good oubord D/A converter, crossover, EQ, preamp all in one.
I am sorry if you got wrong impression in my previous post. I would certainly recommend this unit. Modding output portion is pushing to get more transparency, which is not always audible in every system.
Do not worry about modding. Listen in your system first as is, and later decide if you need to change anything. But... Any serious modding will cost you more than unit itself, because 6 ballanced channels is equal of 12 mono unballanced channels. So, almost anything has to be done x12.
My biggest problem was getting attenuation for six ballanced channels. After that I succeded with my mod simply because I eliminated whole active output circuitry and replaced it with pasive element. More or less whenever the same thing is done in any unit (obviously if possible)
it will result with more transparency.
Also to add what Thylantir started. No question Behringer is one of the best buys I made. Versatility and quality outperforms investment. I used Merchand crossover and nothing is wrong with that crossover. But Behringer gives you much more options, and I wouldn't think twice about it. As long as you have means of attenuating analog signal at the output of Behringer it is better and similarly priced. Just think of it as good oubord D/A converter, crossover, EQ, preamp all in one.
I am sorry if you got wrong impression in my previous post. I would certainly recommend this unit. Modding output portion is pushing to get more transparency, which is not always audible in every system.
Do not worry about modding. Listen in your system first as is, and later decide if you need to change anything. But... Any serious modding will cost you more than unit itself, because 6 ballanced channels is equal of 12 mono unballanced channels. So, almost anything has to be done x12.
My biggest problem was getting attenuation for six ballanced channels. After that I succeded with my mod simply because I eliminated whole active output circuitry and replaced it with pasive element. More or less whenever the same thing is done in any unit (obviously if possible)
it will result with more transparency.
Speadsheet for LeCleach set-up
Regarding the obscurities during DCX setup I finally added an EXCEL-Spreadsheet for a step by step set-up in the ZIP file presented to download. The spreadsheet calculates all the values for the DCX2496 according vour x-over frequencies and the delays measured by auto-setup.
http://freerider.dyndns.org/anlage/LeCleach1.zip
(Server is only up from 8AM to 10PM German time)
I hope this will help you and all the others setting up this genuine LeCleach approach.
AR2 said:
Regarding the obscurities during DCX setup I finally added an EXCEL-Spreadsheet for a step by step set-up in the ZIP file presented to download. The spreadsheet calculates all the values for the DCX2496 according vour x-over frequencies and the delays measured by auto-setup.
http://freerider.dyndns.org/anlage/LeCleach1.zip
(Server is only up from 8AM to 10PM German time)
I hope this will help you and all the others setting up this genuine LeCleach approach.
Today's stupid question:
I've looked,and looked,and looked..and poured over boards and tech manuals. I've come to the conclusion that wasting less time by asking is the way to go.
When the 2496 is accepting data from a digital source, like a cd player, is is using the CS8240 at all, and if it is..is is using any of the onboard crystals?
When the unit is getting purely analog signals, it is obviously utilizing the CS8420...but WHICH crystal is it utilizing? I want to do some clock work, but It's been along time since I poured over manuals and the like. I just want to fix the clock, not back engineer the thing. So which clock is it? the 24meg or the 30.0meg clock?
I feel rude, like I'm cheating or something, but I'm also wasting time, by trying to chase it down on my own. I also have poured over this thread..and not found that info.
Thanks!
PS. the clock implementation from what I have found in casual investigation..the guy who engineered the clocking on this unit should have been tied to a chair, and *****-slapped into last week.
I've looked,and looked,and looked..and poured over boards and tech manuals. I've come to the conclusion that wasting less time by asking is the way to go.
When the 2496 is accepting data from a digital source, like a cd player, is is using the CS8240 at all, and if it is..is is using any of the onboard crystals?
When the unit is getting purely analog signals, it is obviously utilizing the CS8420...but WHICH crystal is it utilizing? I want to do some clock work, but It's been along time since I poured over manuals and the like. I just want to fix the clock, not back engineer the thing. So which clock is it? the 24meg or the 30.0meg clock?
I feel rude, like I'm cheating or something, but I'm also wasting time, by trying to chase it down on my own. I also have poured over this thread..and not found that info.
Thanks!
PS. the clock implementation from what I have found in casual investigation..the guy who engineered the clocking on this unit should have been tied to a chair, and *****-slapped into last week.
Short answer is 24M.
You can find my findings http://www.esken.net/ergo/dcx2496_clock.html
Regards,
Ergo
You can find my findings http://www.esken.net/ergo/dcx2496_clock.html
Regards,
Ergo
Hello Ergo,
I do not know if you are aware but something is wierd regarding your web provider or web page. When trying to connect to your site it switching it to some weird search web page which is www.net.net.
In another word by trying to connect to your page my web browser is routed to that search page. Maybe is something that you are not aware, or maybe is problem on my side, but I better let you know.
I do not know if you are aware but something is wierd regarding your web provider or web page. When trying to connect to your site it switching it to some weird search web page which is www.net.net.
In another word by trying to connect to your page my web browser is routed to that search page. Maybe is something that you are not aware, or maybe is problem on my side, but I better let you know.
While I am here, I do not know if anyone tried new Excel doc. modified by Oehlrich. I did and it is working great. After I did adjustments specified in doc. I got great improvement in imaging and blending of drivers to each other.
Oehlrich, one more time, thank you for sharing this great concept with us.
Oehlrich, one more time, thank you for sharing this great concept with us.
Thanks for the reply. I looked over the web info via the link you provided. As a special 24m clock, is it that the clock is already divided to the 1/2 frequency, or the 'special clock' (XO2 2496 version) has both frequencies on the output?
looking closer at the image of the XO in your 'complete photo' in the pdf (the completed, first image)..that appears to be the case.
It looks as if that..if I cure the issues with the master, I'll still have to deal with the injected jitter of the inverter/divider chip. I'll have to either upgrade that chip and or work with the PS issues on that existing chip...OR ... buy the special xo..and maybe work on that one.
looking closer at the image of the XO in your 'complete photo' in the pdf (the completed, first image)..that appears to be the case.
It looks as if that..if I cure the issues with the master, I'll still have to deal with the injected jitter of the inverter/divider chip. I'll have to either upgrade that chip and or work with the PS issues on that existing chip...OR ... buy the special xo..and maybe work on that one.
LeCleach Measurements
Today I did some measurements on the LeCeach's aproach regarding phase distortions.
The result was that phase distortion is strongly suppressed by the LeCeach approach. Its method is a very good trade off between linear frequency and linear phase.
Hearing a difference between a normal setup and the modified LeCleach setup also means that phase distortion IS audible (a lot of people say it is NOT!).
Some comments regaring the results can be found here:
http://freerider.dyndns.org/anlage/LeCleach.htm
The images from the oscilloscope here:
http://freerider.dyndns.org/anlage/images-LeCleach-Measurements/index.html
(Server is up 8AM to 10PM German time)
Today I did some measurements on the LeCeach's aproach regarding phase distortions.
The result was that phase distortion is strongly suppressed by the LeCeach approach. Its method is a very good trade off between linear frequency and linear phase.
Hearing a difference between a normal setup and the modified LeCleach setup also means that phase distortion IS audible (a lot of people say it is NOT!).
Some comments regaring the results can be found here:
http://freerider.dyndns.org/anlage/LeCleach.htm
The images from the oscilloscope here:
http://freerider.dyndns.org/anlage/images-LeCleach-Measurements/index.html
(Server is up 8AM to 10PM German time)
I found a pro-level isolation box on flea-bay, that has 8 'Jensen' grade (5-10hz to 85-90khz) transformers in it. I'm pulling the entire input section OUT of the crossover, and putting the transformers in there instead, with a 2:1 gain in the wiring of the trannies.
I'm looking for the Schematic for this puppy (the DCX2496). Anyone want to send it to me? I'd very much appreciate it. I can figure it out for myself..but... I get tired of that after a while..
I'm looking for the Schematic for this puppy (the DCX2496). Anyone want to send it to me? I'd very much appreciate it. I can figure it out for myself..but... I get tired of that after a while..
I/O transformers for DCX
hi KBK,
I found a pro-level isolation box on flea-bay, that has 8 'Jensen' grade (5-10hz to 85-90khz) transformers in it. I'm pulling the entire input section OUT of the crossover, and putting the transformers in there instead, with a 2:1 gain in the wiring
I did exactly the same, I removed the I/O board and installed 8 transformers, huge improvement.
I'm looking for the Schematic for this puppy (the DCX2496). Anyone want to send it to me? I'd very much appreciate it.
As already mentioned, the schematics can be found in de yahoo group, there's also a picture of my modified dcx.
regards, Peter
hi KBK,
I found a pro-level isolation box on flea-bay, that has 8 'Jensen' grade (5-10hz to 85-90khz) transformers in it. I'm pulling the entire input section OUT of the crossover, and putting the transformers in there instead, with a 2:1 gain in the wiring
I did exactly the same, I removed the I/O board and installed 8 transformers, huge improvement.
I'm looking for the Schematic for this puppy (the DCX2496). Anyone want to send it to me? I'd very much appreciate it.
As already mentioned, the schematics can be found in de yahoo group, there's also a picture of my modified dcx.
regards, Peter
A somewhat different approach to crossovers using the DCX2496.....
In my two way (soon to be three way) horn system I have been running speaker level elliptic crossovers (very high slope) to greatly reduce driver interaction. The crossovers I am running are the ES600t from:
http://www.alkeng.com/es_xo.html
I wanted to try bi/tri-amping (makes it easier to try different horn/drivers) but I wanted to at least duplicate the slope of the ES600t. The DEQX can do up to 300 db/octave crossovers but since I needed this for L/C/Rs that was way too much money for two of the boxes.
I realized I could simulate elliptic crossovers on the DCX2496 to get much sharper then 48 db/octave it has as a max crossover slope.
This is the line level sum of the high/low pass. A touch of ripple at the crossover point but very minor. Vertical divisions are 10dB, horizontal are 200hz.
This is the high pass:
This is the low pass:
The hash at the bottom of the response is just the noise floor of the FFT getting into the signal, it has around a 70-80db SNR.
Crossover point is 600hz in this setup. In the first 50hz out the drivers are down about 15dB relative to each other and down over 30dB 100hz from the crossover point.
Shawn
In my two way (soon to be three way) horn system I have been running speaker level elliptic crossovers (very high slope) to greatly reduce driver interaction. The crossovers I am running are the ES600t from:
http://www.alkeng.com/es_xo.html
I wanted to try bi/tri-amping (makes it easier to try different horn/drivers) but I wanted to at least duplicate the slope of the ES600t. The DEQX can do up to 300 db/octave crossovers but since I needed this for L/C/Rs that was way too much money for two of the boxes.
I realized I could simulate elliptic crossovers on the DCX2496 to get much sharper then 48 db/octave it has as a max crossover slope.
This is the line level sum of the high/low pass. A touch of ripple at the crossover point but very minor. Vertical divisions are 10dB, horizontal are 200hz.
An externally hosted image should be here but it was not working when we last tested it.
This is the high pass:
An externally hosted image should be here but it was not working when we last tested it.
This is the low pass:
An externally hosted image should be here but it was not working when we last tested it.
The hash at the bottom of the response is just the noise floor of the FFT getting into the signal, it has around a 70-80db SNR.
Crossover point is 600hz in this setup. In the first 50hz out the drivers are down about 15dB relative to each other and down over 30dB 100hz from the crossover point.
Shawn
Ron,
"I'm curious about your crossover slopes, currently I'm using the 48dB/octave. Are you combining a 24dB and a notch filter? Can you give me the particulars?"
You are on the right track.
I don't have the exact settings in front of me (I'll post them tonight if anyone is interested) but that response is a composite of a 48dB/octave (the rolloff after the arc) and three parametric filters.
In a nutshell the 48dB/octave crossover is set a little ways away from the intended crossover point.
Then two parametric notches are set closer to the intended crossover point with Q10 and -15dB. Both are set to the same settings. I needed two here since the DCX can only do cuts of max -15dB and I wanted the initial attenuation higher then that. Then there is a third parametric right about at the crossover point boosting response there to keep it flat to the crossover point and to make the initial slope of the crossover sharper... I was a little limited by the Behringer only allowing a Q of 10 and not sharper.
The setting of each filter was done plugged into a real time dual channel FFT in transfer mode so I could setup the DCX and get immediate feedback on how it effected the filters response and the blend between the two. Without having the feedback from the FFT my initial settings were a mess.
It is possible to get even more attenuation by adding more notch filters (two more) such as:
But the initial attenuation of the crossover is much shallower (more blend region between drivers) and with the arc top coming back up 30dB it seemed kind of a pointless to try for that much initial attenuation. It also took more DSP resources from the DCX of course too.
Shawn
"I'm curious about your crossover slopes, currently I'm using the 48dB/octave. Are you combining a 24dB and a notch filter? Can you give me the particulars?"
You are on the right track.
I don't have the exact settings in front of me (I'll post them tonight if anyone is interested) but that response is a composite of a 48dB/octave (the rolloff after the arc) and three parametric filters.
In a nutshell the 48dB/octave crossover is set a little ways away from the intended crossover point.
Then two parametric notches are set closer to the intended crossover point with Q10 and -15dB. Both are set to the same settings. I needed two here since the DCX can only do cuts of max -15dB and I wanted the initial attenuation higher then that. Then there is a third parametric right about at the crossover point boosting response there to keep it flat to the crossover point and to make the initial slope of the crossover sharper... I was a little limited by the Behringer only allowing a Q of 10 and not sharper.
The setting of each filter was done plugged into a real time dual channel FFT in transfer mode so I could setup the DCX and get immediate feedback on how it effected the filters response and the blend between the two. Without having the feedback from the FFT my initial settings were a mess.
It is possible to get even more attenuation by adding more notch filters (two more) such as:
An externally hosted image should be here but it was not working when we last tested it.
But the initial attenuation of the crossover is much shallower (more blend region between drivers) and with the arc top coming back up 30dB it seemed kind of a pointless to try for that much initial attenuation. It also took more DSP resources from the DCX of course too.
Shawn