Sorry, I didn't realize that the link for the LT-SPICE software doesn't work. Here this one should be OK:
http://www.linear.com/company/software.jsp
http://www.linear.com/designtools/softwareRegistration.jsp
http://www.linear.com/company/software.jsp
http://www.linear.com/designtools/softwareRegistration.jsp
KBK said:
Contact me by mail or by the website, Thank you.
giulio said:
Hi Nicholas and all,
I have done some more work on the board based on the advice of people on this forum plus a useful PM.
Nicholas, basically you would have to terminate with a >10K resistor to ground both the +ve and -ve output. There is no way you can fit 12 Polyprops coupling caps in there, though. So if I were you I would stick with single ended (but I know that Frank will disagree).
Two points:
1) those mundorf caps are good! I get of course more transparency w/out them, but the difference is not that big.
2) The card as it is connects pin 3 on the XLR connectors to ground and lets pin 1 floating. That is irrelevant when run single ended as one would brige -ve and ground on the RCA side, but is a major pain for users who want to turn the card into balanced. Nicholas, if you are doing a second run I suggest you connect pin 1 to ground. This would halve the number of connections that a modder has to establish. Especially given that those ground tracks to the output pins are very fragile!
Giulio
I will email you to ask you what modifications you used
clarification
clarification
hi
I think I understand the benfits of low jitter at the point of DA conversion at the output or for AD conversion at the input of any digital audio device.
But can anybody explain if there is any need or sonic improvmend for low jitter at the SPDIF input.
I am asking because I would like to do some clock mod's and because of the mod Oehlrich has suggested here with his improved digital input stage. This seems to be good not only for having higher input sampelrate capability but also for having a very low jitter performance.
http://freerider.dyndns.org/anlage/Behringer-Input-Stage-E.htm
"http://freerider.dyndns.org/anlage/Behringer-Input-Stage-E.htm"
Greetings
Michael
clarification
hi
I think I understand the benfits of low jitter at the point of DA conversion at the output or for AD conversion at the input of any digital audio device.
But can anybody explain if there is any need or sonic improvmend for low jitter at the SPDIF input.
I am asking because I would like to do some clock mod's and because of the mod Oehlrich has suggested here with his improved digital input stage. This seems to be good not only for having higher input sampelrate capability but also for having a very low jitter performance.
http://freerider.dyndns.org/anlage/Behringer-Input-Stage-E.htm
"http://freerider.dyndns.org/anlage/Behringer-Input-Stage-E.htm"
Greetings
Michael
Hi Janneman,
However, I went back to post 844 to check the context of what I had said.
Extracting that sentence from the context of the post, the preceeding discussion and schematic completely changes the meaning of what I was referring to.
Yes, I could have worded it more carefully to ensure confusion did not exist in the minds of any reader.
BUT, I stand by the message I had intended. The schematic shows the pot and downstream unbalanced and upstream balanced. The pot uses one end as reference and wiper as tapping point. With the phase inversion across the whole pot and the wiper at midpoint then the down stream does NOT see zero signal. There is NOT cancellation of the balanced signal at midpoint.
Explain where I have gone wrong.
I do wish you had come back sooner, as I fear a number of readers may have been misled if I am wrong.
this is so fundamental I could not believe I had said the opposite.
However, I went back to post 844 to check the context of what I had said.
Extracting that sentence from the context of the post, the preceeding discussion and schematic completely changes the meaning of what I was referring to.
Yes, I could have worded it more carefully to ensure confusion did not exist in the minds of any reader.
BUT, I stand by the message I had intended. The schematic shows the pot and downstream unbalanced and upstream balanced. The pot uses one end as reference and wiper as tapping point. With the phase inversion across the whole pot and the wiper at midpoint then the down stream does NOT see zero signal. There is NOT cancellation of the balanced signal at midpoint.
Explain where I have gone wrong.
I do wish you had come back sooner, as I fear a number of readers may have been misled if I am wrong.
AndrewT said:
I knew you knew, and that there was some kind of confusion.
Sorry I didn't notice it earlier.
My point was that in the lower half output part the opamp ALWAYS gets the max neg phase from the DAC output, while ONLY the upper part opamp gets the pot output fed. And that pot output varies between max upper phase through 0 to max lower phase. So, at the output you get some kind of a varying signal between max sum and zero, but is a) totally out of whack with the expected pot setting and b) not have the advantages of the balanced signal wrt supressing cm non-linearity and noise.
Jan Didden
Hi Janneman,
we are looking at the same schematic.
Now explain.
The differantial amp measures the voltage across it's inputs and amplifies it.
One measuring point us at the lower end of the pot.
The other measuring end is on the wiper.
The voltage difference between the two points is amplified. BUT it is not balanced, the lower measuring point is offset from the ground reference. In fact the lower end has become the reference.
The voltage at the top referred to ground is +V and the voltage at the bottom referred to ground is -V but when the reference is moved from ground to the lower end, their voltages become 0 and 2V and halfway up becomes V. NOT 0V.
Now, approaching from a different direction.
If the source side were referenced to the SAME ground as the amplifying side then all is different. It closer matches what you are reading into the schematic.
But in a balanced system one must assume the two signal poles are NOT referenced to ground. That's what makes balanced so good at rejecting spurious signals. Just the wanted differential signal is passed on without reference to the source ground. The schematic as posted shows NO GROUND REFERENCE. But the amplifying end does show a ground reference because it is an electronically balanced input and not a true balanced and galvanically isolated input.
we are looking at the same schematic.
Now explain.
why this should apply?
The differantial amp measures the voltage across it's inputs and amplifies it.
One measuring point us at the lower end of the pot.
The other measuring end is on the wiper.
The voltage difference between the two points is amplified. BUT it is not balanced, the lower measuring point is offset from the ground reference. In fact the lower end has become the reference.
The voltage at the top referred to ground is +V and the voltage at the bottom referred to ground is -V but when the reference is moved from ground to the lower end, their voltages become 0 and 2V and halfway up becomes V. NOT 0V.
Now, approaching from a different direction.
If the source side were referenced to the SAME ground as the amplifying side then all is different. It closer matches what you are reading into the schematic.
But in a balanced system one must assume the two signal poles are NOT referenced to ground. That's what makes balanced so good at rejecting spurious signals. Just the wanted differential signal is passed on without reference to the source ground. The schematic as posted shows NO GROUND REFERENCE. But the amplifying end does show a ground reference because it is an electronically balanced input and not a true balanced and galvanically isolated input.
Hi,
During assembly I realized that there is a lot of tiny SMD work to be done on the UcDs. So I decided to shift some parts of the filter back behind the DAC. There are now only two 0805 caps to be mounted on two 0805 resistors on the UcDs.
Can't complain about sound quality. The only problem left is the pot. At the low end it ends at about 4 ohm. That's enough that you still hear some music. Best would be the shunted relay attenuator KBK (post 898) dreamed of.
Six relays with 6 contacts each (most probably not available?) a PIC and some resistors could be sufficient. Has somebody an idea how to get dreams to become true?
See attached file (change appendix to .asc and use http://www.linear.com/designtools/softwareRegistration.jsp)
Kind Regards,
Frank
PS: The former (post 826) circuit (different pot) you are just talking about might cause some CMRR problem at the instrumental amp.
During assembly I realized that there is a lot of tiny SMD work to be done on the UcDs. So I decided to shift some parts of the filter back behind the DAC. There are now only two 0805 caps to be mounted on two 0805 resistors on the UcDs.
Can't complain about sound quality. The only problem left is the pot. At the low end it ends at about 4 ohm. That's enough that you still hear some music. Best would be the shunted relay attenuator KBK (post 898) dreamed of.
Six relays with 6 contacts each (most probably not available?) a PIC and some resistors could be sufficient. Has somebody an idea how to get dreams to become true?
See attached file (change appendix to .asc and use http://www.linear.com/designtools/softwareRegistration.jsp)
Kind Regards,
Frank
PS: The former (post 826) circuit (different pot) you are just talking about might cause some CMRR problem at the instrumental amp.
Hi,
I read your last posts based on the circuit shown in post 826 more carefully. For my understanding input and output are galvanically isolated due to the two AC coupling caps, as long as you do not connect grounds.
But in real life there is a common main and therefore you might get some noise from the PSUs. So connecting the grounds via the shield of the cable would result in a signal (seen by the opamps) which is identical to the negative output of the DAC (pot setting 0). The difference between the two inputs of the opamps (pot setting 0) is 0 which should result in a 0 output of the opamps. But unfortunately the CMRR of an opamp is not infinite.
Between the low end of the pot and the wiper there is always a 180° phase inverted signal independent of pot setting.
Hi miqe0, (post 905),
I wouldn't worry so much about the CS8420 (digital audio sample rate converter) as long as you do not use your phone cabling. Much more important is the signal quality. So if you have an optical output use a toslink cable and a converter on the DCX side (5 bugs ebay). You can mod the output of the converter (Serial: 74HCxxx + 100n cap + 150 ohm resistor) to improve impedance.
Regards,
Frank
I read your last posts based on the circuit shown in post 826 more carefully. For my understanding input and output are galvanically isolated due to the two AC coupling caps, as long as you do not connect grounds.
But in real life there is a common main and therefore you might get some noise from the PSUs. So connecting the grounds via the shield of the cable would result in a signal (seen by the opamps) which is identical to the negative output of the DAC (pot setting 0). The difference between the two inputs of the opamps (pot setting 0) is 0 which should result in a 0 output of the opamps. But unfortunately the CMRR of an opamp is not infinite.
Between the low end of the pot and the wiper there is always a 180° phase inverted signal independent of pot setting.
Hi miqe0, (post 905),
I wouldn't worry so much about the CS8420 (digital audio sample rate converter) as long as you do not use your phone cabling. Much more important is the signal quality. So if you have an optical output use a toslink cable and a converter on the DCX side (5 bugs ebay). You can mod the output of the converter (Serial: 74HCxxx + 100n cap + 150 ohm resistor) to improve impedance.
Regards,
Frank
Hi Michael,
What I'm talking about is a toslink to cinch converter. E.g.:
http://cgi.ebay.de/Audio-Wandler-To...ryZ36673QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
There might be cheaper ones. Looking at the poor analog input circuit of the original DCX (3 opamps and 2 china caps in a row) I can't believe analog sound quality should be better.
Regards,
Frank
What I'm talking about is a toslink to cinch converter. E.g.:
http://cgi.ebay.de/Audio-Wandler-To...ryZ36673QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
There might be cheaper ones. Looking at the poor analog input circuit of the original DCX (3 opamps and 2 china caps in a row) I can't believe analog sound quality should be better.
Regards,
Frank
Hi
Well me too - but judging from what I hear the original SPDIF input simply sounds worse to me.
It would be fine to know if anybody has similar or opposite experiances. Tough I have to admit that my analog input has different blocking caps and the (original) opamps have bigger bypassing caps - what results in a slight improvement, but not THAT much.
Maybe I should rather go for a HighEnd cabel or for a 110/75 converter than for a SPDIF mod?
greetings
Michael
Well me too - but judging from what I hear the original SPDIF input simply sounds worse to me.
It would be fine to know if anybody has similar or opposite experiances. Tough I have to admit that my analog input has different blocking caps and the (original) opamps have bigger bypassing caps - what results in a slight improvement, but not THAT much.
Maybe I should rather go for a HighEnd cabel or for a 110/75 converter than for a SPDIF mod?
greetings
Michael
music store and bheringer
I worked at a music store last year this time (guitar center) and we had a wide variety of audio devices. Behringer is way on the bottom of the list. dont expect much from it, made in china, euro styling. it works, but you'd be better off with an old analog unit. cheap digital sounds like cheap digital.
I'll be interested if you can make it sound right.
pesky
I worked at a music store last year this time (guitar center) and we had a wide variety of audio devices. Behringer is way on the bottom of the list. dont expect much from it, made in china, euro styling. it works, but you'd be better off with an old analog unit. cheap digital sounds like cheap digital.
I'll be interested if you can make it sound right.
pesky
pesky said:
Well, that's why the DCX got our attention. It's low price is hard to beat - even for a cheap analog xo
pesky said:
But we do! That's our DIY way of thinking
pesky said:
Sorry, that is not true for our application. We would be swamped with op-amp's and caps just to simulate a fraction of what the DCX's DSP has to offer.
pesky said:
This is true most of the time. But many users are surprised how god the stock Behringer DCX2496 actually sounds. That is one reason many of us believe it is possible to create something beautiful with this device.
Another reason is that if you could open the lid on many of the high end stuff you where selling last summer you would be surprised to see that most of the chips responsible for the audio part in those expensive brands are the same Behringer is mounting in they're creap'o China products
pesky said:
There is 'only' 27 pages so far in this thread. New readers interested in the DCX should read though them before they make up they're mind regarding the quality of this device. Many have come to the conclusion that the DCX doesn't colour the sound in they're setup and they therefore use it without any modification.
Ryssen said:All I can say is I have the DEQ2496 before the DCX2496,and I bypassed the DCX taking the signal analog from DEQ to my headphoneamp with HD650,man was that a difference!There must be sommething wrong with the DCX2496!
Ryssen, what are you comparing here ?
What is the DCX configurations when comparing the DEQ analog out with the DCX? Could it be some EQ setting you have forgot to disable on the DCX or something similar?
Is the DEQ flat or bypassed (all modules) when the headphones are used?