Cost 30 cents
The worst that can happen is that I will lose lock or it will sound worse. At which point, I will have only wasted 30 cents worth of resistors. It may also improve my 110R to 75R XLR to RCA conversion from the AES/EBU output of my DEQ2496 to the RCA input of the Big Dac Board.
The worst that can happen is that I will lose lock or it will sound worse. At which point, I will have only wasted 30 cents worth of resistors. It may also improve my 110R to 75R XLR to RCA conversion from the AES/EBU output of my DEQ2496 to the RCA input of the Big Dac Board.
Attachments
110-75 ohm converting cable with attenuators
My new XLR to RCA converting digital cable with built in attenuators is playing. Neutrik silver XLR, Belden 1701a wire, Cardas rhodium RCA. Lock from the XLR digital out of my DEQ2496 into the Big Dac Board configured with DIR9001/ CS43122 running direct out seems to be no problem. 7db attenuation in the XLR input and 6db built into the RCA output gave me the resistor values shown below which I had on hand. Sonic comparisons with my old, otherwise identical cable to follow after some hours of break in tomorrow.
My new XLR to RCA converting digital cable with built in attenuators is playing. Neutrik silver XLR, Belden 1701a wire, Cardas rhodium RCA. Lock from the XLR digital out of my DEQ2496 into the Big Dac Board configured with DIR9001/ CS43122 running direct out seems to be no problem. 7db attenuation in the XLR input and 6db built into the RCA output gave me the resistor values shown below which I had on hand. Sonic comparisons with my old, otherwise identical cable to follow after some hours of break in tomorrow.
Attachments
Digital cables with T pads sound better!
.
It is surprising how much difference adding the attenuators to both ends of the cable makes to the sound even when running into the DIR9001 receiver which has it's own clock. The differences are also very apparent into the same dac when the DIR9001 receiver module is changed to CS8416. My application might be more revealing due to the attenuator on the input of the cable also improving the adaptation from the XLR standard 110R to an RCA 75R output. In this application, adding the attenuator to convert the input/ output impedance is a no brainer compared to a $30 transformer which will save about... $30. I will also do listening tests with an RCA/ RCA cable from my FireWire410 to the SRC2496 upsampler and with an XLR/ XLR digital cable from the SRC to the DEQ2496.
.
The digital cable with T pads on both ends sounds more natural and layered. On Natalie Merchant live San Andreas Fault, the vocal and instrument images float free in a blacker background. Using the standard cable makes the crowd noise sound less realistic and more like white noise. Janis Ian Live, Tatoo, has more body to the vocals and the eerie steel guitar set way back in the soundstage is more fleshed out. The standard cable is flatter and less 3d, less composed. Her live Acousticville shows the only hint of a trade off sounding slightly more dynamic and exciting with the standard cable. None of the other 10+ recordings I played had anything superior to show about the standard cable. The padded cable generally has more ease to the presentation that analog devotees will appreciate. The intro to No God from the Exteme, Waiting For the Punch Line disc clearly favors the new cable as it gives much more body to the drums and bass guitar. On track 7 from Ian Anderson's Secret Language Of Birds, the old cable makes everything sound very artificial by comparison.
.
Adding pads to your digital cable is a clear winner as long as the signal stays strong enough to provide a reliable lock. It will be interesting to hear if the effect can sound even better with the excellent Susumu surface mount resistors replacing the Yageo through hole parts I am using now.
Any recommendations to tweak my values before I order the parts?
OK. I sat and read every post in this thread. 200+ posts, countless hours of theoretical conjecture, and not one person other than the original poster and his compadre doing the excellent scope shots ever actually built and listened to the simplest and cheapest circuit ever proposed on DIYAudio. Sad. You all could have saved a lot of time. It sounds better! I will be rebuilding all three of the digital cables in my signal path.
.
It is surprising how much difference adding the attenuators to both ends of the cable makes to the sound even when running into the DIR9001 receiver which has it's own clock. The differences are also very apparent into the same dac when the DIR9001 receiver module is changed to CS8416. My application might be more revealing due to the attenuator on the input of the cable also improving the adaptation from the XLR standard 110R to an RCA 75R output. In this application, adding the attenuator to convert the input/ output impedance is a no brainer compared to a $30 transformer which will save about... $30. I will also do listening tests with an RCA/ RCA cable from my FireWire410 to the SRC2496 upsampler and with an XLR/ XLR digital cable from the SRC to the DEQ2496.
.
The digital cable with T pads on both ends sounds more natural and layered. On Natalie Merchant live San Andreas Fault, the vocal and instrument images float free in a blacker background. Using the standard cable makes the crowd noise sound less realistic and more like white noise. Janis Ian Live, Tatoo, has more body to the vocals and the eerie steel guitar set way back in the soundstage is more fleshed out. The standard cable is flatter and less 3d, less composed. Her live Acousticville shows the only hint of a trade off sounding slightly more dynamic and exciting with the standard cable. None of the other 10+ recordings I played had anything superior to show about the standard cable. The padded cable generally has more ease to the presentation that analog devotees will appreciate. The intro to No God from the Exteme, Waiting For the Punch Line disc clearly favors the new cable as it gives much more body to the drums and bass guitar. On track 7 from Ian Anderson's Secret Language Of Birds, the old cable makes everything sound very artificial by comparison.
.
Adding pads to your digital cable is a clear winner as long as the signal stays strong enough to provide a reliable lock. It will be interesting to hear if the effect can sound even better with the excellent Susumu surface mount resistors replacing the Yageo through hole parts I am using now.
Any recommendations to tweak my values before I order the parts?
Attachments
Connector
Should I match the inherent impedance of the RCA connector or the 75R termination behind it?
God help you reading through all this thread - it must have been agony 
You are the first from here to actually try this out - well done - it's shameful that 20 pages were spent arguing the theory without one person trying it. Anyway, all in the past!
As you probably read, the way that the attenuators work in this situation seems to be by reducing the reflections caused by impedance mismatches encountered by the SPDIF signal along the signal path (this includes transmitter, outputs stage, connectors, cable, receiver, input stage, etc). The attenuator does a double attenuation of the reflections because they pass through the attenuator twice on the journey back to the DAC where they can cause jitter. So if you have 6dB attenuator the reflections will be reduced by 12dB. Even so it's probably beneficial to try to minimise impedance mismatches as much as possible & where practical.
You would probably need to measure the characteristic impedance of your RCA connectors & make one side of your T-pad match this impedance. This may be just theoretically optimal & the in the real world you may not hear any difference.
You are the first from here to actually try this out - well done - it's shameful that 20 pages were spent arguing the theory without one person trying it. Anyway, all in the past!
As you probably read, the way that the attenuators work in this situation seems to be by reducing the reflections caused by impedance mismatches encountered by the SPDIF signal along the signal path (this includes transmitter, outputs stage, connectors, cable, receiver, input stage, etc). The attenuator does a double attenuation of the reflections because they pass through the attenuator twice on the journey back to the DAC where they can cause jitter. So if you have 6dB attenuator the reflections will be reduced by 12dB. Even so it's probably beneficial to try to minimise impedance mismatches as much as possible & where practical.
You would probably need to measure the characteristic impedance of your RCA connectors & make one side of your T-pad match this impedance. This may be just theoretically optimal & the in the real world you may not hear any difference.
Excellent listening report sendler
Thanks, this is the kind of stuff we need to begin to understand why these pads are making a difference.
The only time that the RCA connector impedance mis-match could be a problem is in a home theater environment. That is when you have a RCA interconnect from the unit to a wall-plate, then an inside the wall RCA interconnect to another wall-plate, and finally another RCA interconnect to the next unit. That would be a total of 6 (or maybe more) connector pairs. And in home theater it would be video not SPDIF audio. The length of a cable mis-match is a very important factor and an RCA connector is very short.
102 ohms cable
.
So I will stick with RCAs for now and also continue to use the Belden 1701a. I am happy with the sound of the attenuation values for now but might hope to make it even better by tweaking the values to get the best balance of performance from my non 75R components.
.
This is also more of a legacy project for me to help other, less diy- more off the shelf system owners as I am quickly moving toward USB to I2S direct connections for my dac boards as all of my music and radio (soon TV?) comes from my laptop and hard drives.
Many of us are stuck with RCAs so the attenuator values may need to be adjusted. Another issue is the 102R impedance of the Belden 1701a solid core in Teflon bonded pair wire that I am using. I use this as it has proven to sound better than any of my other digital cables including Apogee Wideye which is a really excellent ready made cable. Radio Shack Gold video is also surprisingly close for even less money. The Memorial digital cable with foamed coax and Canare RCAs sounds really bad despite the components being supposedly top shelf. I keep it around just to show people how much difference there can be between digital cables. It sounds very slow and opaque. Boring. Generic cat5 pairs don't sound that great either. I have also tried countless scraps of various video coax without good results and anyone using stranded coax for analog interconnects is missing out compared to the bonded pairs of 1701a. It is also interesting that a good spdif cable will still sound a little better than an optical connection from the same source and dac.
.
So I will stick with RCAs for now and also continue to use the Belden 1701a. I am happy with the sound of the attenuation values for now but might hope to make it even better by tweaking the values to get the best balance of performance from my non 75R components.
.
This is also more of a legacy project for me to help other, less diy- more off the shelf system owners as I am quickly moving toward USB to I2S direct connections for my dac boards as all of my music and radio (soon TV?) comes from my laptop and hard drives.
Yes Scott, that was the intention of starting this thread - to do exactly what you are doing & tease out what works & what doesn't work, empirically. I'm just pointing out possible issues that are worth considering & theoretically optimal situations (as best as my understanding allows me) but as you quiet rightly point out the real world is not a physics lab & certain compromises/best fit solutions have to be found. I guess the best rule to follow in this is to avoid as many impedance mismatches as possible & reduce the number of reflections. If you have these T-pads on the ends of your Belden cable would it be worth changing the T-pad impedance facing into the cable to 102R?
Another way of dealing with all this is to choose a cable length that alleviates the problem i.e the reflections don't arrive at the DAC during the critical decision window. But this is another whole can of worms best left unopened.
Also, it has to be born in mind that I've focused on the reduction in reflections effect that these RF attenuators show on the scope & the possible consequent jitter reduction. There may well be other issues that these attenuators are ameliorating - RF noise for one? What DAC are you using? Sorry if you've said already!
I'm glad you found an improvement from the T-pads & if you experiment further, let us know.
Another way of dealing with all this is to choose a cable length that alleviates the problem i.e the reflections don't arrive at the DAC during the critical decision window. But this is another whole can of worms best left unopened.
Also, it has to be born in mind that I've focused on the reduction in reflections effect that these RF attenuators show on the scope & the possible consequent jitter reduction. There may well be other issues that these attenuators are ameliorating - RF noise for one? What DAC are you using? Sorry if you've said already!
I'm glad you found an improvement from the T-pads & if you experiment further, let us know.
Interestingly, I didn't see very much improvement in the overshoot and ringing from the Hiface using the attenuators. When I get back home later this week, I'll post some photos of the scope traces if you don't mind. The comparison with a very cheap Chinese USB-SPDIF converter and a home-built level shifter was instructive.
Do you know what kind of transformer they used? My homebuilt unit uses a Murata DA103.
Do you know what kind of transformer they used? My homebuilt unit uses a Murata DA103.
The SPDIF output trafo used is a 2:1 Murata 78604/1C
I don't mind if you show the scope shots & your comparison with the cheap Chinese USB DAC but also include your listening impressions - that's what it's all about, isn't it? BTW, I thought you were going to do analogue out measurement & analysis - are these for later?
I don't believe the attenuators will do much if anything about the overshoot but certainly will effect the reflections (not sure if this is ringing?) i.e the squiggles on the top of the waveform seen in the scope shots I posted over in the vendor's bazaar.
It's worth showing them here too as I believe they clearly show the effect attenuators have on the waveform:
Without Attenuators first
With Attenuators:
And from Joseph K:
Stock Hiface without attenuators:
Stock Hiface with attenuators:
I don't mind if you show the scope shots & your comparison with the cheap Chinese USB DAC but also include your listening impressions - that's what it's all about, isn't it? BTW, I thought you were going to do analogue out measurement & analysis - are these for later?
I don't believe the attenuators will do much if anything about the overshoot but certainly will effect the reflections (not sure if this is ringing?) i.e the squiggles on the top of the waveform seen in the scope shots I posted over in the vendor's bazaar.
It's worth showing them here too as I believe they clearly show the effect attenuators have on the waveform:
Without Attenuators first
An externally hosted image should be here but it was not working when we last tested it.
With Attenuators:
An externally hosted image should be here but it was not working when we last tested it.
And from Joseph K:
Stock Hiface without attenuators:
An externally hosted image should be here but it was not working when we last tested it.
Stock Hiface with attenuators:
An externally hosted image should be here but it was not working when we last tested it.
Big Dac Board
That makes sense to set the cable side to 102R since I have T pads on both ends of the cable. I will also have to try 50R on any side that is connecting to an RCA. I am intrigued enough by this concept to make some tiny circuit boards to hold surface mount resistors which I will be able to change out easily.
.
.
I have two Big Dac Boards that I can run in identical configurations so the last interesting test would be to change the terminating resistor for the RCA on the dac board to 50R to hear if that sounds any different to the 75R that is there now.
.
http://www.diyaudio.com/forums/digital-line-level/137976-experience-diy-dac-326.html#post2321397
.
.
That makes sense to set the cable side to 102R since I have T pads on both ends of the cable. I will also have to try 50R on any side that is connecting to an RCA. I am intrigued enough by this concept to make some tiny circuit boards to hold surface mount resistors which I will be able to change out easily.
.
.
.
I have two Big Dac Boards that I can run in identical configurations so the last interesting test would be to change the terminating resistor for the RCA on the dac board to 50R to hear if that sounds any different to the 75R that is there now.
.
http://www.diyaudio.com/forums/digital-line-level/137976-experience-diy-dac-326.html#post2321397
.
Attachments
- Status
- Not open for further replies.