A bunch of random questions for the DIY Guru's.

Sep 22, 2005 at 12:38 AM Thread Starter Post #1 of 66

philodox

Headphoneus Supremus
Joined
Dec 12, 2003
Posts
10,244
Likes
17
Hey Guys,

I'll try not to make a habit out of these threads, but I had a few more questions. Not really related per se, but just some stuff that I've been mulling over in my head the last little while.

1) What, in your estimation, is the best single connecter to use for a balanced connection [as opposed to using 2 XLRs]? I'm looking for a connector type, but if you have a specific make in mind share that as well. Basically I'd like something so that there could be one cable that would run the balanced connection from the source to the amps input and also could be used for the balanced headphones. I know that aerius has been using something like this on his amp, but I figured there might be some other options out there and I don't really have any info on what he is using anyways. I would need a recommendation for a cable mounted male and a chassis mounted female... or visa versa I suppose.

1b) Likewise, but the best single connector for an unbalanced connection. [instead of using 2 RCAs]

2) From a technical standpoint what is the best type of digital connection? I figure AES/EBU is better than coaxial RCA or optical toslink, but I'm not sure if there are other options. Are there I2S connectors or is that connection just internal from the transport?

3) Is it possible to have the volume control on a DAC rather than the amp? I think I am missunderstanding, but I've seen lots on digital stepped attenuation and also on equalization in the digital domain and am wondering if this is possible? In DAC's with a volume control is it actually just affecting the analog stage or how does that work?

4) What are those funky circular power connectors that are supposed to be better than IEC's called? Anything else that should be considered?

5) In a balanced amp that has two power supplies is it absolutely essential that there be two power plugs for it or is there some way that you could use only one?

6) What are your thoughts on fuses in audio gear? I've noticed that some hi-fi equipment employs them and some does not and am unsure if they are necissary, a good idea, or detrimental to sound quality.

7) I see a lot of people advocating external power supplies. If a power supply is internal is it possible to put an insane amount of shielding between it and the circuitry in order to achieve the same effect? If so, what is the best material for shielding? If an external supply is used how far does it need to be from the audio component? Can the audio component sit on top of it or is that too close?

8) I've heard some discussion on the proper output impedances and input impedances [probably phrasing this wrong] on audio gear. The whole idea behind the Nitrogen cables of changing the impedances of cables got me thinking about this. What is the ideal output from your analog stage [on the DAC] and what is the ideal input on the amp assuming that no preamp is being used? [a power amp or headphone amp]

Thanks in advance,

Jason
340smile.gif
 
Sep 22, 2005 at 1:16 AM Post #2 of 66
1: 4pin xlr, with "sleve/shell" grounded. just like the k-1000's. i like neutrik, they make both male and female in both panel and cable mount.

3: the flute did this, im not sure how though. there are also chips that function as "stepped attenuators" and are available in audio tapers.
 
Sep 22, 2005 at 1:25 AM Post #3 of 66
1. 4 port xlr connectors
1b. 6.5 jack or 3 xlr connectors
5. i'm preetty sure 1 power plug would be sufficent
6. i don't use the fuse in mine rather i use the inbuilt current protection on my regulators
7. it would be best to have a seperat power supply but by using the right components and adding your own sheilding it would also be as good.
 
Sep 22, 2005 at 2:00 AM Post #4 of 66
Quote:

Originally Posted by nikongod
1: 4pin xlr, with "sleve/shell" grounded. just like the k-1000's. i like neutrik, they make both male and female in both panel and cable mount.


Quote:

Originally Posted by Kenny12
1. 4 port xlr connectors


That is what I was thinking as well, but in most balanced setups don't they use 2 3 pin XLRs? Are the other two pins just not used? Quote:

Originally Posted by Kenny12
5. i'm preetty sure 1 power plug would be sufficent


That is what I figured, but all the balanced amps I've seen have two.
confused.gif
 
Sep 22, 2005 at 2:31 AM Post #5 of 66
Can I guess?

1) Please don't do this. Crosstalk from running both channels in the same jacket sort of defeats the whole purpose of a balanced signal, and then some. If you must, make sure that both pairs are shielded separately. A 4 pin XLR would work, but still a major compromise.

2) I think that ADAT optical is the industry choice. AES/EBU is also very robust, and the cables are much easier to build.

3) Yes, but it is important to step up to a higher bit depth to avoid truncation error and loss of dynamic range. 24 bit or higher is common. Digital attenuation of a 16 bit signal will result in loss of dynamic range and resolution.

4) Neutrik Powercons

5) One is fine, but two is more fun! It allows complete isolation of the 2 channels.

6) Power fuses are a must! Why anybody would leave them out is a mystery to me, especially with an external power supply. Why risk the electronics and your house? All true production gear has fusing.

7) It is easier to isolate PSU noise if it is not in the same housing. I don't know the optimum distance, but the signals drop off with the square of the distance, so that first foot makes a big difference.

8) I am curious about the whole cable impedance thing myself, so no guess.


gerG


edit: DOH! I just noticed that the question was not directed at me. Sorry, I answered out of turn.
 
Sep 22, 2005 at 2:54 AM Post #6 of 66
Quote:

Originally Posted by philodox
That is what I was thinking as well, but in most balanced setups don't they use 2 3 pin XLRs? Are the other two pins just not used?


in a "normal" 3-pin xlr for a ballanced signal, pin1 is connected to 1 side of the signal, pin2 to the other side of the signal, and pin3 is hooked up to a "ground antenna" to absorb noise, pin3 is also typpically tied to the shell of the connector, so that you ground the plug before you coneect the signal.
Quote:

That is what I figured, but all the balanced amps I've seen have two.
confused.gif


the 2 psu's may be used to isolate the chanels.

how bad is the crosstalk in a twisted/braded 4-strand cable driving a headphone sized load? i have started to hear it come up recently, but it seems like it came out of the blue. im sure it is present in high-power signals but we are not bundling the wires to our tweeters with the oo ga cables
wink.gif
going to our washing-machine-sized subwoofers... maybe this just further proves my ignorance.
 
Sep 22, 2005 at 3:05 AM Post #7 of 66
7) The power supply transformer is the most common offender due to magnetic field leakage. Orientation can be varied. Distance can be increased. Even potted toroids are not perfect. In order to avoid contamination of sensitive audio circuits, some companies use measures like MuMetal cans around the transformers, magnetic film wraps around the toriods, and nickel based sprayon coatings. Remember that aluminum does NOT provide a magnetic shield. So, putting an amp on top of a power supply, even if in separate aluminum boxes, may not give you the real performance improvement you are looking for. It takes measurements to know if you've made something better or worse.

Try looking at microphone preamp power supplies from high-end manufacturers. Some ideas may come to mind. All of them tend to be expensive and involve custom mechanical work.
 
Sep 22, 2005 at 3:08 AM Post #8 of 66
1b: 1/4" TRS connectors. ALL my connections are TRS bar my cable box and dj gear. so much better then RCA, more compact then XLR. i can use the same cables balanced or unbalanced and connecting between balanced and unbalnaced. i use some xlr's but only for a few balanced runs.
 
Sep 22, 2005 at 3:35 AM Post #9 of 66
Quote:

Originally Posted by nikongod
in a "normal" 3-pin xlr for a ballanced signal, pin1 is connected to 1 side of the signal, pin2 to the other side of the signal, and pin3 is hooked up to a "ground antenna" to absorb noise, pin3 is also typpically tied to the shell of the connector, so that you ground the plug before you coneect the signal.


you sure? pin 1 is ground....
 
Sep 22, 2005 at 4:15 AM Post #11 of 66
all my xlr gear is wired pin 1 shield, pin 2 hot, pin 3 cold. the pin 1 hot isnt common at all anymore. fwiw 1/4" trs are wired T hot, R cold, S shield.
 
Sep 22, 2005 at 5:50 AM Post #13 of 66
2)

From my readings on diyaudio.com I’ve picked up on a few things -
When it comes to transmitting digital audio, what matters most of all is jitter as the more you get will end up as distortion on the end D to A process

Coax > Optical, though there might be a case where its the opposite but that will be because of a poor coax setup

For a proper coax setup it must have constant impedance along the whole transmission line from start to finish and the specs call for 75ohms here but I’m not sure if you could set it up to work at 50ohm or something else just making sure that it is constant but it is best to stick to 75 imho

Then to also have this constant 75ohms you need your connectors to have the same properties as well and the poor little rca's do NOT have this property (they measure in at 30ohms from what jocko on diyaudio reports) thus this adds reflections in the signal and that equals jitter. So you have to use 75ohm bnc connectors on both ends of your ic

This is all to transmit your signal over the s/pdif protocol which in converting to a more useful digital signal inside the receiver chip to feed the rest of the dac, is also very prone to jitter as the clock is recovered from the s/pdif (hence why you want to get the signal there in as good a shape as you can in the first place)

So if you are feeling adventurous there are a few ways to overcome the s/pdif weaknesses, 1. To transmit a clock on a separate wire either to the dac or from it to the source and have the source slave off that clock, or 2. Directly link your source to dac via I2S which requires 4 wires

There is plenty more on the subject and this is just the basics of it from my understanding of it
 
Sep 22, 2005 at 6:17 AM Post #14 of 66
Quote:

Originally Posted by gerG
Can I guess?

1) Please don't do this. Crosstalk from running both channels in the same jacket sort of defeats the whole purpose of a balanced signal, and then some. If you must, make sure that both pairs are shielded separately. A 4 pin XLR would work, but still a major compromise.



4 Pin is not a major compromise. In fact, the whole using two separate 3 Pin XLR's is just silly. Make sure the shields are grounded to the shell of the XLR. Most (namely Neutrik) have a nice tab to solder to.

Headroom uses Cardas cables that run both channels in the same jacket. I'd have to cut one open to be sure, but I doubt both channels are individually shielded.

-Ed
 
Sep 22, 2005 at 11:25 AM Post #15 of 66
Quote:

Originally Posted by gerG
edit: DOH! I just noticed that the question was not directed at me. Sorry, I answered out of turn.


What do you mean? You gave some great answers, thanks man.
smily_headphones1.gif
Quote:

Originally Posted by gerG
1) Please don't do this. Crosstalk from running both channels in the same jacket sort of defeats the whole purpose of a balanced signal, and then some. If you must, make sure that both pairs are shielded separately. A 4 pin XLR would work, but still a major compromise.


Quote:

Originally Posted by Edwood
4 Pin is not a major compromise. In fact, the whole using two separate 3 Pin XLR's is just silly. Make sure the shields are grounded to the shell of the XLR. Most (namely Neutrik) have a nice tab to solder to.


Ok, so who is right here? I am still confused about how the 4 pin connectors work. If in a three pin setup one is used for hot, one for cold and one for a ‘ground antenna’, what gets shared in a 4 pin setup? How are we moving from 6 wires to 4 wires? Quote:

Originally Posted by skyskraper
1b: 1/4" TRS connectors. ALL my connections are TRS bar my cable box and dj gear. So much better then RCA, more compact then XLR. I can use the same cables balanced or unbalanced and connecting between balanced and unbalanced. I use some xlr's but only for a few balanced runs.


I’ll look into those… I’ve been thinking that it might be better to stick with RCA’s in the end for unbalanced though just for compatibility sake. Those Eichmann bullets and RCA jacks look interesting. Quote:

Originally Posted by gerG
2) I think that ADAT optical is the industry choice. AES/EBU is also very robust, and the cables are much easier to build.


I've never heard of ADAT, I'll have to look into this.
cool.gif
Quote:

Originally Posted by DaKi][er
2) This is all to transmit your signal over the s/pdif protocol which in converting to a more useful digital signal inside the receiver chip to feed the rest of the dac, is also very prone to jitter as the clock is recovered from the s/pdif (hence why you want to get the signal there in as good a shape as you can in the first place)

So if you are feeling adventurous there are a few ways to overcome the s/pdif weaknesses, 1. To transmit a clock on a separate wire either to the dac or from it to the source and have the source slave off that clock, or 2. Directly link your source to dac via I2S which requires 4 wires.



This is exactly what I'd like to do eventually. I think that bypassing SPDIF is a great step towards removing jitter. What sort of connector would you use to transmit I2S?

Quote:

Originally Posted by gerG
3) Yes, but it is important to step up to a higher bit depth to avoid truncation error and loss of dynamic range. 24 bit or higher is common. Digital attenuation of a 16 bit signal will result in loss of dynamic range and resolution.


Exactly what I was looking for. So is there any problem with doing this? Would this be as good as having a stepped attenuator on the amp? Is there any reason that it would be better? Quote:

Originally Posted by gerG
4) Neutrik Powercons


That is the one, thanks.
smily_headphones1.gif
Quote:

Originally Posted by dgardner
7) The power supply transformer is the most common offender due to magnetic field leakage. Orientation can be varied. Distance can be increased. Even potted toroids are not perfect. In order to avoid contamination of sensitive audio circuits, some companies use measures like MuMetal cans around the transformers, magnetic film wraps around the toriods, and nickel based sprayon coatings. Remember that aluminum does NOT provide a magnetic shield. So, putting an amp on top of a power supply, even if in separate aluminum boxes, may not give you the real performance improvement you are looking for. It takes measurements to know if you've made something better or worse.


So is it just the transformers that you have to worry about? Would putting the transformer in a copper box with a hole for the outputs do a good enough job? Quote:

Originally Posted by gerG
8) I am curious about the whole cable impedance thing myself, so no guess.


It is interesting... I'm looking at it from more of the perspective of fixing things on the output of the DAC and the input of the amp rather than patching the problem with a cable though.
 

Users who are viewing this thread

Back
Top