AndrewT,
To clarify, I never meant to imply that CAT5 would not work for audio signals. I do not use it personally for pragmatic reasons. I do not like solid core wire unless it is stationary on a pcb due to its tendency to fatigue (I did not know about this stranded CAT5 you mentioned earlier). My limited experience with CAT5 found it frustrating to terminate and aesthetically offending. Most CAT5 that I have ventured across is not shielded, and I have limitless access to shielded twisted pair cables. I will take in consideration what you said about twisted pair being self shielding, and see if I can confirm this for myself. For short audio interconnect runs, I use standard coaxial RCA plugs from radioshack (if I can't wait) or monoprice (if I can wait).
My response to this post was primarily to correct some misstated information about cross sectional area vs resistance, and to tend to a case of what I diagnosed as acute audiophilia nervosa.
For a succinct reference on audio related transmission, I find this page to be very useful.
Anatech, the engineer in me does wish for matched impedances of the source, input, and cable. I have worked with GPS signals and found the hard way how important matching impedances can be. The pragmatist in me realizes that I can't hear the nanoseconds of phase shift caused by the mismatch in my system, and turns my attention to the much larger issues in my reproduction chain.
Speaking of high speed signals, we just got a new toy in the lab, the MSO4104. 1GHz, 4 analog channels, with 16 logic probes.
To clarify, I never meant to imply that CAT5 would not work for audio signals. I do not use it personally for pragmatic reasons. I do not like solid core wire unless it is stationary on a pcb due to its tendency to fatigue (I did not know about this stranded CAT5 you mentioned earlier). My limited experience with CAT5 found it frustrating to terminate and aesthetically offending. Most CAT5 that I have ventured across is not shielded, and I have limitless access to shielded twisted pair cables. I will take in consideration what you said about twisted pair being self shielding, and see if I can confirm this for myself. For short audio interconnect runs, I use standard coaxial RCA plugs from radioshack (if I can't wait) or monoprice (if I can wait).
My response to this post was primarily to correct some misstated information about cross sectional area vs resistance, and to tend to a case of what I diagnosed as acute audiophilia nervosa.
For a succinct reference on audio related transmission, I find this page to be very useful.
Anatech, the engineer in me does wish for matched impedances of the source, input, and cable. I have worked with GPS signals and found the hard way how important matching impedances can be. The pragmatist in me realizes that I can't hear the nanoseconds of phase shift caused by the mismatch in my system, and turns my attention to the much larger issues in my reproduction chain.
Speaking of high speed signals, we just got a new toy in the lab, the MSO4104. 1GHz, 4 analog channels, with 16 logic probes.
I buy stranded CAT5 cable as 50ft "patch cords" and then cut it up for interconnects. If your not put off by the colors you can make them look nice and professional by running a twisted pair through some small vinyl tubing.
Like this stuff.....
http://cgi.ebay.com/LURE-PROTECTOR-...goryZ794QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
I don't have any of the tubing with me right now, but here is a close up of the stranded CAT5.....
Like this stuff.....
http://cgi.ebay.com/LURE-PROTECTOR-...goryZ794QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
I don't have any of the tubing with me right now, but here is a close up of the stranded CAT5.....
Not too shabby there anonymous1. I guess the orange color could match the home depot extension cord cables! Do you strain relief to the clear tube or the twisted pair? What gauge are the stranded wires?
The summer after I graduated high school I worked in a paper mill as an intern in the process control department. The engineers there taught me the importance of shielding and practical wiring rules. I was able to keep some of the scrap wire from a control project, 4 pair of large twisted pair wires individually shielded in a braided shield jacket. I used this giant bright blue belden water hose as my interconnect in my car and home projects. It was cool to me, as a teenager, to have this audacious expensive cable as an interconnect. I would kick my own rear if I met myself today with that ridiculous looking cable.
The summer after I graduated high school I worked in a paper mill as an intern in the process control department. The engineers there taught me the importance of shielding and practical wiring rules. I was able to keep some of the scrap wire from a control project, 4 pair of large twisted pair wires individually shielded in a braided shield jacket. I used this giant bright blue belden water hose as my interconnect in my car and home projects. It was cool to me, as a teenager, to have this audacious expensive cable as an interconnect. I would kick my own rear if I met myself today with that ridiculous looking cable.
I have no idea what the gauge is; I don't have digital calipers. If I had to guess using my manual calipers I would say maybe 0.2mm x 7 strands.
I just use some heat shrink tubing over the very end of the tube for some added strain relief.
Speaking of funky cables; how about these technicolor cables...
http://i10.tinypic.com/2n8df9x.jpg
Those are 3 twisted pair braided, but I've since switched to just using a single twisted pair.
Keith,
I get the foam from work. It comes as packing material in shipments and usually gets tossed out if I don't take it. It's normally about an inch or so thick and somewhere around the size you mentioned.
I use it for packaging purposes but have thought about using it for speaker building, only that would require me to actually build some speakers.
I just use some heat shrink tubing over the very end of the tube for some added strain relief.
Speaking of funky cables; how about these technicolor cables...
http://i10.tinypic.com/2n8df9x.jpg
Those are 3 twisted pair braided, but I've since switched to just using a single twisted pair.
Keith,
I get the foam from work. It comes as packing material in shipments and usually gets tossed out if I don't take it. It's normally about an inch or so thick and somewhere around the size you mentioned.
I use it for packaging purposes but have thought about using it for speaker building, only that would require me to actually build some speakers.
Hi Andrew,
Why use the wrong wire when the proper stuff is available???
I get all my CAT5 and CAT6 for free, well I used to. In industrial settings, and in home settings, there is an awful lot of electrical noise in the air. The power line harmonics go way up past 2 KHz (from my last test). Therefore, a shielded cable is called for. The problem with a twisted pair is that the CMRR only works if the source impedances are very closely matched, same for the receiving end. So if you are pounding the normal single ended -10 dBu signal down a twisted pair, you are not getting the cancellation you think you are. The situation becomes much worse as your source impedance goes up.
So, you can get a double shielded cable. The inner shield is your signal return and the outer shield is connected at the source end only. This stuff is wonderful!
So, you want to run balanced? I mean real, to the spec., balanced. Not the iffy balanced we see so often in consumer gear. Then, use a shielded twisted pair cable like they've been using for years in recording studios. That stuff works properly and it's been engineered to do the job.
So, will CAT5 work? Yup. So will many other things. I use the proper cable types. One area where CAT5 really shines is in telephone wiring. The tighter twist (than CAT3) eliminates a lot of noise pickup. I've used it for years in telecommunication systems as an upgrade over CAT3 (good) or quad (terrible stuff for even phone work). It's just what I use, not an added cost to the user.
-Chris
Why use the wrong wire when the proper stuff is available???
I get all my CAT5 and CAT6 for free, well I used to. In industrial settings, and in home settings, there is an awful lot of electrical noise in the air. The power line harmonics go way up past 2 KHz (from my last test). Therefore, a shielded cable is called for. The problem with a twisted pair is that the CMRR only works if the source impedances are very closely matched, same for the receiving end. So if you are pounding the normal single ended -10 dBu signal down a twisted pair, you are not getting the cancellation you think you are. The situation becomes much worse as your source impedance goes up.
So, you can get a double shielded cable. The inner shield is your signal return and the outer shield is connected at the source end only. This stuff is wonderful!
So, you want to run balanced? I mean real, to the spec., balanced. Not the iffy balanced we see so often in consumer gear. Then, use a shielded twisted pair cable like they've been using for years in recording studios. That stuff works properly and it's been engineered to do the job.
So, will CAT5 work? Yup. So will many other things. I use the proper cable types. One area where CAT5 really shines is in telephone wiring. The tighter twist (than CAT3) eliminates a lot of noise pickup. I've used it for years in telecommunication systems as an upgrade over CAT3 (good) or quad (terrible stuff for even phone work). It's just what I use, not an added cost to the user.
-Chris
Hi gtforme00,
I was really referring to the accepted practice of having a low source impedance and a higher (at least 10X) receiving end impedance. With balanced signals, both phases must be very close in impedance across the frequency ranges of interest. This includes the noise frequencies we want to reject. Shielding really helps here too.
For a balanced line, you can use receiving end impedances down to about 600 ohms, although they are often higher. Still, lower than single ended (normally 50 K ohms). The driving end therefore needs to run less than 100 ohms - worst case.
-Chris
I was really referring to the accepted practice of having a low source impedance and a higher (at least 10X) receiving end impedance. With balanced signals, both phases must be very close in impedance across the frequency ranges of interest. This includes the noise frequencies we want to reject. Shielding really helps here too.
For a balanced line, you can use receiving end impedances down to about 600 ohms, although they are often higher. Still, lower than single ended (normally 50 K ohms). The driving end therefore needs to run less than 100 ohms - worst case.
-Chris
Hi gtforme00,
I sincerely doubt I taught you anything new. I was clarifying my intent in my earlier post for the benefit of others. I trust you have a better handle on signal transmission than I do. I've just worked in audio for over 30 years and I'm speaking to conventions as they appear in the audio industry.
Is there anything you find fault with in my post?
-Chris
I sincerely doubt I taught you anything new. I was clarifying my intent in my earlier post for the benefit of others. I trust you have a better handle on signal transmission than I do. I've just worked in audio for over 30 years and I'm speaking to conventions as they appear in the audio industry.
Is there anything you find fault with in my post?
-Chris
The the orange cable in the photo does not look like Cat 5 cable to me. Cat 5 cable uses very closely controlled conductor spacing and twisting.
In this day and age of all sorts of new devices generating RFI, using unshielded interconnects in unbalanced circuits will bite you on the butt sooner or later.
In this day and age of all sorts of new devices generating RFI, using unshielded interconnects in unbalanced circuits will bite you on the butt sooner or later.
Hi
(I've been distracted for a few days) Thank you for the posts . .
I've been looking for an economical option, and am now leaning towards shielded video coax. A good list of these:
http://www.bluejeanscable.com/store/shopbycable/index.htm
from which, with favourable opinions elsewhere at this stage I'm favouring either
- AWG 18, RG6 Belden 1694A http://www.bluejeanscable.com/pages/technicaldocs/1694tech.htm (about US$4/ m).
- AWG 22, RG-59 Belden 89259 http://www.bluejeanscable.com/pages/technicaldocs/89259tech.htm (about US$10/ m, probably not worth the extra).
- AWG 24, RG-59 Canare LV-61S http://www.bluejeanscable.com/pages/technicaldocs/canarelv.pdf (about US$2/ m).
AMV8/ Andrew T/ Anatech/ gtforme00/ . .
I know next to nothing about matching/ optimising impedances of eg pre and power amps, and cables . .
"the accepted practice of having a low source impedance and a higher (at least 10X) receiving end impedance"
What happens to sound quality if you're much 'under' this ratio?
Does this apply equally to source > pre, pre > power amp; and eg within a power amp, to the choice of a pot K value?
'Threshold' - At what point of closer source and receiving impedance should something be done, eg if pre > power amp where only 3*, put a buffer in one end?
Interconnects of 5 m plus ~
Don (AMV8),
Apart from shielded cable
"You need to connect the shield at the pre amp, so that you effectively extend the pre amp shield around the interconnect to shield the signal wire inside"
I'm not sure exactly how you'd do that (do you have a picture?), but it sounds somewhat time consuming and fiddly.
For these longer interconnects, might Canare RCA plugs with claimed "superb electrical contact, shielding . . " http://www.bluejeanscable.com/store/component/index.htm (available here locally for c $US 5) sufficiently avoid that?
Thanks again!
(I've been distracted for a few days) Thank you for the posts . .
I've been looking for an economical option, and am now leaning towards shielded video coax. A good list of these:
http://www.bluejeanscable.com/store/shopbycable/index.htm
from which, with favourable opinions elsewhere at this stage I'm favouring either
- AWG 18, RG6 Belden 1694A http://www.bluejeanscable.com/pages/technicaldocs/1694tech.htm (about US$4/ m).
- AWG 22, RG-59 Belden 89259 http://www.bluejeanscable.com/pages/technicaldocs/89259tech.htm (about US$10/ m, probably not worth the extra).
- AWG 24, RG-59 Canare LV-61S http://www.bluejeanscable.com/pages/technicaldocs/canarelv.pdf (about US$2/ m).
AMV8/ Andrew T/ Anatech/ gtforme00/ . .
I know next to nothing about matching/ optimising impedances of eg pre and power amps, and cables . .
"the accepted practice of having a low source impedance and a higher (at least 10X) receiving end impedance"
What happens to sound quality if you're much 'under' this ratio?
Does this apply equally to source > pre, pre > power amp; and eg within a power amp, to the choice of a pot K value?
'Threshold' - At what point of closer source and receiving impedance should something be done, eg if pre > power amp where only 3*, put a buffer in one end?
Interconnects of 5 m plus ~
Don (AMV8),
Apart from shielded cable
"You need to connect the shield at the pre amp, so that you effectively extend the pre amp shield around the interconnect to shield the signal wire inside"
I'm not sure exactly how you'd do that (do you have a picture?), but it sounds somewhat time consuming and fiddly.
For these longer interconnects, might Canare RCA plugs with claimed "superb electrical contact, shielding . . " http://www.bluejeanscable.com/store/component/index.htm (available here locally for c $US 5) sufficiently avoid that?
Thanks again!
a lowish Rs=50r and a highish Zin=50k would be a ratio of 1000:1 and this works very well.otto88 said:
So the ratio can be high without detriment. This applies even more with balanced connections where the suggestion is that Rs=10r and Zin=1M0 would benefit CMRR.
If a passive pot is inserted in the line then I suggest.
200*Rs <= 10*Pot <= Zin as a minimum and 400*Rs<=20*pot<=Zin is even better.
at one time i had a spool of the teflon insulated version of RG-174. and used it for just about everything. the .100" diameter made it useful for going just about anywhere, and the teflon jacket was good for being cut and pinch resistant. it's well shielded too... it's a little bit difficult to work with because it's stiff, but once you get the hang of stripping it it's great. and you never get center to shield shorts from soldering. so you can use shorter pigtails at the ends.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.