Hi, I just wanted to share my personnal experience with Cat5 for analog audio transmission.
Last year, I were asked to plan audio distribution on a street about 2500ft long... Our tests was made with four 1000' Cat5e spools back-to-back (still looped on the box), so about 4000' of cheap, unshielded cable... And the powered speaker (with balanced input) did sound just as if we were plugging the source directly on it!
Everything worked perfectly! Not even a single bug for the whole week of fest.
Cat5 (shielded or not) looks perfect for balanced line-level audio; but don't even try to do it without a balanced signal.
Last year, I were asked to plan audio distribution on a street about 2500ft long... Our tests was made with four 1000' Cat5e spools back-to-back (still looped on the box), so about 4000' of cheap, unshielded cable... And the powered speaker (with balanced input) did sound just as if we were plugging the source directly on it!
Everything worked perfectly! Not even a single bug for the whole week of fest.
Cat5 (shielded or not) looks perfect for balanced line-level audio; but don't even try to do it without a balanced signal.
Line Level Stereo Audio over Cat5
What would the wiring look like for line level audio via Cat5? I've read that I need to use one twisted pair per signal. Geerlingguy's wiring diagrams don't appear to reflect that. What do I do with the unused pairs of wires? Ground them? I want to hookup a computer audio out or iPod to my receiver's Zone 2 audio in (L/R). The laptop or iPod input is on my deck along with Zone 2's speakers.
What would the wiring look like for line level audio via Cat5? I've read that I need to use one twisted pair per signal. Geerlingguy's wiring diagrams don't appear to reflect that. What do I do with the unused pairs of wires? Ground them? I want to hookup a computer audio out or iPod to my receiver's Zone 2 audio in (L/R). The laptop or iPod input is on my deck along with Zone 2's speakers.
Balanced Line Audio over CAT5 with multiple receivers
I'm wanting to set up a system to send line level stereo audio signal over CAT5 cable a fairly long distance, and have multiple receivers at many points along the way. I am thinking of using TI's DRV134 for the Transmitter, and INA137 for the receiver(s) it seems these are matched to each other, since the DRV134 has a gain of 2 and the INA137 has a gain of 0.5. so the two together would produce an overall gain of 1. I like the idea of the voltage on the line being a little higher, seems like noise would affect it less, and it would travel better, and would support multiple receivers better. I am wondering if I should take this idea a little further, and put an op-amp buffer in front of the DRV134, and set it to a gain of 6 or so, then another one set for 1/6 after the INA137, this would boost my line level to something like a maximum of 24V Peak to Peak, Which should be fine with a +/- 15V power supply.
Does anyone have any experience with doing something like this? I want to have a quite a few receivers, perhaps 40 or more, it doesn't seem like multiple receivers should be a problem since the DRV134 can output 85mA and each INA137 has an input impedance of 24K Ohms, seems to me that with a 24V Peak to Peak signal, going into a 24K Ohm load would only draw 1mA, so I should be able to drive 85 receivers with one transmitter. I don't know what the cable impedance is, but it seems I should be able to drive more than 40 receivers anyway. I am also wondering if it needs to be one long cable run with drops along the way, or if I could branch off several times with just the one driver running it all. I would us a lot less cable if I could branch off.
What is the recommended power supply for this kind of thing. at first I thought I could just buy a cheap switching power supply to do the job, but I can't seem to find one that outputs +/- 15V, and with switchers, you can't just put 2 of them together, because they have one end tied directly to the power line, and if you try to connect the + output of one to the - output of the other, it would short out and go POOF. Does anyone have a good way to build my own switching power supply? Or should I just stick with the old center tap transformer going into bridge rectifier, charge a few ridiculous capacitors, then regulate the output with a 7815 and a 7915? I wouldn't need very big transformers, and it would be easy to build, I just notice almost all consumer electronics have dropped the transformer and now have switching power supplies, presumably for better efficiency and lower cost to produce, so I'm wondering if DIY electronics have got into that yet.
Lastly, does anyone have a suggestion on a power amplifier to run speakers from the output of the receivers? It seems to me that since I already have a balanced signal, I could perhaps put two INA137's, with the polarity reversed, and end up with two opposing outputs that I could easily feed each into a power amplifier in a push-pull arrangement. I am also wondering if I should reduce the signal back down to line level, just to be amplifying it again, or if I should just leave it and not amplify it as much as I would need to from line level. I am thinking reducing it then amplifying it a second time would only create extra distortion that I shouldn't need to bother with. is there such thing as a big op-amp chip with power transistors I could bolt to a heat sink and set the gain to whatever I want with feedback resistors? is that over simplifying things? I have experience with small op-amps, but no experience with power amplifiers. seems like it would be possible to put some kind of big transistor, in the feedback loop of an op amp circuit, but I don't really want to start at the beginning, seems like there should be an inexpensive way to make a reasonable quality amplifier. I am thinking my power output would be probably in the 10-20watt range. It seems like I should be able to take advantage of the fact that I already have a differential signal, and that it is already pre-amplified a bit, so I would prefer a solution that takes advantage of those things, however if it works out that I'm better off reducing the signal back to a single ended line level signal, then I could do that as well.
Any suggestions, or comments are appreciated
Zaaphod
I'm wanting to set up a system to send line level stereo audio signal over CAT5 cable a fairly long distance, and have multiple receivers at many points along the way. I am thinking of using TI's DRV134 for the Transmitter, and INA137 for the receiver(s) it seems these are matched to each other, since the DRV134 has a gain of 2 and the INA137 has a gain of 0.5. so the two together would produce an overall gain of 1. I like the idea of the voltage on the line being a little higher, seems like noise would affect it less, and it would travel better, and would support multiple receivers better. I am wondering if I should take this idea a little further, and put an op-amp buffer in front of the DRV134, and set it to a gain of 6 or so, then another one set for 1/6 after the INA137, this would boost my line level to something like a maximum of 24V Peak to Peak, Which should be fine with a +/- 15V power supply.
Does anyone have any experience with doing something like this? I want to have a quite a few receivers, perhaps 40 or more, it doesn't seem like multiple receivers should be a problem since the DRV134 can output 85mA and each INA137 has an input impedance of 24K Ohms, seems to me that with a 24V Peak to Peak signal, going into a 24K Ohm load would only draw 1mA, so I should be able to drive 85 receivers with one transmitter. I don't know what the cable impedance is, but it seems I should be able to drive more than 40 receivers anyway. I am also wondering if it needs to be one long cable run with drops along the way, or if I could branch off several times with just the one driver running it all. I would us a lot less cable if I could branch off.
What is the recommended power supply for this kind of thing. at first I thought I could just buy a cheap switching power supply to do the job, but I can't seem to find one that outputs +/- 15V, and with switchers, you can't just put 2 of them together, because they have one end tied directly to the power line, and if you try to connect the + output of one to the - output of the other, it would short out and go POOF. Does anyone have a good way to build my own switching power supply? Or should I just stick with the old center tap transformer going into bridge rectifier, charge a few ridiculous capacitors, then regulate the output with a 7815 and a 7915? I wouldn't need very big transformers, and it would be easy to build, I just notice almost all consumer electronics have dropped the transformer and now have switching power supplies, presumably for better efficiency and lower cost to produce, so I'm wondering if DIY electronics have got into that yet.
Lastly, does anyone have a suggestion on a power amplifier to run speakers from the output of the receivers? It seems to me that since I already have a balanced signal, I could perhaps put two INA137's, with the polarity reversed, and end up with two opposing outputs that I could easily feed each into a power amplifier in a push-pull arrangement. I am also wondering if I should reduce the signal back down to line level, just to be amplifying it again, or if I should just leave it and not amplify it as much as I would need to from line level. I am thinking reducing it then amplifying it a second time would only create extra distortion that I shouldn't need to bother with. is there such thing as a big op-amp chip with power transistors I could bolt to a heat sink and set the gain to whatever I want with feedback resistors? is that over simplifying things? I have experience with small op-amps, but no experience with power amplifiers. seems like it would be possible to put some kind of big transistor, in the feedback loop of an op amp circuit, but I don't really want to start at the beginning, seems like there should be an inexpensive way to make a reasonable quality amplifier. I am thinking my power output would be probably in the 10-20watt range. It seems like I should be able to take advantage of the fact that I already have a differential signal, and that it is already pre-amplified a bit, so I would prefer a solution that takes advantage of those things, however if it works out that I'm better off reducing the signal back to a single ended line level signal, then I could do that as well.
Any suggestions, or comments are appreciated
Zaaphod
Now that' a real challenge Mr Beeblebrox.
You've done your homework, for sure. I've never tried to drive that many destinations. Close, like maybe 30, but that was with a lot of drivers.
By the numbers it looks like the DRV134 could do it, but my gut tells me otherwise. Just how much cable will there be? You can figure up the capacitive load of the CAT5 to what you get. Will one DRV134 be happy driving all that? It's a strong chip, I've used it. But never in anything so ambitious. Does Edcor have anything that will do this?
The power supply wouldn't be a big deal. You might even find a linear +/-15V already to go. Check the surplus houses.
You've done your homework, for sure. I've never tried to drive that many destinations. Close, like maybe 30, but that was with a lot of drivers.
By the numbers it looks like the DRV134 could do it, but my gut tells me otherwise. Just how much cable will there be? You can figure up the capacitive load of the CAT5 to what you get. Will one DRV134 be happy driving all that? It's a strong chip, I've used it. But never in anything so ambitious. Does Edcor have anything that will do this?
The power supply wouldn't be a big deal. You might even find a linear +/-15V already to go. Check the surplus houses.
I'm guessing I would have a maximum of 2000 feet of cable. I've seen in some earlier posts where much longer cables were used.
Here's some math on the capacitance: From wikiedia I see Cat5e has a capacitance of 52pF/meter so if I had 2000feet that would be 609.6 meters
609.6 * 52 = 31699pF or 0.031μF
If I am reading the spec sheet correctly on the DRV134, it states Load Capacitance for Stable Operation, CL is 1μF
So I am thinking cable capacitance shouldn't be an issue.. unless I am understanding something wrong here.
I am thinking that if that is too many receivers, or too long of a cable, I could run multiple lines, perhaps 2 transmitters running 20 receivers each? I would need more wire to do that. I could also perhaps use op-amp buffers on the inputs of the receivers to raise the input impedance.. I would think if I did that the receivers would each have hardly any affect on the cable at all. a last option would be to put a power amplifiers like LM1875 on the differential outputs of the DRV134, then I could send quite a lot of power over the cable.. higher voltage, and higher current capability, but I would like to do it without additional amplifiers as I would be adding more distortion.
I bought a few DRV134s and some INA137s to mess around with. I need to start somewhere, so I figure I would make a prototype of the basic circuit and see how it performs. I might even try to purposefully induce noise in the cable to see how well the system rejects it. Perhaps I should start a new thread for my experiments in multiple receivers off one transmitter
Here's some math on the capacitance: From wikiedia I see Cat5e has a capacitance of 52pF/meter so if I had 2000feet that would be 609.6 meters
609.6 * 52 = 31699pF or 0.031μF
If I am reading the spec sheet correctly on the DRV134, it states Load Capacitance for Stable Operation, CL is 1μF
So I am thinking cable capacitance shouldn't be an issue.. unless I am understanding something wrong here.
I am thinking that if that is too many receivers, or too long of a cable, I could run multiple lines, perhaps 2 transmitters running 20 receivers each? I would need more wire to do that. I could also perhaps use op-amp buffers on the inputs of the receivers to raise the input impedance.. I would think if I did that the receivers would each have hardly any affect on the cable at all. a last option would be to put a power amplifiers like LM1875 on the differential outputs of the DRV134, then I could send quite a lot of power over the cable.. higher voltage, and higher current capability, but I would like to do it without additional amplifiers as I would be adding more distortion.
I bought a few DRV134s and some INA137s to mess around with. I need to start somewhere, so I figure I would make a prototype of the basic circuit and see how it performs. I might even try to purposefully induce noise in the cable to see how well the system rejects it. Perhaps I should start a new thread for my experiments in multiple receivers off one transmitter
Sounds like you are off to a good, solid start. Agreed, with that small a capacitance, it should not be a problem.
I would not raise the impedance much in the receiver end. There is a reason the telecoms use 600 ohm lines. Lower impedance means lower noise. Will be glad to see what you find in your experiments. Keep us up to date, please.
I would not raise the impedance much in the receiver end. There is a reason the telecoms use 600 ohm lines. Lower impedance means lower noise. Will be glad to see what you find in your experiments. Keep us up to date, please.
I made a prototype circuit on a solderless breadboard using the DRV134. I found a switching power supply with +15 and -15 outputs. I ran the output into a TPA3122D2 configured as a mono BTL amp, and it actually worked... but for some reason, my DRV134 got really HOT. I was not expecting it to get hot at all, and I can't figure out what is causing it. I used the schematic on the spec sheet precisely and my 'cable' length was only 4" (was just to get something started) my amp sounded great, and the amp chip did not even get slightly warm, but my DRV134 was just hot all the time. The TPA3122D2 requires capacitor coupling the inputs, and it also requires a single supply, so I used a separate 24V swtiching supply for the amp chip.
One other thing I was not expecting.. I had to tie my grounds together on the two power supplies to get it to work. I would like to not have to do this. Seems like the whole point of a differential system is to send the signal down a long cable without the ground. The TPA3122D2 is really just two amplifiers, by bridging it, I need one input inverted, but I am wondering if I wanted to run the amp at the end of a long cable, where the differential signal was floating (which it would be if I was using a balun transformer) what is the best way to get it to work with this amplifier? can I just connect a few load resistors from the each differential signal to the amplifiers ground? or will I need something like an op amp buffer? I don't want to over complicate things, and it would be nice to stick to a single supply on the amplifier side. I am wondering if something like the TPA3112D1, which is a mono BTL amp with differential inputs would have this problem, or if it would work with a floating differential signal. I'm hoping for an easy solution with the TPA3122D2 however, because it's a 20pin DIP while all the other ones are surface mount.
One other thing I was not expecting.. I had to tie my grounds together on the two power supplies to get it to work. I would like to not have to do this. Seems like the whole point of a differential system is to send the signal down a long cable without the ground. The TPA3122D2 is really just two amplifiers, by bridging it, I need one input inverted, but I am wondering if I wanted to run the amp at the end of a long cable, where the differential signal was floating (which it would be if I was using a balun transformer) what is the best way to get it to work with this amplifier? can I just connect a few load resistors from the each differential signal to the amplifiers ground? or will I need something like an op amp buffer? I don't want to over complicate things, and it would be nice to stick to a single supply on the amplifier side. I am wondering if something like the TPA3112D1, which is a mono BTL amp with differential inputs would have this problem, or if it would work with a floating differential signal. I'm hoping for an easy solution with the TPA3122D2 however, because it's a 20pin DIP while all the other ones are surface mount.
How's it going, Zaaph? Any progress?
I'm looking at doing something very similar for PA use. Basically, I'd like a set of external line drivers for XLR line level audio and DMX data. DMX is a serial (digital) protocol with a voltage around 5-12v, so I'm curious if this will work without inducing noise on typical 50-200' spans.
I'm just starting my research, so I have nothing to offer yet.
This thread has given me some food for thought, though.
I'm looking at doing something very similar for PA use. Basically, I'd like a set of external line drivers for XLR line level audio and DMX data. DMX is a serial (digital) protocol with a voltage around 5-12v, so I'm curious if this will work without inducing noise on typical 50-200' spans.
I'm just starting my research, so I have nothing to offer yet.
This thread has given me some food for thought, though.I've been tinkering with it on and off.. I have a few ideas in the works, one is to send balanced line level audio down the utp line, the other is to send a pwm digital signal down the utp to a power buffer. see the thread on this idea here:
http://www.diyaudio.com/forums/analog-line-level/196993-floating-differential-signal.html
you may find it interesting.. the idea is to generate a pwm signal, then run it down the utp to a opto isolator, then just power buffer that and run my speaker with it. if you did something like this with your digital signal, you could send a higher voltage down the cable, maybe 24v or more.. make it a balanced signal, then use the opto isolator to run whatever voltage you need on the other end.. should work, the higher voltage balanced signal would travel better and be less affected by noise. I would use something like the opto isolator I show in my circuit, but only hook up on of the photo transistors.. having the second LED in the circuit would keep the impedance balanced on your utp line.
my solderless breadboard is useless for prototyping these amplifiers, mainly because my inductors won't stay in, the leads are too big and not exactly the right space, and they also get a bit warm. So I am working on printed circuit boards. I have a cnc router, and I milled a prototype board with it, but I tried it with an endmill and it came out pretty rough.. so I just ordered the proper engraving bits, that should help a lot, also got some proper carbide drill bits as well. once I get the amp to be something reliable, I can get back to testing driving it with a balanced signal over utp.
I was noticing just recently that the spindle on my cnc router generates a HUGE amount of EMI.. it's a high frequency spindle that runs at 360 Hz, 7HP, and it's PWM is set at 10KHz.. so one of my tests will be to loop a really long UTP cable around my machine while it's running and see how good it rejects THAT noise.
I have done one thing with utp you may find interesting.. I ran a telephone signal down one of the pairs of a long ethernet cable (maybe around 100ft or so) , my ethernet is only 100Mbp/s so it only uses 2 of the 4 pairs.. I needed a phone line at the same location so instead of running another cable, I just picked an unused pair and broke it out and hooked it up. the 70+ v phone line signal does not affect my digital ethernet signal in the slightest, and I pick up no noise on my phone line either. I can pick up noise from the spindle I mentioned from the microphone input on my computer, but it is not picked up on the phone line, and that is a much longer run that does run right past the machine.. yet the phone line is perfectly quiet.
http://www.diyaudio.com/forums/analog-line-level/196993-floating-differential-signal.html
you may find it interesting.. the idea is to generate a pwm signal, then run it down the utp to a opto isolator, then just power buffer that and run my speaker with it. if you did something like this with your digital signal, you could send a higher voltage down the cable, maybe 24v or more.. make it a balanced signal, then use the opto isolator to run whatever voltage you need on the other end.. should work, the higher voltage balanced signal would travel better and be less affected by noise. I would use something like the opto isolator I show in my circuit, but only hook up on of the photo transistors.. having the second LED in the circuit would keep the impedance balanced on your utp line.
my solderless breadboard is useless for prototyping these amplifiers, mainly because my inductors won't stay in, the leads are too big and not exactly the right space, and they also get a bit warm. So I am working on printed circuit boards. I have a cnc router, and I milled a prototype board with it, but I tried it with an endmill and it came out pretty rough.. so I just ordered the proper engraving bits, that should help a lot, also got some proper carbide drill bits as well. once I get the amp to be something reliable, I can get back to testing driving it with a balanced signal over utp.
I was noticing just recently that the spindle on my cnc router generates a HUGE amount of EMI.. it's a high frequency spindle that runs at 360 Hz, 7HP, and it's PWM is set at 10KHz.. so one of my tests will be to loop a really long UTP cable around my machine while it's running and see how good it rejects THAT noise.
I have done one thing with utp you may find interesting.. I ran a telephone signal down one of the pairs of a long ethernet cable (maybe around 100ft or so) , my ethernet is only 100Mbp/s so it only uses 2 of the 4 pairs.. I needed a phone line at the same location so instead of running another cable, I just picked an unused pair and broke it out and hooked it up. the 70+ v phone line signal does not affect my digital ethernet signal in the slightest, and I pick up no noise on my phone line either. I can pick up noise from the spindle I mentioned from the microphone input on my computer, but it is not picked up on the phone line, and that is a much longer run that does run right past the machine.. yet the phone line is perfectly quiet.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.