And it would have to be 2 channels of fiber optic cable to represent the - and + aspects of a pure sinewave signal, so for stereo you would need 4x fiber optic cables (which these days isn't that hard, nor expensive to buy.)
Rather...
AM would be worse. No need for brainstorming. The idea is daft, as I said earlier in this thread. It is one of those ideas which fits the following criterion: anyone who knows enough to make a realistic attempt at doing it, knows enough to know that it won't work so no point in trying.
I suggest you go away and design, build and test a wideband FM receiver which meets or exceeds the current state of the art. Then you will have some idea of the difficulties in scaling performance up to interconnect standards. I have designed an FM receiver, which used the best chips available at the time, but it is nowhere near the bandwidth or dynamic range of a simple piece of wire.
I suggest you go away and design, build and test a wideband FM receiver which meets or exceeds the current state of the art. Then you will have some idea of the difficulties in scaling performance up to interconnect standards. I have designed an FM receiver, which used the best chips available at the time, but it is nowhere near the bandwidth or dynamic range of a simple piece of wire.
Lots of noise on the output of a typical cheap laser diode, especially when you start to vary the current (They usually mode hop if you even look at them funny) and are seldom single longitudinal mode across any sane modulation range, that is before we get into what happens as the die heats and cools. Most lasers are also seriously non linear in their output power as a function of diode current, and often have a smallish range between threshold and excess current causing the optical power density in the cavity to blow its front mirror off (Yes, really a common cause of LD diode failure).
Trying to maintain low noise single longitudinal mode operation gets you into tricky exotica like ECDLs and ring lasers, with a standard diode laser you are doing very well to get a 1M coherence length!
Besides, why on earth would you want a laser for this? LEDS are cheaper and it is not like you are trying to couple to a single mode fibre for a long haul link.
Why this fascination with one channel fer fibre, ever hear of prisims, there is more then one wavelength of IR you know.
Still seems to me that simple twisted pair with good drivers and recevers stomps all over fibre for a short analogue run like this at the sorts of nearly zero bandwidths needed for audio (All audio is as near to DC as makes very little difference!).....
Regards, Dan.
Trying to maintain low noise single longitudinal mode operation gets you into tricky exotica like ECDLs and ring lasers, with a standard diode laser you are doing very well to get a 1M coherence length!
Besides, why on earth would you want a laser for this? LEDS are cheaper and it is not like you are trying to couple to a single mode fibre for a long haul link.
Why this fascination with one channel fer fibre, ever hear of prisims, there is more then one wavelength of IR you know.
Still seems to me that simple twisted pair with good drivers and recevers stomps all over fibre for a short analogue run like this at the sorts of nearly zero bandwidths needed for audio (All audio is as near to DC as makes very little difference!).....
Regards, Dan.
Hi
I talked with a friend of mine yesterday about what we are trying to do. He said that he had unusually good results using a V/f converter for a limited bandwidth. He suggested using a Burr-Brown/TI chip called a VFC110. He also said that using it beyond 2MHz is a loosing battle. Yes I know we will probably blast the 160m - 1.8 to 2MHz Amateur band, but so be it. My HI-FI is more important than 200kHz of amateur radio. Besides by shielding and paying attention to circuit details we should not transmit much more than 100uW or less.
I will have to study the chip specs more carefully, to see if we can get that to work.
Still we would have to put this thing inside a local precision feedback loop to make it work to our enhanced spec.
Using this approach we would transmit the frequency over the fiber optic link, maybe using different IR wavelengths for several channels. I am in favor of using the standard "Toslink" connector and their transmitter and receiver just to stay within reasonable expense-bounds.
On the receiving end we should use the identical f/V converter as used on the transmit side. I do not yet know if the VFC110 can be used as a f/V converter as well, but knowing B-B my guess is that by "pulling some pin sideways" you can.
Back to my paper on the History of Computer Assisted Tomographic (CAT) Image generation, due on Friday.
Later-
Hans J Weedon.
I talked with a friend of mine yesterday about what we are trying to do. He said that he had unusually good results using a V/f converter for a limited bandwidth. He suggested using a Burr-Brown/TI chip called a VFC110. He also said that using it beyond 2MHz is a loosing battle. Yes I know we will probably blast the 160m - 1.8 to 2MHz Amateur band, but so be it. My HI-FI is more important than 200kHz of amateur radio. Besides by shielding and paying attention to circuit details we should not transmit much more than 100uW or less.
I will have to study the chip specs more carefully, to see if we can get that to work.
Still we would have to put this thing inside a local precision feedback loop to make it work to our enhanced spec.
Using this approach we would transmit the frequency over the fiber optic link, maybe using different IR wavelengths for several channels. I am in favor of using the standard "Toslink" connector and their transmitter and receiver just to stay within reasonable expense-bounds.
On the receiving end we should use the identical f/V converter as used on the transmit side. I do not yet know if the VFC110 can be used as a f/V converter as well, but knowing B-B my guess is that by "pulling some pin sideways" you can.
Back to my paper on the History of Computer Assisted Tomographic (CAT) Image generation, due on Friday.
Later-
Hans J Weedon.
You want to introduce crossover distortion into an interconnect?freax said:
I suggested you try to build an FM tuner, to learn about the snags. Maybe you should build a push-pull amplifier too?
I think you may find that the law disagrees. Domestic items using these frequency ranges are usually required to include nulls in emission at amateur band frequencies.HJWeedon said:
Last edited:
Hi I may not just find it I know it.
Regarding RFI, the basic spec is not to radiate more than 1uV/m at a 3m distance. There are lots of concessions in the spec to yield to lobbyists, but that is the basic class B spec. What this really means is that if you take ordinary care in assembly, and round off edges on digital timing signals the spec will in general be met. "Sky-wiring" will be suspect, but laid out on a multilayer PC board it is more probable that you will be quiet enough.
Hans J Weedon.
Regarding RFI, the basic spec is not to radiate more than 1uV/m at a 3m distance. There are lots of concessions in the spec to yield to lobbyists, but that is the basic class B spec. What this really means is that if you take ordinary care in assembly, and round off edges on digital timing signals the spec will in general be met. "Sky-wiring" will be suspect, but laid out on a multilayer PC board it is more probable that you will be quiet enough.
Hans J Weedon.
Hi Guys.
I "cheated" I did an on-line purchase of a Stereo-RCA connector input to Toslink optical A/D converter output and a complement Toslink to RCA stereo D/A converter output. I plan to measure the pairs frequency response, signal to noise ratio and distortion over the next few weeks. It is perhaps a little too much "consumerism" for you guys to do it that way but only if the signal is "junk" will I not use the pair. It will set me back perhaps as much as US$75 plus maybe 4 hours of testing. I have the appropriate test equipment in my Audio Lab (Ha Ha) downstairs I think.---- A Krohn-Hite low distortion sine generator and an "old" repaired HP distortion analyzer. I do prefer my Heath-Kit sine oscillator and the Heath-Kit Distortion analyzer though, because I put them together myself. Who can hear 0.03% distortion anyway. I will write a comprehensive report with nice curves (I think)
Wish me luck, the equipment arrives in about 3 days, but we have Thanksgiving in the way.
Hans J Weedon.
I "cheated" I did an on-line purchase of a Stereo-RCA connector input to Toslink optical A/D converter output and a complement Toslink to RCA stereo D/A converter output. I plan to measure the pairs frequency response, signal to noise ratio and distortion over the next few weeks. It is perhaps a little too much "consumerism" for you guys to do it that way but only if the signal is "junk" will I not use the pair. It will set me back perhaps as much as US$75 plus maybe 4 hours of testing. I have the appropriate test equipment in my Audio Lab (Ha Ha) downstairs I think.---- A Krohn-Hite low distortion sine generator and an "old" repaired HP distortion analyzer. I do prefer my Heath-Kit sine oscillator and the Heath-Kit Distortion analyzer though, because I put them together myself. Who can hear 0.03% distortion anyway. I will write a comprehensive report with nice curves (I think)
Wish me luck, the equipment arrives in about 3 days, but we have Thanksgiving in the way.
Hans J Weedon.
Any luck on your idea? I might be a good idea to simply use a high quality A/D & D/A combination but I feel that you shouldn't have to.
Have you considered using quad shield as a substitute for the shielding problems that you are having? Maybe even earthing only one end of the cable?
Here is my thread on the BNC + Quad Shield combination:
http://www.diyaudio.com/forums/anal...ople-using-bnc-out-there-rg6-quad-shield.html
Quad might do the trick, you never know.
Have you considered using quad shield as a substitute for the shielding problems that you are having? Maybe even earthing only one end of the cable?
Here is my thread on the BNC + Quad Shield combination:
http://www.diyaudio.com/forums/anal...ople-using-bnc-out-there-rg6-quad-shield.html
Quad might do the trick, you never know.
Last edited:
Hi
I want to do this the old-fashioned way, the Analog way. No tubes no digits, no nothing, just intrinsic solutions to the real problem.
I was thinking a lot about this long distance ~ 40 feet or 12meters of signal transport with about 100dB of signal to noise ratio. (A hundred thousand to one !!) Why do I need such huge signal to nose ratio? Maybe I don't, but that is my goal. At low level, normal listening level with NPR on FM with low volume, my wife's morning routine. The volume is set at about 3mW in the speakers. No hum should be audible with the ear to the speaker. Why? because the speaker is next to the breakfast table (tea and croissants). No hum or noise should be heard. When I really play loud at 30W into the speaker no hum or noise should be heard then either.
I assume 3mW to have about 55dB of dynamic range the range from 3mW to 30W is 40dB plus the previous 55dB equals 95dB. I need 100dB signal to noise ratio, period.
When I hook up about 12 meters of dual coax Audio (RCA) cables I can hear 60Hz related signals in the speaker at any volume, I expect that this noise should be reduced by a factor of about 40dB to be inaudible. I make the assumption that the hum and "noise" is due to the 1 Ohm resistive connection between my power amplifiers and the Audio processing center. I want to electrically increase this shield resistance by a factor of 100 to 100 Ohms at most probably 120Hz and above, because a 60Hz sine-wave is practically inaudible. The 1 Ohm resistance becomes a differential mode resistance, I can prove that.
That calls for a 100 Ohm reactance in the cable shields (100 times the resistance of 1 Ohm), reducing audible hum by a factor of 100. That calls for a 133mH inductance in the shielded Stereo Audio cables. I can achieve this by winding some shielded Audio cables on a tape-wound magnetic toroid. This toroid can not be a ferrite toroid because ferrites are difficult to get with more than 10uH per turns squared. This calls for 120 turns, which is impractical. This must be done using a High mu iron core with about 3.5uH per turn squared. That is practical because only about 20 turns of the dual Audio cable should be required.
This is obviously close to a design document, but in desperation I reach out to learn if anyone ever has been able to translate the cryptic specs on toroids to the simple trerm of uH per turn squared. I can handle those specs, but those convoluted specs used by the magnetic guys leave me in the "dust", probably intentionally so that I have to deal with the "application engineer" who obviously knows more than I after my 60 years in the industry.
Frustrated
Hans J Weedon.
I want to do this the old-fashioned way, the Analog way. No tubes no digits, no nothing, just intrinsic solutions to the real problem.
I was thinking a lot about this long distance ~ 40 feet or 12meters of signal transport with about 100dB of signal to noise ratio. (A hundred thousand to one !!) Why do I need such huge signal to nose ratio? Maybe I don't, but that is my goal. At low level, normal listening level with NPR on FM with low volume, my wife's morning routine. The volume is set at about 3mW in the speakers. No hum should be audible with the ear to the speaker. Why? because the speaker is next to the breakfast table (tea and croissants). No hum or noise should be heard. When I really play loud at 30W into the speaker no hum or noise should be heard then either.
I assume 3mW to have about 55dB of dynamic range the range from 3mW to 30W is 40dB plus the previous 55dB equals 95dB. I need 100dB signal to noise ratio, period.
When I hook up about 12 meters of dual coax Audio (RCA) cables I can hear 60Hz related signals in the speaker at any volume, I expect that this noise should be reduced by a factor of about 40dB to be inaudible. I make the assumption that the hum and "noise" is due to the 1 Ohm resistive connection between my power amplifiers and the Audio processing center. I want to electrically increase this shield resistance by a factor of 100 to 100 Ohms at most probably 120Hz and above, because a 60Hz sine-wave is practically inaudible. The 1 Ohm resistance becomes a differential mode resistance, I can prove that.
That calls for a 100 Ohm reactance in the cable shields (100 times the resistance of 1 Ohm), reducing audible hum by a factor of 100. That calls for a 133mH inductance in the shielded Stereo Audio cables. I can achieve this by winding some shielded Audio cables on a tape-wound magnetic toroid. This toroid can not be a ferrite toroid because ferrites are difficult to get with more than 10uH per turns squared. This calls for 120 turns, which is impractical. This must be done using a High mu iron core with about 3.5uH per turn squared. That is practical because only about 20 turns of the dual Audio cable should be required.
This is obviously close to a design document, but in desperation I reach out to learn if anyone ever has been able to translate the cryptic specs on toroids to the simple trerm of uH per turn squared. I can handle those specs, but those convoluted specs used by the magnetic guys leave me in the "dust", probably intentionally so that I have to deal with the "application engineer" who obviously knows more than I after my 60 years in the industry.
Frustrated
Hans J Weedon.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.