OT: "Line level"
Locked
 
|
When an input to, or output from an audio device is specified as being "line
level", what exactly does that mean? Is it a specific RMS voltage perhaps? Its a term that I've heard bandied about over the years but either never knew, or have forgotten the precise meaning. Bill K3UJ _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
At 07:44 PM 7/1/2004, [hidden email] wrote...
>When an input to, or output from an audio device is specified as being "line >level", what exactly does that mean? Is it a specific RMS voltage perhaps? >Its a term that I've heard bandied about over the years but either never knew, >or have forgotten the precise meaning. A "line out" signal is one which provides a nominal .775 V RMS signal (1 mw into 600 ohms = 0 dBm). Peak level is usually +4 dBm, or about 1.23 volts. It has nothing to do impedance, where it's connected internally, or whether or not there's a means of level adjustment (all things sometimes incorrectly associated with the term). Any signal source which provides, or can be adjusted to provide, that level output can properly be called a "line out." The term "line" originates in telephony, and later broadcast audio (here's a reference: http://www.richardhess.com/be/aes-80.htm ). Mike _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Able2fly
Line level for "pro" equipment is generally considered to be a nominal (average) level of +4 dBu,
where 0 dBu = 0.78 volts rms. For typical, uncompressed speech, the rms value of a sine wave equal to the peaks is usually about 10 dB higher than the average level. Compression or limiting will generally reduce that difference to about 6 dB. Translating that to volts, +4 dBu is about 1.23 volts, 6 dB more is 2.5 volts, 10 dB above +4 is 3.9 volts. The key number there is the peak value of the waveform at clip, which is 1.414 x 3.9 volts = 5.5 volts. So an output stage (or input stage) that handles line level would expect to see a peak to peak swing of 11 volts with a nominal line level signal. But VU meters often swing against the peg, so we need some "headroom" -- another 6-10 dB "just in case." 6 dB more is double the voltage, so that's 22 volts peak to peak. For consumer (home) equipment, the "nominal" level is roughly 300 mV, with peaks of 1-2 volts. The same math applies, except that virtually all pre-recorded music has already been subjected to considerable peak limiting and compression, so "headroom" is less important. BTW -- forget all that ancient stuff about 600 ohms. Pro audio hasn't used a 600 ohm reference for at least four decades. 600 ohms came from the characteristic impedance of telegraph lines between cities (spaced at a foot or so) where lines were long enough that they needed to be treated as transmission lines. Audio lines are almost never that long (4000 ft or more). Besides, the characteristic impedance of audio cable is on the order of 60-80 ohms (do the math on conductor size and spacing), so if termination WERE used, that's the value that would be required! Modern audio gear has a low (50-100 ohm) output impedance and a high (10K typical) input impedance. Mic input stages have a much lower input impedance (1K typical) to maximize signal to noise. Jim Brown K9YC http:audiosystemsgroup.com On Thu, 1 Jul 2004 19:44:45 EDT, [hidden email] wrote: >When an input to, or output from an audio device is specified as being "line >level", what exactly does that mean? Is it a specific RMS voltage perhaps? >Its a term that I've heard bandied about over the years but either never knew, >or have forgotten the precise meaning. _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
At 11:32 PM 7/1/2004, Jim Brown wrote...
>BTW -- forget all that ancient stuff about 600 ohms. Pro audio hasn't used a 600 ohm reference for at >least four decades. That may perhaps be true for a limited range of "pro audio" applications, but 600 ohm audio transmission lines are still VERY common both in telephone and PA applications. If you forget that stuff, you WILL end up with real problems. _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Able2fly
Jim, in domestic audio equipment I respect your knowledge on the subject,
but for professional communications equipment where the present digital systems break down to an analog voice channel, 600 ohms is still very much the standard. The problem came with working with different countries was that each authority had its own standards for line levels external to the comms system. When the particular company I worked for until retirement two years back migrated to the now defacto standard of SDH they made a wise decision to change to analog channel levels of 0dBm, TX/RX at test tone levels. This made life a whole lot easier when maintaining the communication systems. It was also now possible to run data links at a higher level rather than the previously accepted figure of -13dB down on test tone which had roots in the old analog microwave systems channel loading effects. After saying all this little use is now made of analog voice channels with most of the links coming through as direct digital links to telephone switches, WAN and LAN systems and the like in 2Mb increments. No 600 ohms audio is not dead yet, at least in the professional communications environment. Bob, G3VVT K2 #4168 _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Mike S-8
On Fri, 02 Jul 2004 06:29:33 -0400, Mike S wrote:
>At 11:32 PM 7/1/2004, Jim Brown wrote... >>BTW -- forget all that ancient stuff about 600 ohms. Pro audio hasn't used a 600 ohm reference for at >>least four decades. > >That may perhaps be true for a limited range of "pro audio" applications, >but 600 ohm audio transmission lines are still VERY common both in >telephone and PA applications. If you forget that stuff, you WILL end up >with real problems. I cannot speak with authority for telco practice, but I am very much up to date on pro audio practice -- I am a member of the Standards Committee of the Audio Engineering Society, and I have recently been an invited speaker on EMC issues to AES, IEEE, and SBE (the Society of Broadcast Engineers). It is not been common practice (or good engineering practice) to drive from a 600 ohm source or terminate with a 600 ohm load for at least 40 years. Yes, broadcast facilities and equipment were built that way 40 years ago, but they have not been built that way for many years. Richard Hess, then at ABC, published an excellent paper on that topic around 1970. Some very old tape recorders and long obsolete passively matched telco interfaces (for example, an equalized dry pair) required 600 ohm (or 150 ohm) terminations to work properly, or for their level matching to be correct. Both practices have been obsolete for many years. In fact, as long ago as 1975, it was hard to find anyone at our local telco who knew what an equalized line was! Some manufacturers of equipment who "don't get it" continue to talk about their equipment as having 600 ohm outputs, but if you measure their output impedance, your instrumentation will tell you they are 50-100 ohms. A quick look at the schematic will tell you the same thing. Further, the output stages of most pro equipment are NOT designed to drive a load. 600 ohms is more of a worst case design spec -- that is, the most current that the output stage can provide -- and the performance of the output stage will be better (a dB or two more headroom, less distortion on peaks) if it looks into the 10K load that is the IEC standard. Jim Brown K9YC http://audiosystemsgroup.com _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by G3VVT
I believe you may be confusing voltage levels with impedance. 600 ohms
has NOTHING to do with it beyond defining the voltage present in a circuit where levels are defined in dBm. dBm means dB with respect to 1 mW, where the impedance must also be defined. dBu, the current most common reference in pro audio, is defined as the voltage with respect to 0.78 volts, independent of the circuit impedance. Of course, 0.78V corresponds to 0 dBm in 600 ohms. There are also issues about how digital guys think about level and how analog guys think about level. Digital folks tend to think of absolute maximum peak levels -- all the bits on, digital clip. This corresponds roughly to 100% modulation of an AM transmitter. But audio levels are DEFINED as the RMS value of a waveform. That 13 dB difference you are talking about is the approximate difference between RMS and peak of an uncompressed audio waveform. Both ways of thinking/talking about it are important, but you absolutely must understand the difference. Jim Brown K9YC http://audiosystemsgroup.com --Original Message Text--- From: [hidden email] Date: Fri, 2 Jul 2004 07:30:49 EDT 600 ohms is still very much the standard. _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Jim Brown-10
G'day,
My interpretation of the point that G3VVT was making is that a blanket dismissal of the 600 ohm reference is hazardous depending upon the circumstances. I also know nothing about pro audio, however, like G3VVT I have had a long and varied career in telecommunications. 600 ohms is the general impedance of 0.5mm diameter conductor poly insulated twisted pair telephone cable at audio frequencies. This drops to about 150 ohms at T1/E1 frequencies over the same pair. All the telco test equipment I have worked with at audio level offers 600 ohm source and terminations as well as high impedance bridging. AF attenuators were all 600 ohm balanced input/output. Impedances in transmission systems commonly change along the line. In our first earth station, built 20 years ago, we went from 600 ohm input to the PCM channel equipment, from there 75 ohm to the up/down converters and 50 ohm for the rest of the SHF link. It was essential to make sure you had the right matching transformers in your test set-up when making measurements. The USA and Europe has a history of doing things differently, a totally different coding law for PCM, different baseband data rates and hierarchy..... and on. Standards are great, so many to choose from! Regards, Mike VP8NO (G3VUI) Ex. GPO/PO/Telecomm and C&W ----- Original Message ----- From: "Jim Brown" <[hidden email]> To: "Elecraft List" <[hidden email]> Sent: Friday, July 02, 2004 10:26 AM Subject: Re: [Elecraft] OT: "Line level" > On Fri, 02 Jul 2004 06:29:33 -0400, Mike S wrote: > > >At 11:32 PM 7/1/2004, Jim Brown wrote... > >>BTW -- forget all that ancient stuff about 600 ohms. Pro audio hasn't used a 600 ohm > reference for at > >>least four decades. > > > >That may perhaps be true for a limited range of "pro audio" applications, > >but 600 ohm audio transmission lines are still VERY common both in > >telephone and PA applications. If you forget that stuff, you WILL end up > >with real problems. _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Jim Brown-10
>--Original Message Text--- >From: [hidden email] >Date: Fri, 2 Jul 2004 07:30:49 EDT > >600 ohms is still very much the standard. At 10:36 AM 7/2/2004, Jim Brown wrote... >I believe you may be confusing voltage levels with impedance. Analog phone circuits are indeed transmission lines. Try to feed a hybrid with an impedance mismatch and you'll get echo and sidetone problems. Your reference to "pro audio" is a bit parochial - telco is the largest "pro audio" industry in the world, whether measured by number of employees, consumers, amount of revenue, whatever. Broadcast, studio and stage likely follow in that order. You may be expert in your corner of "pro audio," but it's a much bigger world, and true 600 ohm audio transmission lines are still quite common throughout the world. The "line" in "line level" refers to telephone lines. Any other use is a misnomer - a broadcast engineer can call a 5v p-p signal "line level," but that usurpation of a well established term doesn't make it correct. _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
On Fri, 02 Jul 2004 12:47:44 -0400, Mike S wrote:
>Analog phone circuits are indeed transmission lines. ONLY if they are long enough between electronic output and electronic input to be transmission lines. Calculate the wavelength (in the cable of interest) at the highest frequency of interest. At 6 kHz, 1/20 wavelength in a 0.66 velocity factor cable (in the ballpark for most real cable used for audio) is nearly one mile; at 20 kHz it is nearly one half mile. At 3 kHz, the limit of baseband audio on POTS, 1/20 wavelength is nearly two miles. A line must be 1/20 wavelength at the frequency of interest for transmission line effects to be just perceptible. If a line is less than 1/20 wavelength (or at frequencies where it is less than a wavelength), it can be completely characterized by a simple lumped parameter model. That is, all of its inductance in one series L, all of its parallel capacitance as one C, and all of its wire resistance as one R, and all of its leakage resistance as one parallel R. One of the most important limiting factors on real audio lines (including telephone lines) is how much parallel capacitance that the output stage can drive. Real lines typically have 40-60 pF/ft. It is not unusual for a line to be long enough for the capacitive reactance to fall well below 600 ohms. When this happens, there can be distortion as the output stage clips prematurely. And the characteristic impedance of a transmission line is defined by the physical construction of the wire, NOT a paper standard. The characteristic impedance of virtually all cable used to carry any form of audio signals is on the order of 50-100 ohms. Nothing you can write down as a "standard" or a practice will change that physical reality. See any EE text on transmission lines, the ARRL Handbook, or the ARRL Antenna Book. Modern telephone lines (from a central office to a home) are NOT 600 ohm lines because they don't use 600 ohm cable. So calling any audio line, including telephone line that isn't a pair of open wire spaced at something on the order of a foot between insulators a transmission line is simply wrong. Jim Brown Audio Systems Group, Inc. Chicago http://audiosystemsgroup.com _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
Hi
First of all I am not an expert on telephone systems or "pro audio" although I have worked with both a little from time to time. With that said it seems to me I read somewhere that 600 ohm in the telephone system came about because that was the internal resistance of the old carbon microphones used in all telephones up through the early 60's. This resistance was also needed in conjunction with the off hook current to activate the holding relay in the central office when the phone was off hook. Later phones with amplified microphones used transformers or IC's matched to 600 ohm impedance so the central office equipment was compatible. Perhaps this is where the 600 ohm standard in audio originated. I have a HP 200C audio generator and it has a 600 ohm output impedance and I have seen many other audio generators that also have 600 ohm outputs. In consumer and pro audio the line levels that I remember used on most amplifiers was 47K input impedance and about .5 to 1 volt RMS for full rated output. Likely 0 dbm. Of course I have not messed with this for many years back when many amplifiers had 6L6's and big output transformers so the memory may be failing. I also checked the specs on a high end sound card that I have and the line levels are +4 dbm or -10dbm jumper selectable. So to answer the original question that started this thread, I assume to know what level to send to a sound card for PSK31 and such, it looks like 0 dbm to +4 dbm is about right when using the line inputs. If you are using a meter about .78 volts RMS or looking at it with a scope 2.2 volts pk to pk would be 0 dbm. The main thing is to not over drive the sound card and cause clipping but supply enough drive so the signal to noise is minimum. If any of this wrong any of you experts feel free to correct it <G> Don Brown KD5NDB _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
On Fri, 2 Jul 2004 15:37:33 -0500, Don Brown wrote:
>I read somewhere that 600 ohm in the telephone system came about >because that was the internal resistance of the old carbon microphones The more likely scenario is that the carbon mics were made to whatever spec Ma Bell needed to make their phone work well. The carbon mic in a Ma Bell phone did NOT hang directly across the line. Rather there was a network called a hybrid (it included a transformer) that hung across the line and combined the mic with the earphone to do what became known as "sidetone cancellation." On a phone line, the local signal is far stronger (on the order of 20 dB) than the distant signal because the distant signal is attenuated by the line. The sidetone network reduces the level of the local signal in the earphone so that the user hears him/herself, but not at a lot higher level than the caller . 600 ohms originally comes from the old intercity transmission lines. >I also checked the specs on a high end sound card that I have and the line levels are >+4 dbm or -10dbm jumper selectable. Yes, and they really mean dBu, whether they say it or not. Most spec sheets are written by marketing types, not engineers, and many are wrong. >So to answer the original question that started this thread, I assume to know what >level to send to a sound card for PSK31 and such, it looks like 0 dbm to +4 dbm is >about right when using the line inputs. Most computer inputs are much closer to consumer levels than pro levels. You're lucky to have a pro sound card with switchable level matching. I have one on one of my machines, but the others are the plain consumer cards. >If you are using a meter about .78 volts RMS or looking at it with a scope 2.2 volts pk >to pk would be 0 dbm. The main thing is to not over drive the sound card and cause >clipping but supply enough drive so the signal to noise is minimum. You got it right, but you really need to look at it with a scope to see the peaks. That's because what really counts is keeping it out of clip. I recently hooked up my computer to generate PSK31. I was running the sound card close to full output, and padding it down before going into the rig's mic input. I got good level matching, but some guys on the Top Band list suggested that the sound card might be generating some IM near its max output, even though it wasn't clipping. Always trying to learn new stuff, I hooked up some fancy instrumentation and learned that they were right. I dropped the output level of the rig about 6 dB and increased the input gain of the radio, and the IM produced by the sound card dropped from roughly - 30 dB re carrier to about -50 dB re: carrier. If you spend any time around the PSK31 part of the band, you'll see some occasional signals that are producing significant distortion sidebands. Those same guys observed that one of the good reasons for running a rig well below max output on PSK31 is to reduce the IM that is produced. I haven't seen any data for a rig's IM vs. power output, and I haven't measured my rig, but I suspect they are correct. Many rigs, including the K2 and K2/100, are rated for 3 dB less power on 100% duty cycle modes (RTTY, PSK, etc.) to prevent excessive dissipation. Jim Brown K9YC _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Jim Brown-10
At 02:47 PM 7/2/2004, Jim Brown wrote...
>On Fri, 02 Jul 2004 12:47:44 -0400, Mike S wrote: > >>Analog phone circuits are indeed transmission lines. > >ONLY if they are long enough... > At 3 kHz, the limit of baseband audio on POTS, >1/20 wavelength is nearly two miles. According to Verizon, over 20% of US subscriber loops are more than 18,000 feet long. The balance of your post continues to demonstrate a lack of familiarity with real world audio transmission lines. For instance, you're apparently unaware that loading coils are inserted on a long subscriber loops to offset the parallel capacitance. Nonetheless, throwing out non-sequiturs in defense of an incorrect statement is pointless. Your original statement was "forget all that ancient stuff about 600 ohms. Pro audio hasn't used a 600 ohm reference for at least four decades," which is demonstrably false. "Line level" is a telco term which has well established meaning. "Line level" had meaning before broadcasting or electronic sound recording even existed. You, in your response, gave an ambiguous answer of at least two significantly different values, which illustrates that the term has lost any real meaning in your environment through imprecise use and ambiguity. The original questioner deserved a correct answer. For all you know, he was trying to build a phone patch, for which your "impedance doesn't matter" response would cause him no end of grief, as would your ambiguous description of what constitutes "line level". >Modern telephone lines (from a central office to a home) are NOT 600 ohm lines >because they don't use 600 ohm cable. Your implication is incorrect. By your own statements, what matters isn't what's on paper, but what's measured. The measured "impedance is in the region of 500 to 1000 ohms" for normal telco twisted pair cable - http://www.qsl.net/vk5br/TransLines.htm Real world and by design, telco twisted pair is functionally 600 ohm. There's no doubt a Bellcore standard for all of this, too. And yes, this does matter beyond formal transmission line theory. You're used to unidirectional signal transfer (4 wire, in the telco world). A telephone uses two wires full duplex. The hybrids at each end of that line which make this possible must be impedance matched, or you get problems like echo, bad sidetone or improper levels. I'm done. My intent was to correct an obvious error, not get into a pissing match. _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Mike Harris-9
I did work on some audio systems - half a century ago - so I'm no expert. We
did use 600 ohm "line level" output for long distance lines (e.g. from a microphone mixer just off stage to the power amplifiers up in the projection booth in theatres). That might involve 200 or 300 feet of cable by the time the got from one place to another. The connection was made with a shielded and balanced line with - and this I assumed was the critical part in the 600 ohm specification - a 600 ohm center-tapped transformer at each end to suppress common-mode noise the line might pick up. I thought it was interesting that Elecraft resurrected this application specifically to avoid noise pickup by the circuit traces within the K2 carrying audio. That's why the K2 has a fully balanced audio system until the signal gets to the LM-386 output I.C. The problem there isn't external noise, it's the abundant digital noise floating around inside the K2 that can easily get into the low-level audio lines. Ron AC7AC _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Jim Brown-10
Hi
You are right I had forgotten about the hybrid. It has been about 30 years ago when I messed with this stuff and then it wasn't much. <G> Thanks for jogging my memory Don Brown ----- Original Message ----- From: Jim Brown<mailto:[hidden email]> To: Elecraft List<mailto:[hidden email]> Sent: Friday, July 02, 2004 4:16 PM Subject: Re: [Elecraft] OT: "Line level" On Fri, 2 Jul 2004 15:37:33 -0500, Don Brown wrote: >I read somewhere that 600 ohm in the telephone system came about >because that was the internal resistance of the old carbon microphones The more likely scenario is that the carbon mics were made to whatever spec Ma Bell needed to make their phone work well. The carbon mic in a Ma Bell phone did NOT hang directly across the line. Rather there was a network called a hybrid (it included a transformer) that hung across the line and combined the mic with the earphone to do what became known as "sidetone cancellation." On a phone line, the local signal is far stronger (on the order of 20 dB) than the distant signal because the distant signal is attenuated by the line. The sidetone network reduces the level of the local signal in the earphone so that the user hears him/herself, but not at a lot higher level than the caller . 600 ohms originally comes from the old intercity transmission lines. >I also checked the specs on a high end sound card that I have and the line levels are >+4 dbm or -10dbm jumper selectable. Yes, and they really mean dBu, whether they say it or not. Most spec sheets are written by marketing types, not engineers, and many are wrong. >So to answer the original question that started this thread, I assume to know what >level to send to a sound card for PSK31 and such, it looks like 0 dbm to +4 dbm is >about right when using the line inputs. Most computer inputs are much closer to consumer levels than pro levels. You're lucky to have a pro sound card with switchable level matching. I have one on one of my machines, but the others are the plain consumer cards. >If you are using a meter about .78 volts RMS or looking at it with a scope 2.2 volts pk >to pk would be 0 dbm. The main thing is to not over drive the sound card and cause >clipping but supply enough drive so the signal to noise is minimum. You got it right, but you really need to look at it with a scope to see the peaks. That's because what really counts is keeping it out of clip. I recently hooked up my computer to generate PSK31. I was running the sound card close to full output, and padding it down before going into the rig's mic input. I got good level matching, but some guys on the Top Band list suggested that the sound card might be generating some IM near its max output, even though it wasn't clipping. Always trying to learn new stuff, I hooked up some fancy instrumentation and learned that they were right. I dropped the output level of the rig about 6 dB and increased the input gain of the radio, and the IM produced by the sound card dropped from roughly - 30 dB re carrier to about -50 dB re: carrier. If you spend any time around the PSK31 part of the band, you'll see some occasional signals that are producing significant distortion sidebands. Those same guys observed that one of the good reasons for running a rig well below max output on PSK31 is to reduce the IM that is produced. I haven't seen any data for a rig's IM vs. power output, and I haven't measured my rig, but I suspect they are correct. Many rigs, including the K2 and K2/100, are rated for 3 dB less power on 100% duty cycle modes (RTTY, PSK, etc.) to prevent excessive dissipation. Jim Brown K9YC _______________________________________________ Elecraft mailing list Post to: [hidden email]<mailto:[hidden email]> You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft<http://mailman.qth.net/mailman/listinfo/elecraft> Help: http://mailman.qth.net/subscribers.htm<http://mailman.qth.net/subscribershtm> Elecraft web page: http://www.elecraft.com<http://www.elecraft.com/> _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
Re: OT: "Line level" - End of Thread
|
Administrator
|
Interesting topic and lots of good info in these postings - but we're now in
the 2nd phase of discussion where it is getting beaten to death ;-) Plus its probably a bit of email overload for many of our list readers. Let's take this off list for now :-) 73, Eric WA6HHQ Elecraft List Moderator Don Brown wrote: > Hi > > You are right I had forgotten about the hybrid. It has been about 30 years ago when I messed with this stuff and then it wasn't much. <G> > > Thanks for jogging my memory > > Don Brown _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by Mike S-8
On Fri, 02 Jul 2004 17:21:57 -0400, Mike S wrote:
>My intent was to correct an obvious error So was mine. >According to Verizon, over 20% of US subscriber loops are more than >18,000 feet long. The balance of your post continues to demonstrate a >lack of familiarity with real world audio transmission lines. For instance, >you're apparently unaware that loading coils are inserted on a long >subscriber loops to offset the parallel capacitance. I'm not unaware of it all. That technique is practical BECAUSE the line can be viewed as a lumped capacitance. 18,000 ft is still less than a tenth of a wavelength at 3 kHz in 0.66 Vf cable. I learned about such techniques in an EE circuit analysis course around 1962. Then they were called "peaking circuits" when applied to the outputs of active electronics. They form a resonance with a broad peak near or above the end of the desired response limit, then roll off more sharply than they otherwise would due to the capacitance alone because of the resonance. If you have doubt about the characteristic impedance of audio cable, try measuring some. Or look in the printed copy of a Belden catalog -- the one that is a half-inch thick (if you can find one). You will find the characteristic impedance listed for some of their twisted pair cables. They are all in the 60-80 ohm range. Or look at the some of the excellent applications notes and tutorials on the Jensen Transformers website. You will also find Vf (velocity factor) listed for some cable types. Gepco (a good small cable mfr based in Chicago) also lists this data for some products. http://www.jensentransformers.com/ Jensen still makes repeat coils like the old 111C used by Ma Bell (except that their coils have better specs than the original, thanks to the greater precision of modern manufacturing and computer-aided design). This thread is, indeed, rather far off topic, and this will be my last post to the list on it. My apologies to any who might have been bothered by the length of it. Jim Brown K9YC _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
«
Return to Elecraft
|
1 view|%1 views
| Free forum by Nabble | Edit this page |