klieb@tekigm2.MEN.TEK.COM (Kurt Liebezeit) (07/30/89)
Some questions from an audio neophyte: I recently purchased a portable tape deck that happens to have a bias trim adjustment available. The owner's manual is somewhat vague on how to use this feature; it says to monitor the tape during recording of program material with lots of high frequency sound, and adjust to my liking (it's a three head deck, by the way). I've tried this method, A-B'ing between source and tape while recording some hammered dulcimer music. At one end of the trim's range the tape was muddy compared to the source, and at the other end of the range it sounded bright. It was hard to choose a setting in the middle that was clearly the best. I thought there probably ought to be a more scientific way, and so I brought the recorder into work. A review article said that the line input expected 70mV. <Question #1: is it customary to quote rms? pk-pk?> I hooked up the line in input to a function generator, and set it for 70mV peak to peak, 440 Hz sine wave. <Question #2: should I have used a DC-blocking capacitor?> The VU meter barely registered the tone, though I could hear it easily through the monitor speaker. I adjusted the recording level to get a couple divisions showing on an oscilloscope I hooked up to the line out jack. Then I set the function generator to 12.5kHz, still at 70mV. The VU meters jumped higher, the tone got louder, and the signal filled the screen on the scope! Clearly, the recorder plus my setup did not have flat frequency response! <Question #3: the function generator has a 50 ohm output impedance, and I initially reasoned that it would look like a reasonably ideal source to the deck inputs. Perhaps the line in expects a higher output impedance source that would roll off the high frequency response? Or does the output of the line out expect some sort of equalization to be supplied by the amplifier? Why didn't 70mV produce something close to 0 VU?> Perhaps someone with experience in this area could suggest a method of adjusting the bias trim. Is it possible to measure frequency response with simpe generators and scopes? I have access to good T&M equipment thru work; I suspect that impedance matching is my problem. I'd also appreciate any pointers to magazine articles that might help. The recorder in question is a Marantz PMD430. This is my second tape deck, BTW. I'd like to use both in conjunction with my amplifier, but the amplifier has a limitation: one can dub from TAPE 1 to TAPE 2, but not vice versa. The two decks have incompatible noise reduction systems, so I can't simply dedicate one deck as the source. Radio Shack sells a little switch box that can mechanically connect either tape deck to the amp, or dub either way. There is no buffering in the box, just switches that connect the line in and line out of each deck to each other in various ways. <Question #4: are line in and line out directly compatible? The review article says that line in is 70mV, and line out is 400mV. On the face of it I would be wary that the line out of deck #1 will swamp the line in of deck #2. Is this a problem?> For convenience, I'd like to monitor the source deck through the amplifier line in jacks, while the signal also goes to the destination deck (the amplifier won't allow me to pass the signal from source to amp to destination, unfortunately). <Question #5: is it a OK to expect the source deck to drive cables and the impedances of two destinations (amplifier input and destination deck input)?> <Question #6: this whole setup will involve longer RCA-RCA cables than I am used to using. Although they are standard shielded cables, will the build- up of capacitance cause high-frequency roll-off problems? Are fancy audio cables significantly better in this respect?> Thanks in advance for any advice you may have. I suspect that most people share my ignorance in these matters, so posting may be more appropriate than e-mail. Kurt Liebezeit (views expressed are those of the author only) klieb@tekigm2.MEN.TEK.COM ...!tektronix!tekigm2.MEN.TEK.COM!klieb
durham@w2xo.UUCP (Jim Durham) (07/30/89)
> Some questions from an audio neophyte: > > I recently purchased a portable tape deck that happens to have a bias trim > adjustment available. The owner's manual is somewhat vague on how to use > this feature; it says to monitor the tape during recording of program > material with lots of high frequency sound, and adjust to my liking (it's > a three head deck, by the way). I've tried this method, A-B'ing between > source and tape while recording some hammered dulcimer music. At one end > of the trim's range the tape was muddy compared to the source, and at the > other end of the range it sounded bright. It was hard to choose a setting > in the middle that was clearly the best. > > I thought there probably ought to be a more scientific way, and so I brought > the recorder into work. A review article said that the line input expected > 70mV. > > <Question #1: is it customary to quote rms? pk-pk?> > RMS is usually quoted. > I hooked up the line in input to a function generator, and set it for 70mV > peak to peak, 440 Hz sine wave. > > <Question #2: should I have used a DC-blocking capacitor?> In most cases, it isn't necessary, but it can't hurt. Use at least .5 mfd. > > The VU meter barely registered the tone, though I could hear it easily > through the monitor speaker. I adjusted the recording level to get a > couple divisions showing on an oscilloscope I hooked up to the line out > jack. Then I set the function generator to 12.5kHz, still at 70mV. The > VU meters jumped higher, the tone got louder, and the signal filled the > screen on the scope! Clearly, the recorder plus my setup did not have > flat frequency response! > > <Question #3: the function generator has a 50 ohm output impedance, and I > initially reasoned that it would look like a reasonably ideal source to the > deck inputs. Perhaps the line in expects a higher output impedance source > that would roll off the high frequency response? Or does the output of the > line out expect some sort of equalization to be supplied by the amplifier? > Why didn't 70mV produce something close to 0 VU?> This is more than one question..you're cheating! First (3a), most freq response problems dealing with running a generator, amplifier, etc into a *higher* impedance load were in the old transformer-coupled days. Most modern gear doesn't exibit freq-response problems due to a higher impedance load, but the voltage level may rise. (3b). I've never known anyone to design anything so that you expect random things like input loading to produce the correct freq response. Things in the real world just vary too much. (3c). If you are setting up for 70mv p/p, you would get only .707 times the value rms. However, it sounds like you are *way* down. You may have a problem. Low gain and rising response vs frequence usually spells dried-out electrolytic coupling capacitor somewhere. Is this a new unit? > > Perhaps someone with experience in this area could suggest a method of > adjusting the bias trim. Is it possible to measure frequency response with > simpe generators and scopes? I have access to good T&M equipment thru work; > I suspect that impedance matching is my problem. I'd also appreciate any > pointers to magazine articles that might help. The recorder in question > is a Marantz PMD430. > Yes, you can use simple equipment, with cautions. You can probably assume that any decent scope is flat over the audio spectrum, so hook up the gen to the tape and check the level on the scope, with the scope bridged across the output of the generator to tape deck input line. Check at , say 50 hz thru 15khz. If there are any deviations from flat, then life gets more complicated. If there are, then hopefully you have a dual-channel scope. Leave one probe on the input line and put the other on the output line. Check the voltage out of the generator at 1000hz(leaving the generator connected to the tape deck input at all times). Adjust the variable input knob on the scope for a convenient reference as nearly full screen as possible. For every frequency, adjust the generator to this reference voltage. Now, any differences in the output are caused by the deck. The "rule of thumb" for bias adjustments(and this is *very* rough), is to put 700 hz into the tape deck while watching the output. Decrease the bias as much as possible(trimmer at minimum capacitance). Now, increase the bias until you go *thru* the maximum output. Continue until the output drops 1db. Now, you adjust the HF equalization for flat response. > This is my second tape deck, BTW. I'd like to use both in conjunction with > my amplifier, but the amplifier has a limitation: one can dub from TAPE 1 > to TAPE 2, but not vice versa. The two decks have incompatible noise > reduction systems, so I can't simply dedicate one deck as the source. > Radio Shack sells a little switch box that can mechanically connect either > tape deck to the amp, or dub either way. There is no buffering in the box, > just switches that connect the line in and line out of each deck to each > other in various ways. > > <Question #4: are line in and line out directly compatible? The review > article says that line in is 70mV, and line out is 400mV. On the face of > it I would be wary that the line out of deck #1 will swamp the line in of > deck #2. Is this a problem?> > Dunno, maybe 70mv is the *maximum* it will handle. If the first thing the signal hits is a pot, then you don't have a problem. > For convenience, I'd like to monitor the source deck through the amplifier > line in jacks, while the signal also goes to the destination deck (the > amplifier won't allow me to pass the signal from source to amp to destination, > unfortunately). > > <Question #5: is it a OK to expect the source deck to drive cables and the > impedances of two destinations (amplifier input and destination deck input)?> > Usually.... > <Question #6: this whole setup will involve longer RCA-RCA cables than I > am used to using. Although they are standard shielded cables, will the build- > up of capacitance cause high-frequency roll-off problems? Are fancy audio > cables significantly better in this respect?> > Most modern equipment has a pretty low output impedance. You shouldn't have any roll-offs for reasonable cable lengths. Almost all these problems are ideas held-over from the tube days. When you are working with impedances of 500 ohms or so compared to the old 500k ohms, there is a 1000/1 increase in the length of cable the system will tolerate. > Thanks in advance for any advice you may have. I suspect that most people > share my ignorance in these matters, so posting may be more appropriate > than e-mail. > > Kurt Liebezeit (views expressed are those of the author only) > klieb@tekigm2.MEN.TEK.COM ...!tektronix!tekigm2.MEN.TEK.COM!klieb Hope this helps.... Oh.. One very important point when doing record-to- playback measurements on tape gear.. keep the level down. If you try to use full recording level, you will clip the upper spectrum due to pre-emphasis. Keep the level down to at least 10db below nominal level. -Jim Durham
strong@tc.fluke.COM (Norm Strong) (08/02/89)
In article <107@w2xo.UUCP> durham@w2xo.UUCP (Jim Durham) writes: [ lots deleted] } (3c). If you are setting up for 70mv p/p, you would get only .707 times } the value rms. However, it sounds like you are *way* down. You may } have a problem. Low gain and rising response vs frequence usually } spells dried-out electrolytic coupling capacitor somewhere. Is this } a new unit? Correction: 70 mv p/p is 28mv rms. You're off by a factor of 2. -- Norm (strong@tc.fluke.com)
strong@tc.fluke.COM (Norm Strong) (08/02/89)
In article <10032@fluke.COM> strong@tc.fluke.COM (Norm Strong) writes: }In article <107@w2xo.UUCP> durham@w2xo.UUCP (Jim Durham) writes: } }[ lots deleted] } }} (3c). If you are setting up for 70mv p/p, you would get only .707 times }} the value rms. However, it sounds like you are *way* down. You may }} have a problem. Low gain and rising response vs frequence usually }} spells dried-out electrolytic coupling capacitor somewhere. Is this }} a new unit? } }Correction: 70 mv p/p is 28mv rms. You're off by a factor of 2. }-- } }Norm (strong@tc.fluke.com) Correction of correction: 25mv rms -- Norm (strong@tc.fluke.com)