Line level balun
in [Live Sound]
|
From: Phil Allison on 12 Dec 2006 18:14 "Rupert" > A properly impedance balanced output is what's primarily required to > for CMR work. If you tie the ground of an unbalanced output directly to > pin 3 and the hot to pin 2, the relative impedance of the ground to > cold path is near 1 ohm whereas the impedance to hot is whatever the > circuit is designed at, say 50 ohms. This means that noise common to > both wires feeding the balanced input is not going to be of the matched > amplitude an will therefore not cancel when the cold polarity is > flipped and the signals are added. There may be some reduction of > noise, but not in the order of matched impedance hot and cold. > ** I already covered this problem days ago in the same thread: " That only works OK with transformer balanced inputs ( ie very high CMRR and non susceptible to RFI) which are rare as hen's teeth in modern audio. With typical active balanced line level inputs ( ie limited CMRR) - ground loop hum will probably still be audible PLUS you are just asking for horrible RFI problems as the cable screen is earthed only at the send end. " ........ Phil
From: Rupert on 12 Dec 2006 19:10 Phil Allison wrote: > "Rupert" > > > > A properly impedance balanced output is what's primarily required to > > for CMR work. If you tie the ground of an unbalanced output directly to > > pin 3 and the hot to pin 2, the relative impedance of the ground to > > cold path is near 1 ohm whereas the impedance to hot is whatever the > > circuit is designed at, say 50 ohms. This means that noise common to > > both wires feeding the balanced input is not going to be of the matched > > amplitude an will therefore not cancel when the cold polarity is > > flipped and the signals are added. There may be some reduction of > > noise, but not in the order of matched impedance hot and cold. > > > > > ** I already covered this problem days ago in the same thread: > > > " That only works OK with transformer balanced inputs ( ie very high CMRR > and non susceptible to RFI) which are rare as hen's teeth in modern audio. > > With typical active balanced line level inputs ( ie limited CMRR) - ground > loop hum will probably still be audible PLUS you are just asking for > horrible RFI problems as the cable screen is earthed only at the send end. > " > > > > > ....... Phil I saw your excellent post. Mr. Krueger has stated in the past that his method was virtually as effective as a balanced output, so I was providing an explanation to the contrary. That method may work in some situations, but it's not good practice. I'm sure he will disagree. Rupert
From: Phil Allison on 12 Dec 2006 19:33 "Rupert" > > Phil Allison wrote: > >> >> >> > A properly impedance balanced output is what's primarily required to >> > for CMR work. If you tie the ground of an unbalanced output directly to >> > pin 3 and the hot to pin 2, the relative impedance of the ground to >> > cold path is near 1 ohm whereas the impedance to hot is whatever the >> > circuit is designed at, say 50 ohms. This means that noise common to >> > both wires feeding the balanced input is not going to be of the matched >> > amplitude an will therefore not cancel when the cold polarity is >> > flipped and the signals are added. There may be some reduction of >> > noise, but not in the order of matched impedance hot and cold. >> > >> >> >> ** I already covered this problem days ago in the same thread: >> >> >> " That only works OK with transformer balanced inputs ( ie very high >> CMRR >> and non susceptible to RFI) which are rare as hen's teeth in modern >> audio. >> >> With typical active balanced line level inputs ( ie limited CMRR) - >> ground >> loop hum will probably still be audible PLUS you are just asking for >> horrible RFI problems as the cable screen is earthed only at the send >> end. >> " >> > > I saw your excellent post. ** My gawd - someone here actually comprehends them !!! > Mr. Krueger has stated in the past that his > method was virtually as effective as a balanced output, so I was > providing an explanation to the contrary. ** Arny's method ?? The resident " Eeysore " was the utter fool who claimed that ( two conductor) unbalanced outputs were as good as regular three conductor, balanced ones. > That method may work in some > situations, but it's not good practice. I'm sure he will disagree. ** It is important to know that an " impedance balanced" output CANNOT be created from typical unbalanced one by just adding one resistor in the ground circuit. It has to be designed from the very start to work in that mode. The output HOT signal must have a precisely fixed source impedance, with virtually no amplitude variation or PHASE SHIFT over the whole audio band - means any output coupling cap must be very large or omitted. Even then, the practical CMRR of typical ( one op-amp) active balanced inputs is lucky to be 40 dB at low audio frequencies and usually much less at high ones. Given that ground loop hum voltages can be from 1mV to over 100mV, even achieving 40 dB rejection is not enough to render the humming noise silent in bad case. ....... Phil
From: Arny Krueger on 13 Dec 2006 10:08 "Rupert" <foodsteaks(a)linkline.com> wrote in message news:1165960287.503920.247090(a)80g2000cwy.googlegroups.com > Arny Krueger wrote: >> "Phil Allison" <philallison(a)tpg.com.au> wrote in message >> news:4u6q8rF16rc26U1(a)mid.individual.net >>> "Eeysore" >>>> Geoff Wood wrote: >>>>>> >>>>>> Why do you even need one ? >>>>> >>>>> I'd rather not send an unbal signal 50 metres. >>>> >>>> If the receiving end is wired balanced it makes no >>>> odds. >> >>> ** What asinine BOLLOCKS !! >> >> IME, there is quite a bit of truth to it. A lot of the >> benefit of balanced I/O is provided by the balanced >> input. >> >>> So balanced outputs are suddenly unnecessary >> >> A very high proportion of so-called balanced outputs are >> merely impedance balanced. IOW there is not really an >> active balanced output, just a low-value resistor >> connected to ground. > > A properly impedance balanced output is what's primarily > required to for CMR work. If you tie the ground of an > unbalanced output directly to pin 3 and the hot to pin 2, > the relative impedance of the ground to cold path is near > 1 ohm whereas the impedance to hot is whatever the > circuit is designed at, say 50 ohms. Ever do the math and see how much this really matters? > This means that > noise common to both wires feeding the balanced input is > not going to be of the matched amplitude an will > therefore not cancel when the cold polarity is flipped > and the signals are added. There may be some reduction of > noise, but not in the order of matched impedance hot and > cold. No argument that impedance-matching is ideal, but you might be surprised if you run some numbers and calculate the actual benefits. Remember, that the CMRR of many so-called balanced inputs is also limited by internal parts tolerances.
From: Phildo on 13 Dec 2006 13:11 "Rupert" <foodsteaks(a)linkline.com> wrote in message news:1165968648.622421.266390(a)j44g2000cwa.googlegroups.com... > I saw your excellent post. Mr. Krueger has stated in the past that his > method was virtually as effective as a balanced output, so I was > providing an explanation to the contrary. That method may work in some > situations, but it's not good practice. I'm sure he will disagree. Of course. Arny is incapable of ever admitting he is wrong in anything, even to himself. Is all part of his PPD affliction. Don't bother arguing with him, it will just cause another long and pointless thread. Everyone here already knows to ignore pretty much anything he says. Phildo
First
|
Prev
|
Pages: 1 2 3 4 5 Prev: ebay auction for a dcx 2496 mod Next: BEHRINGER EURORACK UB2442FX-PRO Problem/Repair |