Unbalanced to balanced connection

[Tubetec]

So , I have a battery powered mic/preamp combo , its unbalanced out ,
can I connect the unbalanced output (signal and ground)to the +/- of a balanced input and just connect the screen to the input end ground ?
Ive seen the term 'impedence balanced' single ended output , is that what it is ? is there extra resistors required ?

 

[scott2000]

Not sure if this has any useful info?

https://web.archive.org/web/2016021...resonus.com/news/articles/balanced-unbalanced

 

[Moby]

I would use quasi balanced connection

 

[Tubetec]

Thanks gents , I'll have a proper look tomorrow and most likely end up with more questions .

 

[andYz00m]

Here is the connection bible, might help you.

 

[Tubetec]

The two recomended ways seem to be the impedence balance connection that Moby showed and the other method I was thinking of below with screen only connected at the input end and signal applied between the phases. Will be interesting to look at how the different schemes measure up in terms of noise .

 

[Matt Syson]

Interconnecting from balanced to unbalanced is always an issue and will depend on your exact circumstances. If you are working in an area with significant 'airborne radiation' )poorly routed and wired mains cabling in the building, many switchmode supplies or large power transformers that are not in steel cases, the only way to (almost) 'guarantee' a low interference interconnection is properly balanced to balanced. The 'impedance balanced' output (a resistor in the 'cold' leg matching the hot leg impedance is a cheap way of possibly lowering interference although again it depends what the interference actually is and how much is being picked up. It also the throws the responsibility for good balance (and CMMR) onto the receiving end (input). Technically this is often a bit easier to do electronically as even a basic 1 op amp 4 resistor setup using 1 percent resistors probably has a CMMR of around 40dB which is a lot better than an unbalanced 0dB!. Whether you actually have a problem is the next question. That picture put up by Moby is only the 'simple' version as in reality the 'balance' of such is questionable as the resistor on it's own probably has a constant impedance to 'ground' whereas the resistor in the amplifier output has the variable output impedance of the amplifier be it increasing HF impedance or if it has an output capacitor, increasing LF impedance. The capacitor affecting balance against hum (mains interference) and the op amp balance at supersonic frequencies.
Thus the answer to the original question is really pretty complicated
Balancing at audio frequencies is a different matter compared to 'RF' completely as you have to be very strict with stray capacitances to maintain good 'balance' and even then the CMMR of the input stage may be ineffective.

 

[ccaudle]

with screen only connected at the input end

With screen only at the input end the screen is not connected to the source common mode node (aka source "ground"), so the capacitance between shield and signal wires acts to filter the signal. With shield connected to the source ground, the common mode voltage of the signal driver(s) and the shield move together, so no effective capacitance.
If you want to connect the shield at only one end the best end is the source end.
Either way with a battery powered source (or a double-insulated line powered source with no safety earth connection) you should have some connection between the source and receiver grounds to control the common mode offset.

 

[Tubetec]

Thanks Matt and CC . I'll need to chew down on that a bit for it to sink in .

 

[Matt Syson]

Having the shield connected at only one 'end' will produce different results. Whether they are noticeable for what you are trying to do in your environment is difficult to say. When putting the unit together you should be prepared to experiment, so 'design in' the ability to change if you find it necessary.
Best of luck with it anyway.
Matt

 

[Monte McGuire]

If you connect the shield only at one end, you can use a small RF quality cap to connect the shield to chassis at the "open" end to provide continuous RF shielding. This can be a good approach if you expect high RF, and it will prevent large LF fault currents flowing between chassis. Axial leaded SMD ceramic caps are a good choice as they are mechanically pretty tough.

 

[ccaudle]

If you connect the shield only at one end, you can use a small RF quality cap to connect the shield to chassis at the "open" end to provide continuous RF shielding. This can be a good approach if you expect high RF, and it will prevent large LF fault currents flowing between chassis. Axial leaded SMD ceramic caps are a good choice as they are mechanically pretty tough.

Good point, and Neutrik even make a set of cable connectors with a low inductance capacitor arrangement built in:
Neutrik EMC series connectors

You could build up cables and choose whether you wanted pin 1 connected (I think through a ferrite bead in those Neutrik connectors), or pin 1 not connected, but shield still connected to the shell capacitance.

 

[Matt Syson]

Unless this is only for your sole use you should design so that any length of any quality of connection cable (bought cheaply from a 'music shop' can be used which will give correct and consistent results.
Matt S

 

[Tubetec]

Thanks again Andy , Matt and CC ,
A little more detail about what I'm planning to do might be worthwhile .
So I have a pair of B&K 2203 precision sound level meters , Ext filter out is designed to match into greater than 600 ohms and puts out around 1.6 volts at full deflection on the meter ,output Z is in the order of 5 ohms , I plan to mount one of Ivan's (E1DA) Cosmos ADC's on each meter so the audio cable will only measure a few inches . Proprietary connectors on the B&K to XLR on the ADC , both level meter and ADC will be battery powered and connect to laptop pc via USB .

The EMC sheilded XLR's definately look like a worthwhile addition for the sake of a few euros extra cost , plenty of food for thought from all the wiring suggestions , but I guess 'Adjustment on test' and measurement via REW will be the only true measure of performance .

Im aiming for a very neutral and flat response which the B&K units already do very well compared to most studio condensers , a little digital EQ could render out any of the small peaks and troughs even further . The spare inputs on the ADC's will give me plenty of flexibillity to add a pair of directional mics to the set up where required .

 

[ccaudle]

I plan to mount one of Ivan's (E1DA) Cosmos ADC's on each meter so the audio cable will only measure a few inches

A few inches? Then you are overthinking all this too much, center pin of the connector to pin 2 on the XLR, shell of the connector to pin 3 of the XLR, shield connected to shell of the connector and pin 1 of the XLR. For a cable of just a few inches you could just use coax cable, and connect pin 3 to pin 1 at the XLR connector. If the buffer output is inverting then switch the pin 2 and 3 connections at the XLR.
Making the cable should take about the same amount of time as reading through all these posts.

output Z is in the order of 5 ohms

25 Ohms per the manual. Should still be low enough, although the Cosmos ADC has a strange input, the impedance is very low for higher gain (or perhaps lower attenuation) settings. Only 640 Ohms input impedance for 1.7V max input, which is really low for modern equipment.

Note that according to the 2203 manual the output is 0.3V for full scale deflection of the meter, so you may need to check your expectation of 1.6V into the ADC (see page 17 of the manual, it is designed to feed weighting filters that accept 1V max signals as input).
Note also that the output of the 2203 is designed to feed through weighting filters that are specified to drive into a 146k Ohm input, so it is not guaranteed that the output buffer is designed to source enough current to drive a 640 Ohm input.

The short of it is you should probably experiment a bit with your setup before expecting you can just connect this to a Cosmos ADC and go.
Do you not have an existing audio interface of some kind you can connect and check? One of those little Focusrite or PreSonus USB connected interfaces would have a more appropriate input impedance, and has adjustable gain so you can check how much gain you need for various sensitivity settings on the SPL meter.

 

[Newmarket]

This seems to be getting way over complicated tbh. Do you have access to the output circuit ? If you can convert that to any flavour of balanced output- and I suggest a simple impedance balance - then it will simplify things enormously.

 

[Tubetec]

So for an output of 316mv at the filter in socket you get 3.16 volts at the output stage , which corresponds to full scale deflection on the meter , 16 volts is the max output so around 12 db above the 3.16volt level . by my calculations that should mean around 1.3volts max at the filter in socket .

The later model unit specs the output of the filter in socket as what you see below , I think the earlier units may have a 25 ohm impedence though .

Thanks for the input Newmarket , I'd rather tap the signal from the pre-filter stage if I can , I was thinking maybe a step up transformer from 600-2k4 might interface better with the Cosmos ADC