DIY line level summing

[ChrissA]

This is very debatable. True only if the reflected impedance at the primary is significantly smaller than the mixing resistors.
In the absence of relevant data this is open to speculation.

It's the aux output of an M32 going into a Sennheiser G4 IEM transmitter

 

[ruffrecords]

Given the TRS are the sources then the first one is correct.

Cheers

Ian

 

[Newmarket]

It don't know what Cp and Cs models are, but I am using a DMM (digital multi meter, right?)

Yes re DMM. The p and s refer to parallel and serial respectively. One models a resistance as being in parallel with the capacitance and the other...well you can guess

 

[abbey road d enfer]

Its late here so I might understand in the morning, but I'm not following. To prevent confusion like there has been before I have made some (extremely janky) diagrams. This is how I have soldered everything up now:

View attachment 135568

That is correct.
Now regarding the HF loss, the enormous capacitance values suggest about 1000+ meters length.
I seriously doubt the veracity of your measurements.
If the signals are fed into these cables via resistors, a huge HF attenuation is bound to happen.
Alternatively, if the cables are fed directly from the sources, the HF loss will be that resulting from the internal impedance of the source, which is so far unknown..

 

[ChrissA]

That is correct.
Now regarding the HF loss, the enormous capacitance values suggest about 1000+ meters length.
I seriously doubt the veracity of your measurements.
If the signals are fed into these cables via resistors, a huge HF attenuation is bound to happen.
Alternatively, if the cables are fed directly from the sources, the HF loss will be that resulting from the internal impedance of the source, which is so far unknown..

So what you are saying I should solder it up like the second diagram? Isn't it a problem I am directly connecting two outputs together?

 

[abbey road d enfer]

So what you are saying I should solder it up like the second diagram?

Did I ever say that?
What I'm saying is that the HF loss depends on the way the cable is connected.
As I said earlier I seriously doubt your capacitance measurements.
I would like to see an exact diagram of where the resistors are, where the connectors are and where the cables are. That makes a significant difference.

 

[ChrissA]

Did I ever say that?
What I'm saying is that the HF loss depends on the way the cable is connected.
As I said earlier I seriously doubt your capacitance measurements.
I would like to see an exact diagram of where the resistors are, where the connectors are and where the cables are. That makes a significant difference.

Pardon me, that's what I understood. I can understand why you doubt the measurements, I've been told with a DMM they can be off. Diagram 1 (posted earlier) is how I have soldered it up. 2 lengths of 50cm tops of microphone wire, resistors in the TRS plugs and summed at the XLR

 

[abbey road d enfer]

Pardon me, that's what I understood. I can understand why you doubt the measurements, I've been told with a DMM they can be off.

Most audio cables have a capacitance varying from 70 to 180pF/meter. Your results in hundreds of nF make no sense.

Diagram 1 (posted earlier) is how I have soldered it up. 2 lengths of 50cm tops of microphone wire, resistors in the TRS plugs and summed at the XLR

That mean sthat the resistors see the cable's capacitive load, which results in a low-pass efect, but certainly well beyond audio frequencies.

 

[ChrissA]

Most audio cables have a capacitance varying from 70 to 180pF/meter. Your results in hundreds of nF make no sense.

That mean sthat the resistors see the cable's capacitive load, which results in a low-pass efect, but certainly well beyond audio frequencies.

The cable I'm using according to it's specs is 55pf/m, which tracks with the measurements of the cable not plugged into anything: 74pF between + and - 100-115pF between + and gnd, and - and gnd measured at the XLR. Measurements most likely are not correct, I was searching for the answer of the lowpass filter in capacitance but it almost can't be that since no measurement makes sense, going from 10k to 2k should have fixed everything but it didn't, etc etc. Right now I have 1k resistors inside the jack plugs on the tip and the ring (which plug into the outputs), and they sum at the xlr where I do nothing except sum the two, and there is a -0.5dB drop at 20kHz which is more than manageable. If there is another (better) way of doing this I'm more than happy to try ofcourse, or if you can tell me it's bad for equipment to do it like this for example, please do tell me!

 

[kags]

 

[abbey road d enfer]

The cable I'm using according to it's specs is 55pf/m, which tracks with the measurements of the cable not plugged into anything: 74pF between + and - 100-115pF between + and gnd, and - and gnd measured at the XLR. Measurements most likely are not correct, I was searching for the answer of the lowpass filter in capacitance but it almost can't be that since no measurement makes sense, going from 10k to 2k should have fixed everything but it didn't, etc etc. Right now I have 1k resistors inside the jack plugs on the tip and the ring (which plug into the outputs), and they sum at the xlr where I do nothing except sum the two, and there is a -0.5dB drop at 20kHz which is more than manageable. If there is another (better) way of doing this I'm more than happy to try ofcourse, or if you can tell me it's bad for equipment to do it like this for example, please do tell me!

0.5dB roll-off at 20kHz suggests the receiver to which your summer is connected has a rather large input capacitance, about 1.2nF, which is not unusual for a mic input designed to receive 150-200 ohms sources.

 

[abbey road d enfer]

View attachment 135621

This is not very different than connecting microphones in parallels. The main advantage is the galavanic isolation, but for the rest, the two microphones are essentially in parallels, each one seeing the other as load, which may not be a terrible problem if they are identical, but becomes one if they have different impedance curves.
I would not use that for combining a Shure SM57 with a Sennheiser 441.
The only mitigating factor is the DC resistance of the transformer, which somehow buffers the effect, at the cost of some level loss.
Note that JT112L specified as a Line Output transformer.
If each input is driven by an active stage, these two stages end up being paralleled, which we know can be problematic, particularly when summing uncorrelated signals or worse, out-of-phase signal.
The only limiting factor would be the DC resistance of windings.

 

[ChrissA]

This is not very different than connecting microphones in parallels. The main advantage is the galavanic isolation, but for the rest, the two microphones are essentially in parallels, each one seeing the other as load, which may not be a terrible problem if they are identical, but becomes one if they have different impedance curves.
I would not use that for combining a Shure SM57 with a Sennheiser 441.
The only mitigating factor is the DC resistance of the transformer, which somehow buffers the effect, at the cost of some level loss.
Note that JT112L specified as a Line Output transformer.
If each input is driven by an active stage, these two stages end up being paralleled, which we know can be problematic, particularly when summing uncorrelated signals or worse, out-of-phase signal.
The only limiting factor would be the DC resistance of windings.

Just to be clear, I am not summing microphones, nor am I summing out of phase signals, both signals are completely discrete so uncorrelated and line level. They also do not go into a mic preamp, they to into the line in of a sennheiser G4 IEM transmitter.

 

[abbey road d enfer]

Just to be clear, I am not summing microphones, nor am I summing out of phase signals, both signals are completely discrete so uncorrelated and line level.

Hence the jusrtification for separation resistors.

They also do not go into a mic preamp, they to into the line in of a sennheiser G4 IEM transmitter.

This input is globally unspecified, except for max level and a HF limitation of 15kHz (tolerance unknown).