talkback-to-hp-monitoring

Hey Guys,

I need to sum a switched TB signal into a HP monitoring chain and was thinking something like this.

The TB audio will be coming from either a shure-automatic-mic-mixer, or from a switched TB out of a console. The HP amp is an HP4.

The goal is to maintain good monitoring signal and have the TB audio mixed in when audio is switched to that line.

I was looking for some feedback and input on resistor values.

Thanks!

- jonathan

You may have some crosstalk between L & R.
You must check that source impedance of TB source is low enough and CMRR of HP amp is good enough.
I would suggest 2.2k resistors everywhere, should be compatible with most console/soundcard outputs.

Thanks Abbey, The monitor audio source and the TB audio source are both 100 Ohm and the HP amp is probably 10k so 2k2 fits nicely in there as far as bridging goes. Not sure about the CMRR of the HP amp but if I use 1% resistors and match them as best as possible then wouldn't the CMRR pretty much be the same as without the R's?

Not sure how to figure the cross-talk or attenuation with this setup but cross talk won't be that much of an issue in this case (as long as it's not too bad, and I would like to optimize anyway) since they are currently monitoring in mono! Attenuation could be an issue. My goal is to keep a simple passive solution because it should be less wires and maintenance but can always go with one of those RDL stick-on things even if I don't particularly like the RDL approach.

Best,
jonathan

well, the attenuation is about 7dB with a 10k input, I think it should be ok.

Hey Abbey, Thanks!

Looks to me like you came to -7db by looking at this a 2.2k/2.2k divider into 10k. At least, when I do that I come with about 7db as well.

For instance, a typical divider consists of R1, R2, and RL or even simpler R1 & R2T (where R2T=R2||RL).

R1=2.2k, R2 = 2.2k, RL=10k. VIn=1.23V,
VOut = VIn*R2/(R1+R2)
VOut=.554V,
db=20*log10*(vout/vin)
db=-7

But this is only considering the MON-L and TB part of the network. Simplifying it to include the MON-R impedances makes it more confusing to me about which to parallel, etc.. So when in doubt, block it out and simplify.

So I added in the other path (MON-R), simplified, and it appears to be a wash for attenuation.

Now, for crosstalk I am coming up with about -61dB based on my simplifications. Does that sound like the ballpark?

I appreciate it!
best,
j

The voltage divider is 2.2k + 2.2k series resistance into (2.2k + 2.2k) // 10k shunt resistance. The attenuation is (4.4.10/4.4+10)/ 4.4 +(4.4.10/4.4+10) => 0.41 20log(0.41) = -7.74 dB.
Because of the relatively low impedance of the TB source, the Right half can be considered as negligible. In fact the R half imposes a resistance of about 8 kohms in parallels with the TB source Z. The influence is about 1% - insignificant.
X-talk is defined by the amount of let's say the R signal that goes into the TB source Z (approximately 100/8.8k =>0.011) then that goes to the L output with 7.7dB of attenuation. But remember that the L signal is also attenuated by the same amount, so the effective xtalk figure is -38.8dB (20log0.011)