Studio Cue System Project

[Potato Cakes]

The ground lift test did not fix the issue. I thought it might as the signal is getting into the ground so the thought was to remove it or isolate it at the input or maybe at the INA134.

I did not change the pan pots to 100k. I thought about as I was looking at the schematic. Time to order more parts!

I am only testing one box right now. No other boxes are connected.

Same leakage with different lengths of cable.

The schematic for this project is based off what I've seen on several mid and small sized consoles which I've used. There is usually a buffer op amp between the fader and pan pot. Those consoles all used 10k pots and I haven't had any issue with leakage on those desks. I'm probably forgetting a key resistor or cap somewhere as am in the habit of doing lately.

Regardless, I will swap the pan pots out and see what happens.

Thanks!

Paul

 

[abbey road d enfer]

Same leakage with different lengths of cable.

This seems to eliminate ground current as the probable cause.
Maybe it's the capacitive leakage in the CAT5 cable, but this would also vary very much with cable length, and teh leakage would be high-passed.

The schematic for this project is based off what I've seen on several mid and small sized consoles which I've used. There is usually a buffer op amp between the fader and pan pot. Those consoles all used 10k pots and I haven't had any issue with leakage on those desks.

There is nothing wrong with this schemo, but the devil is in the details, parasitic resistance, parasitic capacitance, unfortunately we've reached the limits of what's achievable remotely.
Maybe the pots themselves do not cut properly, have you tried shorting wiper to ground on the unused channels?

 

[Potato Cakes]

I have not tried shorting the ground of the wipers. Maybe that will be next. As you pointed out we are at the limits of remote help. I'll tinker around with this some more and stare at the schematic to see if anything new comes to mind.

As I said earlier, it works as intended and JLM 12PAK amp being used to drive the headphones has more power than anyone should ever need. Audio mixed through it sounds excellent. I just obsess with the imperfections of my builds and it is hard to leave things well enough alone.

I'll report back after swapping out the pan pots.

Thanks!

Paul

 

[Potato Cakes]

Swapped the pan pots. Still the same results. Signal is getting through the ground with all the volume pots all the way, which is high passed above 3kHz. Again, once any volume is actually turned up, the signal is so much louder than any bleed and more importantly, it sounds great and has much more power than most any other commercial studio cue system that I've used. This problem could probably be solved with some minor reworking of the circuit, I'm just not going to bother doing so.

Thanks!

Paul

 

[abbey road d enfer]

Swapped the pan pots. Still the same results.

Have you also increased accordingly the bus resistors? (And the summing amp's FB resistor, of course)

 

[Potato Cakes]

I did change the bus resistors to 47k as well.

 

[Potato Cakes]

I think what I may have to do is just add some mute switches that break the connection of the summing resistors so there is no return path for the ground. I'll have to verify this works but it seems to be the easiest solution without working the entire circuit.

 

[abbey road d enfer]

I think what I may have to do is just add some mute switches that break the connection of the summing resistors so there is no return path for the ground. I'll have to verify this works but it seems to be the easiest solution without working the entire circuit.

It's possible. I must admit it's a little puzzling, so experimentation is needed.

 

[Potato Cakes]

Scratch that idea. Just realized that would be changing the bus impedance every time some thing was muted. I'll just stick to shorting the audio to ground like it is in console. My only concern is that since I'm getting this leakage that I would still have the same issue. Like you said, experimentation is needed. I'll do some tests when I get later this week.

Thanks!

Paul

 

[Potato Cakes]

I think I found the cause of the issue. On my Tac scorpion, the makeup gain for the goes to the non-inverting input and the inverting side is used for feedback gain. Neither side is making direct contact with the ground. So now the question becomes what is the affect on the output signal if I simply put a 10k between the non-inverting inputs on the NE5532 and ground? I am certain that having the ground mixed into the + sides of the NE5532 is where I am getting the faint signal. I just need to do some testing later this week if I get a chance.

Thanks!

Paul

 

[abbey road d enfer]

I think I found the cause of the issue. On my Tac scorpion, the makeup gain for the goes to the non-inverting input and the inverting side is used for feedback gain. Neither side is making direct contact with the ground.

You can't compare a non-inverting circuit with an inverting one. Of course all inputs need a galvanic connection to ground or to a fixed reference.

So now the question becomes what is the affect on the output signal if I simply put a 10k between the non-inverting inputs on the NE5532 and ground?

It increass the noise gain of the stage.

I am certain that having the ground mixed into the + sides of the NE5532 is where I am getting the faint signal. I just need to do some testing later this week if I get a chance.

You may try it but don't expect too much.

 

[Potato Cakes]

After making a Rev B boards with the above mentioned changes the overall results are worse. I've tested and prodded and almost caused some hearing damage inject signal into the wrong place. I even shorted all pins on the input XLR and I could not get the signal to be reduced to levels that would be considered turned off. I did find a post where thor uploaded a simulation similar to what I am doing with the passive mixer but the cross talk level is similar to what I am trying to achieve. I'm going to go back to Rev A and rebuilt the mainframe and focus on the mixer section. I think tried to short the signal to ground to simulate a mute switch but I don't think that worked either. I will retest this when I compete the mainframe rebuild.

I did think about redesigning everything and have the mixer sections receive power down the Cat6 cable and building cue stations as active mixers, but big part of this whole exercise is having powerful amps to drive the headphones. From earlier in this thread I mentioned that I was using a JLM Audio 12 PAK which is twelve channels of 75W amplifiers, which will drive any set of headphones cleanly and beyond what the more hearing challenged musician could want. I don't know how well a pair of individual Cat6 leads could power two of these amps over a 25' run.

Thanks!

Paul

 

[drtechno]

After making a Rev B boards with the above mentioned changes the overall results are worse. I've tested and prodded and almost caused some hearing damage inject signal into the wrong place. I even shorted all pins on the input XLR and I could not get the signal to be reduced to levels that would be considered turned off. I did find a post where thor uploaded a simulation similar to what I am doing with the passive mixer but the cross talk level is similar to what I am trying to achieve. I'm going to go back to Rev A and rebuilt the mainframe and focus on the mixer section. I think tried to short the signal to ground to simulate a mute switch but I don't think that worked either. I will retest this when I compete the mainframe rebuild.

I did think about redesigning everything and have the mixer sections receive power down the Cat6 cable and building cue stations as active mixers, but big part of this whole exercise is having powerful amps to drive the headphones. From earlier in this thread I mentioned that I was using a JLM Audio 12 PAK which is twelve channels of 75W amplifiers, which will drive any set of headphones cleanly and beyond what the more hearing challenged musician could want. I don't know how well a pair of individual Cat6 leads could power two of these amps over a 25' run.

Thanks!

Paul

I would take the signal part of it out of the remote mixing equation, and use VCA's in the head unit so that only dc voltages are going through the CAT5 instead of signal.
Something like an AS3360 VCA with a TLE2074 as op amp buffers for 2 channels.

I use a Yamaha TF1 dante PA for que mixes.

 

[Tubetec]

Did you say 75W per channel on a headphone amp ?
Sounds like enough to make peoples head catch fire if it doesnt render them stone deaf first ,
Putting that much power at the finger tips of the musicians , headphone leakage might become an issue .

 

[abbey road d enfer]

After making a Rev B boards with the above mentioned changes the overall results are worse. I've tested and prodded and almost caused some hearing damage inject signal into the wrong place. I even shorted all pins on the input XLR and I could not get the signal to be reduced to levels that would be considered turned off. I did find a post where thor uploaded a simulation similar to what I am doing with the passive mixer but the cross talk level is similar to what I am trying to achieve. I'm going to go back to Rev A and rebuilt the mainframe and focus on the mixer section. I think tried to short the signal to ground to simulate a mute switch but I don't think that worked either. I will retest this when I compete the mainframe rebuild.

I did think about redesigning everything and have the mixer sections receive power down the Cat6 cable and building cue stations as active mixers, but big part of this whole exercise is having powerful amps to drive the headphones. From earlier in this thread I mentioned that I was using a JLM Audio 12 PAK which is twelve channels of 75W amplifiers, which will drive any set of headphones cleanly and beyond what the more hearing challenged musician could want. I don't know how well a pair of individual Cat6 leads could power two of these amps over a 25' run.

Thanks!

Paul

don't thhink it's the mixer that creates the issue.
I would look at how you distribute the amplifier's outputs to the headphones. It's not clear.

The return currents from the headphones may impact the "ground" cable.

 

[Potato Cakes]

I'll have to see if I can get a schematic from Joe for the 12PAK. It's 12 of his single amp boards that use an LM3886 and is wired for balanced input. The output are wired in pairs to XLR connectors on the front, Pin 1 is both pairs of the (-) output of the LM3886, and Pin 2 at 3 are the (+) connections of pairs 1 and 2 respectively. Pin 1 on the output is isolated from the rest of the circuit ground. Should the (-) connections be tied to the circuit ground? I've noticed that is the case with his headphone amplifier board but I did not think about that when combining two different amplifiers for a headphone output. I will try this when I rebuild the boards.

Thanks!

Paul

 

[Potato Cakes]

Did you say 75W per channel on a headphone amp ?
Sounds like enough to make peoples head catch fire if it doesnt render them stone deaf first ,
Putting that much power at the finger tips of the musicians , headphone leakage might become an issue .

That is up to them. Halfway up is plenty of volume. Plus the outputs are current limited. The point is so the headphone amp doesn't distort. That is the biggest complaint that I have received with the other units in a closed, live band tracking situation.

Thanks!

Paul

 

[Potato Cakes]

I would take the signal part of it out of the remote mixing equation, and use VCA's in the head unit so that only dc voltages are going through the CAT5 instead of signal.
Something like an AS3360 VCA with a TLE2074 as op amp buffers for 2 channels.

I use a Yamaha TF1 dante PA for que mixes.

I don't know enough about VCAs yet to do this, but it would be an excellent idea. I have another project where I am trying to implement a single VCA but haven't started testing.

Digital is not an option with this cue system. Especially when it comes to singers.

Thanks!

Paul

 

[drtechno]

Digital is not an option with this cue system. Especially when it comes to singers.

Thanks!

Paul

I don't know where you would get that because its only if you are using a computer interface to monitor it. Digital mixer latency is in the nanoseconds and people are talking out their butt if they can tell less than 10 milliseconds. Modern DAW monitoring is horrible because its twice this latency, but old pro tools wasn't because it was a hardware mixer so after that, that monitor button in DAWs never worked correctly after 2003.

 

[kags]

I don't know where you would get that because its only if you are using a computer interface to monitor it. Digital mixer latency is in the nanoseconds and people are talking out their butt if they can tell less than 10 milliseconds. Modern DAW monitoring is horrible because its twice this latency, but old pro tools wasn't because it was a hardware mixer so after that, that monitor button in DAWs never worked correctly after 2003.

People seem to forget that even stage monitors can be several feet away from a performer (or even further for sidefills) so the delay from distance to the monitors can be at least a few milliseconds. DAW latency can be awful, but monitor systems have never been an issue for me in studio or live situations.

In L.A. most newer studios use Avioms, while many older installs still use the Furman systems, and the top studios still use the Manley or Mytek cue systems. I've used Berhinger systems that seem to work well, and I've had mixed results with Hearback systems.

A few studios still use the old Mackie 16-channel mixers (or similar) for cues. Every musician knows how to push faders around until they get a mix they can tolerate. There are some analog 8-channel mixers with decent headphone outs for around 100 bucks, but then you need multicore snakes and power for each station.