Thursday, 15 May 2014

Monitor Matters

The next stage of the mixer build is the wiring and commissioning of the monitor panel. As I mentioned in a previous post, it is surprising how much wiring there is even in a small mixer and half of it seems to be associated with the Monitor Panel. The monitor in this mixer is quite straightforward. It consists first of a 6 way rotary switch that selects one of the four direct outputs or the two AUX sends. The output of this switch feeds a three way switch that selects between the 6 way switch, the main stereo bus  and a two track playback input. The output of this switch then feeds the two VU meters and two pots. One pot is the monitor level control which feeds the monitor out sockets at the rear of the mixer and the second one is the headphones level pot.

Nothing really complex but it turns out to be a lot of cable. There are six cables going to the 6 way switch plus four more for the bus and  two track playback going to the 3 way switch. Then there are two outputs to the meter, two to the headphones amp and lastly the two monitors outputs themselves for a total of 16 cables. All these need to be run under the main mixer sub-rack to the rear of the mixer and the headphones and meter cables need to reach even further to the meter bridge. Here is a picture of the completed wiring of the Monitor Panel:

To help route the cables neatly to the rear of the mixer they are gathered together in bundles using tie wraps. To make sure I could identify the cables when they reached the back of the mixer I colour coded the tie wraps as follows:
  • The bundle of 6 going to the 6 way switch were coded using the resistor colour code to represent 1 through 4 direct outputs with 5 and 6 repenting AUX 1 and 2 respectively.
  • The remainder are all stereo pairs so each pair was colour coded and a the far end of the cable, the right hand channel had a red tie wrap added to it.
Commissioning the monitor section actually highlighted a number of other faults which I was then able to correct. In fact the monitor section is very useful for overall mixer  and channel commissioning. I first checked the inputs to the 6 way switch. I used the known working Helios 69 module in these tests. First I checked the four direct outputs by placing the Helios 69 module in each of the four channel slots and making the appropriate selection of the 6 way switch. Then I checked the two AUX sends.

Next I checked the main left and right buses. This was useful because it showed up several problems with pan controls. It turned out the pan on the Helios and one other channel amplifier and also on one of the AUX returns were not working as expected. In all three cases a simple wiring error was found and corrected. Checking the AUX returns was also useful as it highlighted another issue. The AUX return is completely passive. The signal passes through a transformer, a level control and lastly a pan pot before being fed to the left and right buses. There is perhaps one dB of loss in the input transformer so with the pan hard left or hard right, the level on the bus might be a dB or so lower than the level from a channel. However, it turns out to be over 9dB less. The reason for this is the step down transformer used for the direct outputs. This steps down by 6dB so to get a nominal 0dBu level from a direct out means the channel amplifier is outputting +6dBu and this is the level it feeds to the bus. If we allow a 1dB loss in the direct out transformer then this, plus the 6dB extra level and the 1dB loss in the AUX return input transformer, accounts for 8dB of the 9dB difference.

The solution is to use larger value bus feed resistors for the channels. Currently the bus feed resistors are set at 47K. If we increase them to 150K then the bus level from the channels drops by 9dB. Normaly this would make the noise level about 9dB worse, but since the channel level is already 6dB higher, the overall effect is only a 3dB worsening of noise. The extra signal loss on the bus is made up by the bus amplifier which has plenty of gain in hand as this is only a 4 channel mixer.

The next problem to be highlighted by the Monitor Section was with the AUX sends. Each one worked fine on its own but, with the amplifier gain set to 60dB and with both AUX sends turned up full, some sort of oscillation occurred as could be seen by the VU meter trying to wrap itself around the end stop. Checking the wiring from the AUX send pots to the bus resistors in the channel amp revealed the wires were laid flat along the top surface of the PCB. There was a good chance this would couple the signal from the AUX sends to other parts of the amplifier circuit. Simply re-routing all the bus feed signals in the air an inch or so above the PCB cured the problem. In future it would probably be best to use screened wires for the feeds from the pan and AUX send controls to the bus resistors. This modification has been applied to all the channels amplifiers.

The last problem identified by the monitor section was with the meters themselves. I tested the 2 track playback input which is simply a direct connection from the connectors at the rear of the mixer via the switches to the meters. Feeding the same level into both the 2 track playback inputs should give identical readings on the meters but it does not. One meter is about 1.5dB below the other. I checked this again by monitoring the direct and AUX sends where the one signal is sent to both meters and got the same result. Clearly one meter is faulty so the next step is to feed in exactly +4dBu and see what each meter reads.