neil's webbly world

me@njohnson.co.uk
Academic Pages
learn Russian
  

DDA D-Series (16:8:2) Mixer


Saw this fixer-upper come up on the usual place, and after some negotiating with the seller I got it for a very reasonable price. Untested, without power supply, ex-theatre company, looking a bit sad. Perfect!!!

It's not your usual mixing desk. As well as the obvious PPM meters in place of the more common VU meters, it turns out that nine out of the sixteen input channels have Neutrik mic transformers fitted. It also has custom multi-way connectors on the rear, so it will need a complete new set of rear channel plates making and fitting, and it did not come with a power supply, so that needs sorting out as well.

And So It Begins...

First task was to remove all the channels, cabling, meters, and metal work down to the bare parts. Then clean everything and start rebuilding. So here is a photo that explains why this mixing desk is so heavy!

Yep, that's right, a welded steel tubular chassis that must weigh close to 20kg on it own! This 30-slot frame is 1m (40 inches) wide, to give you some idea of scale.

Casework

The first stage is now completed and the metalwork is all back together again. Main changes were to replace all the old screws with new blackened ones, clean out some of the threaded inserts, treat the rust patches on the base plate with rust neutraliser, and give everything a thorough clean with disinfectant wipes and sticky label remover. There was plenty of dust, fluff, and a dead spider.

Once the main metalwork was complete it was time to reupholster the armrest and refit the reconditioned wooden side panels. The meter panel has also been stripped down and cleaned (loads of masking tape goo on the rear!):

Next task is to start on the wiring harness and rear panels. Then start on the modules (master, inputs, outputs, and auxes). Power supply will likely be last as I can test the system on a bench PSU for now.

Meters

First foray into the electrical side of things is to test and refit the meters. This desk was fitted with the BBC PPM (Peak Programme Meter) option (white needle against a black background). I originally thought to replace these with Sifam VU meters, which would be a drop-in replacement. But after some digging around and costing it up I decided to stick with the present PPM meters. A gentle introduction to programme meters is found in BBC Training Supplement No. 6, together with technical details of the classic ME12/9 and ME12/10 PPM meters. You can find more information about BBC PPM meters at the excellent BBC Eng.info website in the "ME" section.

Most PPM meters require a driver board to correctly drive the needle with the correct response and ballistics. DDA were no exception, and every PPM meter has a driver board bolted on the back of the meter:

Before refitting them I decided to test them, using a bench supply and a signal generator courtesy of my colleague Bruce.

Unfortunately two issues were found. Firstly, whoever re-wired the mixer with the weird Bendix connectors messed up the meter power wiring: instead of connecting them to +18V and 0V they wired them directly across the +18V to -18V rails, putting 36V onto the meters. While not a problem for the TL071 and TL072 opamps (although rather close to their abs.max limit), the mid-rail circuit was no longer mid-rail, and the op-amps got rather toasty. And secondly, the meters for channel 5 and Right were faulty, so needed to be fixed.

Fixing requires a schematic, and unfortunately the excellent DDA Consoles website did not have it. So some time spent with a photocopier, coloured pens, and TinyCAD resulted in a schematic (click to view). Note that the -4 trim corresponds to "3" on the meter scale, and +8 corresponds to "6", with "4" being the 0dBu level in accordance with the standards.

The meter input is fully differential and capacitively coupled simplifying the signal connection (no need to worry about which pin is ground or signal). The 9V1 zener diode generates the mid-rail. After that it's a fairly standard diff.amp into full wave rectifier into meter driver with piecewise logarithmic function generator feeding the meter.

The only issue to note is that the polarity of the input capacitors requires the local mid-rail reference to be positive with respect to the mixer's 0V. This means I cannot run some of the meters off -18V and 0V as that would put a rather unhealthy negative DC bias across the input capacitors which would probably let the magic smoke out. An option to be explored is to generate a local -6V/+12V supply for the meters, a couple of beefy zener diodes would be good enough as the load is fairly constant.

With the meters all working electrically, attention turns to the way they are mounted. This is done with a black plastic bezel and two little spring things on either side of the meter. Unfortunately the combined weight of the meter and driver board looks like it is a wee bit too much for the two springs, so an application of Dow 744 RTV silicone is enough to hold them in place without causing damage to the meter or metalwork:

Meters held in place with RV silicone

Finally, with new wiring harnesses assembled and fitted the meter bridge can now be refitted to the frame:

Meter bridge refitted

The Curious Case of the Cue Switches

The first module for bringing up is the monitor master module, for two reasons: it brings the power into the mixer and onto the mix bus, and it has the headphone amplifier and test tone generator. It also has two switches for operating the cue lights:

Cue light switching

Tracing out the four combinations of buttons shows that while the operation of the LEDs is as expected, the behaviour of the external connections does not exactly follow the panel indicators:

CUERed/GrnLEDsConnections
UpUp----
UpDown--2 - 3
DownUpGreen1 - 2
DownDownRed2 - 3

The issue is the connection to the red/grn switch. Swapping two of the PCB traces round will remove the odd state and make the cue and light controls work as expected (cue lights on only when the CUE switch is down, and the colours match the front panel indicators). A couple of small bodge wires tacked down to the board should fix this little oddity.


Copyright © 2001-2019 Neil Johnson