audio development AD146_feat

Audio developments is well known UK manufacturer of very high quality location (portable) mixing consoles that find their use in location sound mixing/recording situations like tv shows, film sound, and music recording.

Audio development equipment is also known for their reliability, good set of features, power autonomy (battery power option), and great sound quality.
This is short story about AD146 , which is model that features eight mic/line input channels, four master outs, and various monitor and talk back features.

This piece was purchased on second hand market in pretty good working and cosmetic condition, but as all electronics are prone to aging and parts wear, it was decided that mix console first goes through cleaning and refresh process.
Before I continue, I want to say that basic idea behind all of this is realted with long time wish and passion that my good old friend Petar Milic and I share in idea of making highest quality analog recordings of good music with best available music artists.

I will write about those recording sessions in separate articles, now lest go back to AD146.
Before any actual work on this device, I purchased service documentation from Audio developments, and I must say that they have been very helpful and fast in response.

By looking at schematic I have seen space for upgrades in whole signal path.

So let go with the signal flow.
On every channel strip (Mic or line) input we have RF filtering composed of LC low pass filters on both hot and cold balanced signal nodes, then signal goes to low tuned HP fiter that removes ultra low freq/DC from input audio transformer for avoiding any chance of transformer core saturation. Before this transformer we also have L pad attenuator that have function of attenuating line level signals to mic level
because first input stage is actually high gain mic level stage composet of low noise dual SSM2220 transistor in long tail pair connection with OP-amp. This stage has variable gain which is set in global NFB(negative feedback) via switch selectable resistor array plus big value DC offset decoupling cap.
Second stage is three band active EQ that also drives channel fader(Penny&Giles). Fader, as you look at next stage, is actually variable current souurce for this next stage which is OPamp connected as current to voltage converter. After that signal is passed to signal splitting resistors to pan control/pot which works as variable shunt resistor for those resistors(no direct signal goes through this pan control), works like L pad.
One side of signal is than passed to output node of channel strip,and other side is passed to buffer and other output node(for final stereo paninag), those panned(or not) signals are sent to one or pair or quad assigned summing amps that sends signal to four separate master out channels. Those summing amps are also in current to voltage converter mode so OPamps from channels can be connected via high impedance resistors that prevens channel strip output OPs from interacting with each other. By-product of current signal transfer is also better signal to noise ratio and smaller signal crosstalk.

from summing amps signal is passed to another buffer amps that feeds master output faders(also used as varable current signal sources) after which we have OP I/V converters which than feeds hybrid output modules(ceramic based pscb´s) that are also in output limiter function via bootstrap nodes conected to time base regulated AC to DC converters(audio AC signal is converted to time delayed DC state in attack and relase domains). Those modules drives output audio transformers via big value decoupling caps for master balanced outputs.

Whole internal electronic signal path is single ended but all audio inputs/outputs are transformer unbalanced/balanced.
Now, what can be upgraded in signal path?

In respect of one mic/line input to master out signal path, I have counted ten(!!!) decoupling caps. All those ten caps were Philips(Vishay) generic quality unipolar electrolytics in almost no DC bias(read article about cap biasing) ) environment, I´ll say baaaad situation for signal quality,and those caps are prone to specs degradation over time.
In adition there was about eight caps, les´t say, indirectly(paralel) connected to signal path, in various compensation and RF fitering positions. all those were of ceramic (disc) types, also bad choice. Ceramics are great for some jobs in high freq and impulse domain,but not so great for audio circuits as they want to add sound grain and microphony.
All OP-amps (DIP8) were TL061/62 types. OK, but now we have better types with grater speeds, faster settling times, and pretty important, lower input noise specs and beter output drive.

After detail exam of topology there was pretty clear that there is no need for all those caps in signal path, but you never know what can go wrong if you bypass them without testing.
So I start to bypass one by one coupling cap and test behaveour of whole mix console with all options switched in and out of signal path. Electrical tests were done in frequency and phase domain, for any variations of DC offest, especially on legs of faders, potentiometers and switches.

At the end of testing, result was clear, only two(!!!) coupling caps are needed for stable work without clicks and humm or undesired low frequency energy build up through gain stages which can generate undesired clipping moments.

One after mic pre stage, and one before master output transformers.
First was replaced with good quality film cap, and output caps that needs bigger capacity are replaced by top quality wet tantalum caps that are military grade components with no aging effect, practically last forever.
All other places were bypassed.

All OPamps are replaced with J-Fet input types TLE2071(single) and TLE2072(dual). Those were selected as best choice regardless of topology, stability and given sound quality.
All DIL8 sockets were removed and ICs were directly soldered to pcb for better long life contacts.

All ceramic caps in compensation and parallel to signal positions were replaced by precision high quality film types, and tantalum on bootstrap positions.

Further on, mixer use top quality Penny&Giles faders that are practically indestructible, they only need periodic maintenance and cleaning.
All twelve faders were disassembled, sliding rails, conductive plastic resistive elements and wiper elements were carefully cleaned and faders were reassembled.
All power supply rail filter caps were replaced with new, low ESR, high temp, long life types from Panasonic.
Actual power supply module is simple but effective switching type, it generates +/-12V for audio electronics, additional 12V for LED indication and 12 V phantom power, and 52V for filtered 48V phantom.
Down the stream is multiple RC filtering.
As new Ops suck little more current than original OPs, power supply is modified for higher current limiting.
All XLR input and output connector contacts were polished, cleaned, and conserved with anti oxidation fluid.
All open style switches were also cleaned and conserved for long trouble free operation.

Next,  adding of channel strip direct out option, which is not factory option without loss of other inputs, if it can be done.

I choose to add new set of connectors that are relative standard in pro audio, small, and high quality.
My choice fell on mini xlr (or TA-3) connectors from Neutrik (REAN).

In one step of this whole job, mixer was completely in peaces, so I had back panel free from rest of housing. So I made precise positioned drilled holes for mini xlrs, without distroying model numebring on that panel.
It was pretty tight run as you can see.

Connectors were linked to main boards with high quality wiring.
I had to develop balanced signal source for direct outs.
As there is no balanced signals inside mix console, and there is no enough space for eight high quality line outtput audio tranformers and driver stage, so I choosed to go with electronic balancing and driving of those outputs.
As those direct outputs in practice are connected to recorder that have 48V phantom power option on his inputs, protection against incoming phantom power must be used.
Texas instruments have nice high quality DRV134 IC for space limiting solutions, which is also simple to implement and ask for almost no external components.
I designed small custom PCB for direct out module that contains DRV134, local power supply decoupling, and Phantom power protection.

Those are installed in each channel strip. Audio signal for direct outpus was taken from post fader point so you have individual control of signal level on those direct outputs.

At the end mix console was tested with Rohde&Schwarz UPL audio analyzer.
S/N ratio on mic inputs was about 10dBs better, THD was lower for one decimal point with better (more neutral) harmonic structure.
In practice sound is far more clear with richer timbre, far more low level and space information on recorded material, dinamics are firs rate both in micro and macro region.
Mix consle is now like new, much better sounding, with better specs and added direct out feature.

Thank´s for reading.