Some have expressed interest in how my Delta rebuild project progresses.
So, here's a summary of what's been done so far.

If interested, please come on along -- helpful comments, questions, and
suggestions always welcome. If you're not interested, please killfile now,
either on my name or on the subject keyphrase "Delta Rebuild Update" (I'll
always use that in the subject line).

==
Acquired console August 2004 based on its functions, flexibility, and
clean design for our main application: location live and location session
work, mainly acoustic.

Basic build quality seemed solid; best of all it was modular with lots of
room on the boards -- easy to work on.

Acquired PDF of schematic set. Ran off two sets of 20"x24" prints on the
big plain-paper printers at the local blueprint house. Started studying as
time allowed.

Cursory tests showed that everything worked; further tests indicated most
switches crunchy, but all faders in excellent health and only a few pots
were bad.

Ran 100 Hz through all channels to find bad switches (channel on, pan,
pfl, eq on, line, phase, etc -- mostly all bad).

Got 100+ new DPDT and 4PDT switches from Harmon. Good to know they still
carry full spares for this model. Talked to Frank and Rene there; nice,
helpful folks.

Replaced bad switches, retested. Switching now solid and silent. TB
function restored (another bad switch; but replaced electret TB capsule
anyway based on age).

Console used on a few sessions; intermittent phantom and PFL on some
channels. At first assumed switching but was perplexed as most had been
replaced. Instead, eventually concluded bad sockets on the master loom
(ribbon cable assembly called "loom" by Soundcraft). Most contacts pretty
relaxed on those 34 pin female connectors probably due to age; some were
so loose they just fell off the board connector once the hold-down levers
were released.

Master loom many months on back order. Rene finally said he'd just build
one. Next time I will probably get the $20 connector press and do my own.
An audio systems designer friend suggests hardwiring; probably a good idea
but mostly impractical; I'd rather just build a new loom every couple of
years.

New loom installed, all documented weird intermittent problems now seem
gone.

==

Added a 2-gang pot to RH master so as to provide variable monitor dim,
along with DPDT ON-OFF-ON toggle to provide switchable monitor dim,
normal, and mute functions via a switch. (a small 20K linear is what I had
in the parts box; an 50K log taper would likely be better.)

Added a momentary mono switch. Very handy.

It's convenient that these modules are built so that without too much
hassle the PC board can be removed from the metal frame. Made using the
drill press for switch and pot mounting holes much simpler.

Also added PS rail LEDs. One unsettling thing about this unit is that the
power supply shoebox has NO pilot lights, and there's nothing on the
console that glows if it's just idling. Aside from running a signal, you
have to press a channel "on" or PFL button to see if there's power; I
prefer to know at a glance that incoming voltages are present.

Located dim pot just above and to the right of the master faders (wiper
and one leg from each dim pot segment replaced R151 and R51 on RH master).
Mono and monitor switches nearby.

Did little paper strip labels and glued to panel, matching the finish and
lettering colors. Done with ink jet on plain paper; coated with artist's
fixative. Not a perfect visual match, but very close and easier in this
instance than some sort of press-on transfer.

==

After at first thinking it not possible, was finally able to break apart
the TB switches and remove the latching assemblies, then reassemble -- all
while the switch was still on the board. I prefer not to have a latching
TB switch -- much more comfortable to push-and-hold/immediate-release the
TB. Double-tapping a latching switch is annoying for TB.

==

Dropped in 2134s in the right and left hand master section. Outputs probed
for HF oscillation; everything seemed fine, though aux sections might have
some gunk. Replaced headphone jack (normalling jack in the control room
monitor circuit). Much improved sonics, even doing nothing else.

==

Looked at 1KHz square wave into mic inputs all the way through to program
outs. Initial sense of feeble LF now quite visible as downward tilt on
trailing edge of square wave. Got worse and worse the farther along in the
path. And the LF restriction seemed quite "hard" in that dialing in a few
dB of bass boost via EQ would mostly level out the slope, but the last
little bit always drooped, no matter what.

So, replaced coupling caps with much larger Panasonics and bypassed with
Wima films. Program summing amp input caps replaced with Black Gates.

No more LF loss. Squarewave nice and level all the way through. Top end
clean. Sounds relatively sweet even with a lot of top end dialed in. Not
exactly warm, but acceptably neutral. Way better than what it was (thin
with a layer of fuzziness).

==

Beware slip-on jumpers! A couple of the submasters were not passing
signal, or would cut in and out as the board was tapped. Brushed against
bus select jumper and signal came through loud and clear. Removed/inserted
several times to firm up connection. Might go ahead and hardwire these
jumpers. Places where aux send pre/post jumpers are used should be checked
as well.

==

Replaced insert (normalling) and line in jacks. Replaced mic XLRs on those
channels I'll use to supplement my outboard preamps. Those XLRs looked
pretty bad and didn't want to clean up. Might be okay, but would rather
not risk it.

==

Added EQ bypass switches to standard input modules where the EQ is
otherwise in the circuit all the time. Nice to have that option as the
kind of work we do does not require EQ a fair amount of the time.

==

Added mechanical coupler on master L/R program faders so that they
cross-track perfectly all the time, at least to the limits of the fader
electrical tolerances. This consisted of a precision drilling through the
plastic fader caps. Hole sized for a tight fit of a 1" 4-40 machine screw;
nut applied between the fader caps to provide mechanical support. Works
great! No more inadvertent L/R channel imbalances, minor as they typically
might be.

Actually, this is pretty important as we do the 3-fader mid-side decode,
and to work optimally the Side-minus and Side-plus signals need to stay
perfectly equal in their levels. Locking the master faders means one less
place to have to tweak for this.

Also, board fades are a lot easier.

==

Aux send 1 had 7 volts DC on the output!! New caps and input transistors
cured this.

==

Still to come..,

Replace main PS with new 18V International Power linear. Use original PS
as "hot backup", the two supplies wired together via diodes. Drop through
diodes will then put the main rails at 17.4V; the backup rails at 16.4V. I
assume I can get away with these slight voltage unders/overs.

Socket rest of ICs with machined sockets and add 0.1 decoupler caps to
each op amp supply rail throughout. Upgrade to handle "faster" chips. (47
pf in series with 100 ohms across op amp inputs, per Jim Williams.)

Investigate noise floor and find optimal gain staging. Dynamic range is
probably around 90-95 dB, but I'm only using 70 dB of that (0 dBFS at
A-D). Most likely can push the console 5-10 dB harder, but haven't wanted
to get into higher distortions. LED ladders on Delta match metering on
external recorders/converters quite closely. That is, 0 dBFS on an
external box is all LEDs lit on the Delta, both in and out.

==

Many thanks to Jim Williams, Scott Dorsey, Stephen Sank, Monte McGuire,
Paul Stamler, Phil Allison, and others who provided valuable advice in
recent months. This project would not have been possible without your help.

==

That's it for this update. Too long? Not enough?

Frank Stearns
Mobile Audio
--