D-Control

D-Control, known internally as Buckley, PC II or ProControl II, put a large Pro Tools control surface on a configurable stand. I designed the stand and took over Matt Cho’s mechanical work early in the program. The responsibility grew into console integration, manufacturing recovery and sustaining changes. Matt’s earlier work remains part of that history: the surviving 2002-2003 PCB interface drawings carry his authorship. Electrical design, software, industrial design, purchasing, quality and production were shared team responsibilities.

One console, several configurations

The product combined a main unit with separate 16-channel fader modules. The D-Control guide describes expansion to 80 faders: five fader modules plus the main unit. Each main or fader unit had its own power connection. That modularity made the stand and the interfaces between neighboring units central engineering problems. A configuration had to work as an assembled structure, with the surface controls, meter bridges and cosmetic edges meeting correctly.

The main-unit assembly drawing shows the physical layers: a lower pan, side structure and rails, internal supports, and upper control-panel assemblies. PCB dimensional-control drawings located board outlines, mounting holes, controls and connectors within that package. They defined mechanical interfaces to the electronics; they were not circuit-design ownership.

The stand began with feature tradeoffs

The 2003 correspondence follows the stand before the production problems. In April I extended the heel of the foot cover to support the rear steel foot when the shelf was loaded, departing from the earlier industrial-design cutout. By May, unresolved price points were holding up the stand and shelving work. The June feature review put construction choices in front of the team: a separate bolt-on heel, a thick flat foot with a formed saddle, two-piece legs, shelf fasteners and cable routing.

Those were interacting choices. A cable path inside the leg changed fabrication work; concealed shelf fasteners affected the shelf structure; the heel had to support the loaded arrangement while fitting the intended appearance. The June exchange records proposals and cost questions, rather than a single frozen design. The later stand packet carries those interfaces into part and assembly definitions. A July 6, 2004 drawing, for example, releases a stiffened main foot plate with a welded reinforcement under ECO 6325.

The stand kept changing as the team learned from assembly. A July 7 recap of the mechanical changes described a welded foot stiffener to address reported flex and larger openings through the top brackets and feet for cables. It also recorded a remaining compromise: the cable exit at the foot provided no shear protection if someone moved the console. The recap called the changes effective immediately while warning that parts might take another month to become available.

Configuration work continued after the initial release. In August 2004, Tom Oiwa questioned the C-Mid and D-Mid crossbar lengths and the precision of their dimensions. I explained that the crossbars were intentionally shorter than the overall dimensions of the units they supported. The exchange is useful evidence of design intent and manufacturing review. Its four-decimal calculations do not establish a measured system tolerance.

In the follow-up I agreed to three decimal places, but asked which feature Tom wanted to loosen and said I did not want parts less accurate than those already being received. The custom-configuration rails also had a separate handoff: I designed the parts, Tom completed drawings and submitted ECO 6371, and Vicky Moreno described her team’s planned pre-staging and custom stand builds. The surviving ECO requests Rev A for four C-Mid and D-Mid front and rear rail parts. Its blank approval and effectivity fields leave their release and delivery unresolved.

Custom requests also had a defined approval boundary. On October 7, 2004, Richard Entenman requested B-Mid supports for a customer’s three-B stand arrangement and proposed an in-house test before shipment. Neal Breitbarth corrected the premise: only configurations on the chart supplied to production had engineering approval. Vicky Moreno confirmed the distinction between those configurations and custom stand requests. That exchange records a request and an approval boundary; it does not establish a tested or delivered three-B stand.

Later B-Mid work exposed a different handoff problem. In December 2004, Tom Oiwa reported that MASS read the rail models as 84.478 inches until regeneration changed the reading to the intended 82.83 inches. I identified unregenerated files and asked whether the pan mounting holes had been adjusted for the shorter length. Tom described adapting the existing B-End models; the B-End PDFs I supplied still defined the longer parts. He later reported sending cleaned files to Terri Merrell with my help. Terri required the ECO, files and PDFs to go through DOC IN for signatures, and Vicky Moreno was still asking whether the ECO was complete on December 13. This records diagnosis and a reported file handoff; it does not establish approved release or delivery of corrected rails.

The later drawing package makes the next definition explicit. ECO 6581, dated December 9, changed the B-Mid length from 82.83 to 82.669 inches. Tom prepared the revised front and rear drawings, and I checked and approved those sheets. The ECO also requested a revision to the associated kit. Its approval and effectivity fields remained blank in the surviving copy, and the December 13 email still requested approval. Corrected drawings are therefore documented; production release and delivery are not established by this package.

The stand was still being defined at the component level as its cost model developed. The July 2003 and January 2004 sheets both describe a one-main, one-fader configuration, but list different crossbar and bracket quantities and a more detailed list of support parts. The later sheet still marks several prices as estimates awaiting quotes. These records show the configuration taking shape; they do not establish a production-cost reduction.

Two surviving configuration worksheets show changing part definitions, not a verified production-cost trend. Part scope and total-cost bases differ; these sheets do not establish realized savings or prove that the listed variants were built.

The January 13, 2004 Mass Precision pricing sheet lists unit costs for quarterly releases against annual planning quantities. Several of those prices recur in the January 28 configuration model, while the crossbar bracket carries a separate estimate pending re-quote. This connects the working model to supplier pricing without establishing what was ordered or delivered.

The stand’s modular structure is visible inside the crossbar assemblies. The A crossbar’s Rev1 bill of materials lists four different panel parts: a small-slot and a slotted version for both the main and fader sections. Rev A lists two of each slotted panel instead, retaining two crossbar extrusions. The exploded drawing shows those four panels fastening between the extrusions. This is a change in the documented assembly definition; the surviving lists do not establish the reason for it or a manufacturing saving.

The stand also needed the right revision of each part at assembly. An undated parts-tracking sheet lists Rev 3 for the D front and rear pan supports, but records that Rev 2 had arrived. The front-support drawing shows a concrete difference: its February 20, 2004 revision adds a second visible slot while retaining the same overall length. The front-support drawing’s March 31 table identifies Rev A with Rev 3. These records establish the intended definition and a reported receiving mismatch; they do not show when replacement parts arrived or that the discrepancy was closed.

Revision control extended beyond the stand. A February 2005 supplier tracker kept Digidesign’s drawing revision, the revision ordered at Mass Precision, purchase-order quantities, production balances and quantities still due in separate columns. Several console parts still showed different drawing and ordered revisions. The March correspondence used that comparison to chase missing files and correct a part-number typo. It records the work of aligning design and supplier records, rather than a completed shipment or a closed revision discrepancy.

Geometry had to survive assembly

The manufacturing instructions make alignment part of the build sequence. Both main and fader assemblies started with screws loose enough for adjustment. Before installing the covers, the process called for flatness and squareness checks across the top, front and inside surfaces. Assemblers were to loosen, realign and recheck a unit that failed those checks. The main-unit instructions also controlled harness routing and delayed the cosmetic overlays until after burn-in.

The installation guide used the same principle at system scale. Alignment pins joined neighboring units while their attachments to the rails remained loose. A ratcheting nylon strap closed the stand legs; if necessary, it also brought the units together before the remaining fasteners were tightened. The guide explicitly kept the strap clear of switches and encoders. Adjustment was designed into the assembly method.

Service access depended on those interfaces too. The service guide describes reaching internal assemblies through removable control panels and separating cable connections in sequence. To reach the side plates on a console in its stand, it calls for loosening the stand joints before moving the legs apart. The structure had to support disassembly as well as the initial build.

Inspection then tested whether the parts and process delivered that intent. Tom’s July 14, 2004 report measured ten completed main units and ten fader units with a precision square and thickness gauges. The widest top-panel gap reached 0.028 inch in both groups. He was collecting data to help choose an allowable manufacturing gap; the report does not establish an adopted acceptance limit. On July 30, he called for further squareness checks with units both separate and installed before ordering fixtures. The inspection photographs locate the interfaces under review, without proving that every condition was resolved by one change.

The inspection photographs recorded different local conditions: one panel stood above its neighbor, a panner panel showed unequal gaps, and another joint was marked as meeting well. Together with the underside assembly views, these images locate the interfaces being checked. They do not establish a single cause or a common resolution for every condition.

Recovering the manufacturing process

The first 42 steel fader pans exposed a basic forming problem. PEM hardware insertion stretched the sheet locally and bowed the pans enough that more than half were rejected; the side-plate interfaces would not line up. My immediate response was to demonstrate cold work over a table edge to recover flatness. The supplier correction redistributed the insertion sequence and added flattening after insertion. Immediate salvage and the process change were separate parts of the recovery.

The June 2004 manufacturing and DFM campaign was estimated at two months; my retained account records completing it in 2.5 weeks, including home-built assembly jigs. The immediate recovery got usable parts and tooling into the process, while fit and production work continued afterward.

Prototype board containment

A separate prototype-board investigation found V-2 material in 90% of one unnamed vendor’s prototype population, where V-0 was required. Four shop logos exposed uncontrolled subcontracting. We quarantined the affected prototypes, checked customer exposure and terminated that vendor for prototype work. The retained correction record states that no affected boards shipped. This was a bounded containment result, separate from the pan and stand work.

Sustaining work after launch

The engineering responsibility continued into installed systems and production changes. In June 2004 I volunteered as the mechanical group’s point person for assembly and installation issues, working with installers, QA and manufacturing. The 2005 correspondence shows why that role still mattered: a successful part test did not automatically settle rework, documentation or release.

The replacement side plates had a prototype definition before their production transition. My January 5, 2005 drawings identify separate left and right parts, both at Rev 1 and released to prototype. They replace the older fastening detail with a standoff at a formed recess.

For those replacement side plates, Brandon Cammack reported an aggressive manual test on February 15, 2005. He loaded and moved the side cap while driving the screw, damaging the cap while the fastening interface continued to hold. I approved the parts after that report. It was a useful abuse test, not a quantified qualification of every production unit.

A February 22 supplier exchange still treated rework as a feasibility question. Ed Stegall proposed removing the old insert, flattening and forming the sheet, installing the new standoff, and repainting. Mary asked whether that process could convert the old plates to the new design; the exchange was not an instruction to start work.

Getting the change into manufacturing took further work. The team still needed second-operation drawings and an ECO; the supplier’s first rework setup scrapped parts. By April, QA reported acceptable reworked parts. Tom initiated urgent ECO 8045 to replace the old side plates with my 9420-13080-01/-02 designs, and I delivered the drawings and models on April 15. Purchasing, QA, manufacturing and mechanical engineering each owned part of that transition.

The Swedish Television installation raised a different interface question. In April 2005 I reviewed the stand-flange condition and explained that the flange was intended to lie parallel to the underside of the unit. The concerning case was a flange sitting too high and preventing the pan from seating squarely on the rails. The selected correspondence establishes that diagnosis; it does not establish a completed replacement or inspection-gauge rollout.

The plastic changes had their own verification trail. ECO 6310 enlarged mounting holes from 0.200 to 0.250 inch, enlarged their bosses and raised five bosses by 0.050 inch. Deviation 1212 covered interim reaming; the Rev B part record specifies Noryl. On August 1, 2005, I confirmed that the T1 samples incorporated the ECO changes and sent the signed approval to PPI, the molder. That is a documented sample-approval endpoint, not proof that all field fit issues had disappeared. ECO 8197’s side-plate relief remains a separate change.

What the record establishes

These are documented stages, not exact employment or responsibility-transfer dates. I designed the stand and took over Matt’s work early; the precise handoff date remains unspecified. The selected record establishes involvement through August 2005 without making that my final D-Control deliverable.

D-Control shipped and appears in the installed-system archive. The engineering outcome was a modular product supported by detailed part definitions, an adjustable assembly sequence, containment when parts failed and continued production support. The stand, console and cosmetic interfaces had to work together in the factory and again when assembled at a customer’s studio.

Source trail

The eight remaining native stand workbooks distinguish supplier pricing, planning quantities, blank costing fields and leg definitions; their source-state and arithmetic limits remain in the private research record.

The December B-Mid follow-up joins the native model and file-identity discussion to the December 9 ECO 6581 package. That package documents corrected 82.669-inch front and rear drawings and a requested kit revision. Full ECO approval, effectivity and physical delivery remain unresolved; its B-End reuse instruction also conflicts with the earlier B-End length definition.

The later stand check joins four native parts workbooks, the August 25 dimension-drawing transmission, the September notch correspondence and the October 7 custom-configuration approval boundary. The examined record still does not close ECO 6371 or the corrected 48-fader drawing delivery.

The August rail account uses the native C-Mid correspondence and its ECO 6371 attachment. Earlier drawings with the same part numbers describe different crossbars and lengths, so they are not treated as the August release definitions.

The stand revision account also uses the undated parts-receiving sheet and the front-support Rev 2, Rev 3 and Rev A drawing states. Those definitions do not establish replacement-part receipt or closure.

The account draws on the stand and PCB drawing packets, the main/fader manufacturing instructions, the D-Control installation guide, contemporaneous 2003-2005 correspondence and my retained corrections and role clarification. Assembly illustrations, design renders, inspection photographs and installed-system photographs serve different purposes; their captions preserve those distinctions.