• Please review our updated Terms and Rules here

Wrangling an Apple ADC Studio Display onto a G3 Powermac DVI (Part 2)

RickNel

Veteran Member
Joined
Apr 24, 2009
Messages
641
Location
Canberra, Australia
CONSTRUCTION ISSUES

1. DVI and ADC connectors.


ADC is fully obsolete and DVI increasingly rare. I needed right-angle through-hole sockets of both types. On a lead from one of the DIY blogs I found them available at www.surplusales.com at a good price.

2. Prototype board.

Sadly DVI and ADC connectors are built on a 1.9mm grid-pattern for which no standard proto-board exists. The nearest is a 2mm grid known as "Euro-board" among other names. Only one specialist supplier in Australia carried a board manufactured in Germany by Roth-Elektronik. A small board cost me more than the G3!

In order to seat the connectors in that board, I needed to drill out the holes on the board from .5mm to 1mm. This gave just enough play for the difference between 1.9 and 2mm spacing. But it also halved the area of the solder pads, and left most pins out of contact with their pad, adding to soldering problems.

I knew I would not be using the analog signals so I pulled the 4 analog signal pins surrounding the ground blade back through the connector bodies and removed them, leaving only the GND blade connectors in place. This saved some drilling and soldering.

View attachment 10216

3. Wire selection.

The pin layout meant that there was hardly a single straight-through connection, but an almighty ratsnest. Twisted pairs for the video signals (8 pairs) would have been desirable but proved impossible for me to solder. Steel multi-strand wires with any spring in them
were unmanageable in the tight spaces, even under high magnification. Loose strands were uncontrollable. In the end I used single-strand insulated copper, recycled from standard telephone hookup wire. The copper could be formed into shapes to get around corners so it could be held down while I applied heat from the iron.

4. Soldering hell.

Several factors made this the most difficult soldering job I have ever fumbled through. The copper wire and the long pins of the 90-degree connectors sank a lot of heat. The reduced area of the drilled-out solder pads meant that the remnant pads often melted before the connecting wires or even before the resin-core solder wire I was applying to the joints reached fusing heat.

I minimised the number of solder points and wire crossings as best I could by running a common ground loop, using a bare copper wire. On second thoughts I separated the return line for the 24DC power circuit, but I'm not sure that this was necessary.

Due to closeness of the pins, several times the heat would spread across pad zones, and solder would flow between an existing joint and a new joint next to it. In a few cases, I had to push a 3mm stub of wire into the hole beside the target pin in order to get a firm contact to solder. One pin, the TDMS clock on the ADC (pin30), receded so far into the plastic mount under heat that no contact could be made in the hole. I had to create a new path by adding a wire from the component-side of the ADC connector and feeding that through the protoboard and across to its connection on the DVI connector pin.

Many iterations of continuity checking were required, tracking down shorts and opens, resoldering, cutting, solder-sucking... Finally all contacts were according to plan.

The solder side of the adapter is not a pretty sight, but it functions.

View attachment 10227

5. Hotplug sensor.

One special connection is required in order for the DVI video card logic to recognise that a monitor is attached. In DVI this is pin 16, labelled "hotplug". ADC uses the USB monitor feedback for that purpose. So the DVI side of the adapter needs to be pulled up to logic high in order for the video card to start communicating with the monitor.

This is handled by bridging the DVI +5v pin 14 to DVI pin 16 via a 1Kr resistor. That DVI +5v is otherwise not connected to ADC.

6. Soft power

On a true ADC connection, a power button on the ADC Display functions not to turn the monitor itself on or off, but rather as an extension of the host soft-power button circuit.

The ADC connector carries a soft-power signal on pin 13, but I have found no information as to how this can be connected to a non-ADC host. It bypasses the USB signal path that is used to invoke video adjustments. If anyone can tell me how the soft power connector should work, it would be nice to get 100% ADC compatibility.

7. USB

The USB hub on my 15" Studio Monitor is fully functional but the hub on the 17" Studio Monitor is not recognised. Any suggestions why this might be would be welcome.

Continuity tests on the ADC cable suggest there may be a break affecting the USB signal alone. I suspect that the position of the USB pins at one corner of the 30pin matrix on the connector may cause wires at the connection to break under fatigue after years of flexing. The ADC plug includes a 90-degree flexing hinge between cable and plug, which could induce that cumulative effect. If I can discover how to open the plug safely I'll check that out.

As usual, the Apple user community forums just say "It happens - get a new one".

8. Power to the adapter

The G3 has a switched AC port on its main PSU. It made sense to connect the adapter PSU to this so the monitor would switch on with the host. The G3 cabinet is quite tight. An AC lead would need to pass into the cabinet, along with the DVI lead and a USB lead, to the input side of the adapter, inside the case. The least destructive method was to cut out, from the base ventilation grid area, a port large enough for the DVI connector to pass through. Since this port will also be carrying a full AC power lead, it is essential to pad this hole to protect against cable fraying. I used 1/4" plastic tubing, sliced lengthwise, to form a grommet protecting the metal edges of the port. This is how the inside of the case looks with everything installed (case open):

View attachment 10228

First choice for PSU was to get the cheapest available 24v DC 1Amp powerpack, available online. I got one for $8 including postage (from Hong Kong). The power was sufficient, but cheap powerpacks designed for isolated laptop powering often neglect EMS precautions. If connected to an earthed appliance, in this case via the common ground and shielding of the video and USB cables, they emit offensive levels of radio interference.

Second thought was to construct a small linear PSU using 24v transformer, diode bridge and a couple of smoothing capacitors. On investigation, the components costs and need to provide robust insulation for the AC mains input made this a less attractive option.

Best and final option was to lay out $30 on a regulated 24DC wall-wart power unit with good insulation, compact size and negligible radio emissions.
 
Back
Top