DOM Hub Service Board (DSB) Revisions

The DOM Hubs with more than one DOR card require a common clock to be fanned out through the DOM Hub Service Board (DSB). The card distributes a common clock, a phase coherent one-pulse-per-second time tick, and under certain circumstances, a time string from an external source

Revisions 1.0

Mechanical

No particular mechanical modifications have been suggested.

Electrical

Internal Sync. push-button

The four-pin pin-out for the push-button switch was assigned in the schematic before a physical part was selected for the layout. Two of the pads on the selected part comprised the support frame pads formed from a single piece of metal. The frame shorted out the switch contacts. The Rev 0 work-around is to sever at least one of the traces that connect the switch contacts to the frame contacts.

The Rev 1 schematic will have a revised pin assignment.

Harmonica Connectors

The DSB schematic indicated that a single 8-socket, ganged RJ-45 connector is used for signal output. Two 4-socket, ganged RJ-45 sockets were substituted into the layout during the layout process. Functionality remains unaltered.

The Rev 1 schematics will be revised to indicate the alternate part selection and pin-out.

Duty Cycle accommodation

The signal processing of the one pulse per second input of the DSB accommodated the typical 20 us wide pulse of an external source (like a GPS clock), however, a 1 Hz, 50% duty cycle clock was not adequately accommodated. The trailing edge of the signal output from the DSB chattered, and returned to the neutral value prematurely. The work-around for Rev 0 is to remove C40, and move R63 into its place. The input is thereby converted from AC coupled to DC coupled.

The Rev 1 schematics will have a revised input circuit.

Input Terminations

R66 and R68, 50 ohm input termination resistors, in the layout were 0603 size due to an error in the footprint attribute of the original netlist. The Rev 0 workaround requires the larger parts 1210 parts to be tacked onto one of the two pads, thomb-stone fashion, with the other end connected to the pad by a short piece of 30AWG jumper wire.

In Rev 1, the termination resistor footprints for these parts will be corrected in the schematic, and verified in the layout.

Transformer Inputs

The original schematic and netlist identified a particular ethernet transformer which was not readily available. The available replacement had a different footprint,. The Rev 1 schematic has been modified to reflect the changed part.

Revisions to the DSB are not yet complete. G. Przybylski, LBNL, for IceCube