N-BCAM Head (A2083) Manual

Contents

Description

The N-BCAM Head (A2083) is a LWDAQ device built on a rigid-flex circuit board that provides readout for one ICX424 image sensor and control of two laser diode light sources on a Dual Laser Head (A2074C/D). The circuit fits into an N-BCAM chassis as shown below. Connection to the A2074 is through a flex extension.

Figure: Black N-BCAM Head (A2083A). The A2083A is the main circuit board with RJ-45 connector, flex circuit to ICX424 image sensor, and flex circuit to flash the two lasers on a Black N-BCAM Side Head (A2074C) mounted on the near wall. The left-hand laser, as seen from the front of the BCAM is source element three (ON3) and the right-hand laser is source element four (ON4).

The main rigid section of the circuit board sits in the bottom of the chassis. A smaller rigid section holds the image sensor on the back wall of the chassis, connected to the main rigid section by a short flexible circuit. A six-way flex cable emerging directly from the bottom side of the main ridig secion plugs into a dual laser board on the front wall. For N-BCAM assembly instructions see the N-BCAM Assembly Manual.

Figure: Assembled Blue N-BCAM. The A2083B is inside with a Blue N-BCAM Side Head (A2074D). In both types of BCAM, the central source is element number four (ON4) and the outer source is element number three (ON3).The serial number on the lid identifies this device as destined for the ATLAS New Small Wheel.

The N-BCAM provides two laser-diode light sources. One is close to the center of the front face and the other is close to the edge. The outer source we flash with source element number three (ON3) and the inner sourcde with element number four. Looking at the front face of the Black N-BCAM, the outer source is one the left, but in the Blue N-BCAM, the outer source is on the right.

Design

Operation

The A2082 is a LWDAQ device of type ICX424 (6) when we read it out one pixel at a time, and device type ICX424Q (7) when we read it out with quadruple pixels. When we flash one of its lasers, we can use device type ICX424 (6) or ICX424Q (7), because these have the same allocation of bits for flash jobs. The A2071E and A2037E LWDAQ Drivers support both device types, but you will need firmware version 4+ for the A2071E and 16+ for the A2037E.

The table below gives the sixteen-bit commands we can send to the A2086 to turn on one or more of its light sources. As we turn on more than one, they must share the available current. When we calibrate lasers, we turn on only one laser at a time to make sure that we are driving the laser to full power.

Looking at the front face of the BCAM, the outer laser is ON3 and the center laser is ON4. For the Black N-BCAM, this means that the left laser is ON3 and the right laser is ON4. But in the Blue N-BCAM, the right laser is ON3 and the left laser is ON4.

Figure: Assembled Blue N-BCAM. The serial number on the lid identifies this device as destined for the ATLAS New Small Wheel.

The arrangement of the two lasers on the front of the Black and Blue N-BCAMs is consistent with the Black and Blue Azimuthal BCAMs (A2048L/R). Command bits DC12 (ON3) and DC13 (ON4) turn on sources three and four respectively. The logic on the A2083 directs ON3 to signal L2 and ON4 to signal L1 (see S2083_2). The signals L1 and L2 turn on LD1 and LD2 on the Dual Laser Head (A2074C/D). Thus we see that ON3 is LD2 and ON4 is LD1. If we flash a laser on the A2083 while capturing an image from one of its own image sensors, we must be sure to use device type ICX424 (6) or ICX424Q (7). Otherwise, we my encounter the White Rectangle Problem.

Quiescent Current

Here is the current consumption of a fresh A2083A in various states.

Figure: Black N-BCAM Head Current Consumption with All Auxiliary Circuits attached..

The A2083 with image sensor and two lasers is supported by the Analyzer Tool V18+.

Image Dynamic Range

We discuss the image intensity dynamic range of this readout circuit in our Bar Head Manual.

Radiation Tests

All the components uses on the A2083 are also used in the same way on the A2082. We discuss the radiation tolerance of these components in the Bar Head Manual.

Development

[22-APR-15] Circuit diagram first draft done. Initial radiation tolerance tests complete.

[02-JUN-15] We have the A208301X printed circuit board. We measure all hole spacing and flex locations and they match the mechanical drawing we designed from. We load the board into a chassis. We find a mismatch between the image sensor hole spacing and the holes on the chassis. But it looks like the error is in the chassis, not the circuit board, because the circuit board matches the drawing.

[03-JUN-15] We have the A208301X fully loaded with parts. We capture images. We solder a flex cable to the flex extension and flash two lasers. There appear to be no modifications necessary to the layout. We increaes the DA and DDA delay times in the firmware so as to make the DA and DDA delays 125 ns and 250 ns.

[09-SEP-15] A couple of weeks ago we received 108 A2082A assemblies, populated with all SMT parts and built on the A208301A rigid-flex printed circuit board delivered. The boards work well and load into the Black N-BCAM Chassis. We receive A208301Z circuit boards, with blue solder mask, 20-mil thick, all-rigid to test the Blue N-BCAM layout. All is well except we must invert the order of the pads on the six-way flex plug. We do this and send A208301B Rev2 files to a manufacturer.

We assemble two Black N-BCAMs and measure current consumption. We create firmware P2083A02.abl with the lasers arranged for ON3 and ON4 as for the Black Azimuthal BCAM.

[10-SEP-15] We increase solid-wire root cable length from an A2071E to a Black N-BCAM and take images. We obtain sharp images at 140 m total length, but none at 160 m.

[23-JUN-16] We have a batch of 440 A2083A circuits assembled in February of this year. We just made our second set of 100 Black N-BCAMs. In doing so, we discovered a dozen circuits all in the same package upon which U9 looks something like the example shown below.

Figure: Burn Marks Around U9. Every circuit with these burn marks has a broken via beneath U9.

All the circuits with discolored solder and solder mask around U9 have the same electrical problem: U8 and U9 do not receive −15 V. Beneath U9 is a via that brings −15 V power to U8 and U9. This via is broken in the faulty circuits. Neither U9-4 nor U8-4 are connected to −15 V, nor to one another. We correct the circuit by running a wire from our −15 V test pad to U8-4 and U9-4.

[10-JAN-19] We find that the A208301B circuit board fails to connect H1C, H2C, and OUT to the test point pads on the image sensor portion of the board. This omission does not affect the function of the circuit, but does affect fault-finding in broken circuits.