[Hbi-svt] HFCB Design Considerations

[Marc McMullen]

Hi All,*
*

*
HFCB Design Considerations*

The design of the HFCB is well underway. During the process several 
issues have had to be carefully researched. There have been six issues 
in particular which have been considered, four of these have been 
resolved. The remaining two are being worked on by Hall B 
Instrumentation and Engineering in conjunction with our collaborators 
from MSU.

The last of these issues is the most critical as it determines the 
layout of the design. Currently, the design is consists of a single 
flexible circuit. This circuit addresses three aspects of electronics 
readout, both top and bottom ASIC readout and first level disconnect 
without the use of any connectors to join these areas.

Michael Merkin has proposed a module assembly process which is done in 
two halves. This assembly process accommodates the current module 
assembly tools at MSU. However, the assembly process may not be suitable 
for the current HFCB design using the tools currently onsite at MSU.

Hall B Instrumentation has researched possible strategies of separating 
the areas of the HFCB and using connectors to attach the top hybrid area 
to the bottom hybrid area. The proposed connector would have to be 
densely populated due to the limited area and number of circuits needed. 
Additionally, high voltage requirements for the circuit exceed the 
limits of the available connector products.

Hall B Instrumentation has requested Michael Merkin to review the 
current HFCB design as well as the module assembly process at MSU and 
determine what equipment would be required to augment the assembly 
process in order to accommodate the current HFCB design. This 
information can be used to weigh the cost of design change versus 
assembly process change, providing a solution to this critical decision.

In evaluating design change the following must be considered:

· Creating separate areas of the HFCB will require more than one circuit 
design; additional manufacturing cost and NRE should be considered.

· Incorporating friction based connectors may add impedance to the 
circuit, in addition to the possibility of having to design a special 
connector.

· Connectors used in the bore area of the solenoid would present an 
additional point where a problem may occur. Due to the physical 
dimensions of the area and the spacing of the components in the area, 
repairs could become time consuming.

1. _LV trace widths and calculated voltage drop_:

The LV trace widths along the length of the cable between the top hybrid 
area and the L1 disconnect is defined by the number of necessary low 
voltage channels and the overall width of the HFCB cable. Two channels 
are needed, each requiring a positive and negative voltage. The given 
width is 38mm. Calculations were made, determining the voltage drop of a 
trace between 250mils and 332mils range between .02V and .014V for and 
internal trace made of half ounce copper for a 500mm length. The design 
will incorporate a LV trace width of approximately 300mils with 74mil 
space between each trace to achieve a voltage drop of <.02V.

2. _The addition of a LV regulator to the L1 Disconnect_:

A voltage regulator will be added to the design to provide stable 
voltages to the HFCB at to the L1 disconnect. There will be 2 channels 
of regulated voltage, and two separate regulators. Considerations for 
these regulators will include stability, noise, and power dissipation in 
the form of heat.

3. _The calculated impedance of a microstrip in the design_:

The impedance of the LVDS microstrips currently uses a prepreg thickness 
of .062mm (2.44mils). For polyimide which has a dielectric constant of 
2.5, the resulting differential impedance is 53.28 ohms. Doubling the 
prepreg thickness will result in a differential impedance of 78.1 ohms.

4. _The process of thermal transfer from the ASIC pad to the copper tab 
built into the module_:

The design will provide for an arrangement of the stack-up using thermal 
vias and copper floods to conduct heat from the chip through the 
polyimide substrate and the FR4 stiffener to the copper tab. The design 
then calls for the electrical isolation between the HFCB and the cooling 
system. Dyconnex has been requested to investigate a material which will 
be thermally conductive yet still maintain electrical isolation between 
the polyimide and the FR4.

5. _The overall length of the design_:

The current length of the HFCB (525mm) is the max length possible for 
Dyconnex using the largest panel size available. Any changes in this 
length has to be determined before routing on the signal layers, changes 
to the length after this point will result in having to re-route the 
whole cable.

6. _Fastening the HFCB to an SVT module_:

Michael Merkin of MSU has provided design guidance for the basic module 
of the SVT. This design calls for the module to be built in two halves. 
This is to accommodate MSU’s current assembly process. These halves 
extend to the copper cooling tab used for thermal conduction from the 
HFCB. The halves are to be joined with fasteners at the HFCB end, which 
will also and align the HFCB to the module. Without a modification to 
this module assembly process the HFCB design would have to be made as 
two separate circuit boards which would be joined after module assembly.

In conclusion, as stated previously these design considerations have 
been researched and for the most part resolved. Hall B Instrumentation 
will proceed with the current design. In parallel with this, we have 
requested information from MSU on the possible modification to the 
current assembly process or an MSU electronics readout design plan. This 
issue needs to be resolved in order to produce a first article as soon 
as reasonably possible.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: mcmullen.vcf
Type: text/x-vcard
Size: 252 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/hbi-svt/attachments/20100528/52f3255f/attachment.vcf