Progress Report of Region3


               General: Click on image to enlarge to higher resolution

1) Frame related

The machined G-10 frame parts are expected to be delivered end of June by Atlas Fibre. G-10 scrap material of various sizes (leftovers from the cutting process) is expected to arrive within next week.

There has been a cutting problem with the long boards (they don't know why - they are attempting to figure out why) -  but the width at one end starts off at the correct 6 inches,
but it "tapers" on one side so that at the other end it is up to 40 mils too narrow. It seems to vary a bit from board to board, but 40 mils is the worst case.
Workaround:  The short and long board width of 6" has been reduced by 1/16" (1.5875mm) which is uncritical.  The tapered boards will be routed (CNC  milled) to this width so the boards are squared-up again.
                      The length of the short board have to be shorten by 1/8" (3.175mm) accordingly. The reduction of the frame dimensions does not affect the inner dimensions of the drift chamber.



A1) Reamed hole diameters specification for dowel pin

For alignment and assembly purposes the G-10 frames/parts have several positions for dowel pins. Some are press fit and some are slip fit. So we thought Atlas Fibre would figure out
to which hole diameter these dowel pin holes have to reamed ... (The Machinery's Handbook and others don't provide tables for G-10).
Apparently Atlas Fibre is shifting responsibility to us and we had to figure this out. Details on http://dilbert.physics.wm.edu/elog/Construction/420



A2) Wire Jig






The outer dimensions of  the 8 wire jig boards have been measured with a 40" long caliper with a resolution of 0.5mil (12.7um). The alignment of boards were done using precise 1-2-3 blocks
(flushing jig plate against aluminum profile).  Gage blocks were used for the correct pin-to-pin spacing of  adjacent jig boards (spacing between the last pin of previos board to first pin of next board).

The characterization of the dowel pin location on the wire jigs has to be done with the wire scanner. Last week we made a 20' long extension cable that connects the wire scanner with the motor controller
and computer.

A2) Glue Jig Design
 
The CAD model/design is complete, just finished the translation of the model into technical drawings that will be posted on the web like I did for the VDC design.
The (top/bottom) 71.5"x8" long and the (left/right) 35"x11" short alunimum boards are identical. This job has be outsourced due to limitations of the W&M machine shop.

Technical drawings of Short Glue Jig Board (pdf)
Technical drawings of Long Glue Jig Board (pdf)


Pic below: The 0.5" thick machined aluminum (tool) plates are mounted on a 80/20 aluminum profile frame. In the corners 6"x6" large heat pads can be mounted for accellerating the expoxy curing time.
                  The location of the G-10 boards is indexed by dowel pins.


Pic below: Top view of  the 4'x10' large laser table with the wire frame mounted on the glue jig, surrounded by the wire stringing jig


Pic below: Full setup showing a wire frame mounted on the glue jig, surrounding wire stringing jig ,and the wire scanner above the laser table


B) Epoxy glue tests: Araldite AY 103 + Hardener HY 991. Mixing ratio by weight is 100:40

Lesson learned:
After mixing the resin and the hardener you have ~45min left in order to apply it. Otherwise the viscocity of the glue is getting too high.
A lot of air gets trapped in the mixture and the glue dispenser will not work with this kind of foam. In order to get defined glue dots the glue has to be degassed first.
The size of a glue dot is defined by the duration of the applied air presure ("shot") on the piston inserted in the syringe. As long as the glue is not incompressible the
air shots get absorbed by the foamy glue. For gegassing we are using a centrifuge which is much faster than any vacuum degassing.
We are spinning a 3cc syringe at 5000 rpm for 4-5 minutes for a complete degassing   


Up Left: Resin and Hardener cans                                                                                Up Right: Air pressure driven glue dispenser with syringe and needle



Pic above: Air bubbles trapped in the epoxy mixture                                                      Pic above: Centrifuge used for glue degassing.

We did some epoxy bonding tests (24h curing time at room temperature, 48h recommended by data sheet):
   Kapton foil on G-10:     weak bonding. I could very easly peel off the Kapton foil   
   PCB          on G-10:     strong bonding. I could break it off but with a lot of force. The used PCB has a green solder mask on it that provides a very smooth surface
   25um wire on G-10:      strong bonding. You have to wait at least ~16-24h, otherwise the wire slips through the glue dot.

  G-10 on G-10         :      very strong bonding. We inserted 4mil shim stock pieces in order to maintain a required gap between the G-10 boards
                                         for a stronger bonding, see http://dilbert.physics.wm.edu/elog/Construction/429 for details.

Pic below: Gue tests of G10 with Kapton and PCB


Pic below: Glue test G-10 on G-10 with a 4mil gap in between for stronger bonding.



C) UV glue tests:

For the construction the 281 wires will be strung across the wire frame. Wire position, spacing, and wire angle are defined by the
surrounding glue jig. Each wire will be tensioned with 60g. Rather than having 561 dangling weights we would like to fix/bond the tensioned
wire on the wire jig. This will avoid the situation that the dangling weights will shear off the wires over time.

A wire can be attached on the wire jig in two independent ways:
 - clamp the wire using a fastener (Nylon  standoff + washer)  , or
- glue the wire on the glue bed  using a UV curing glue

W&M posseses an UV light curing system which allows to use glues that cure under UV light within seconds,
see for details here: http://dilbert.physics.wm.edu/elog/Construction/418

Pic below: Details of a wire jig board. A thread is used to illustrate the stringing. One pin to the right a 25um wire has been strung.
                  All wires can be attached using a fastener and/or can be glued on the glue bed (below bottom pin row).



C) Wire hookup 101

How to apply a 60g weight on a fragile 25um wire ? We tried several approaches:

a) Scotch tape the wire and use a standard paper clip. Works for a short time only . After some minutes or hours the wire will slip through the Scotch tape.
b) Bond the wire on a washer and attach a hooked weight. Up to now the fastest method and it is very reliable.
c) Solder the wire to a small piece of a PCB. Soldering is easy and provides a strong bond. However very labor intensive: you have to cut a PCBs into small pieces
    and drill a hole for hooking up the weight.

 

 

2) Electronics:

a) MAD chips

   - radiation hardness:
           http://dilbert.physics.wm.edu/elog/Construction/426

   - worst case scenario and fall back plans:
    a) BigBite is using the W&M design
 
        My personal favorite. The MAD chips will be soldered only once. Requires a high priority status in the
        Electronics Group (Chris Cuevas et al.). To be discussed with Chris tomorrow at our bi-weekly Electronics meeting.

    b) W&M is using BigBite design
        In case the don't have our custom 16 channel boards on time (e.g. due to delay in design/production and/or 200 MAD chips
        got fried by Hall A):
        Use an inexpensive adapater board to hookup the BigBite 16 channel MAD boards.
        Shown below is such an adapter card of the HKS chambers using the old Nanometrics cards. A BigBite adapter board
        would be similar (6V instead of +5V and -5V).

       Disadvantage: "No bells and whistles" (Roger's comment)

 -> Bells and whistles are : adding two chips and one 4-pin connector on a BigBite board which allows the following:
         chip #1 : Setting the threshold per 16 channel board directly using an DAC. Monitor threshold and 3 other on-board voltages

                       => We measured BigBite's external threshold (circuit) setting which involes a non-linear threshold attenuation factor
                               

         chip #2 : Disable/Enable the wire inputs into the MAD chip individually. Great for *remote* debugging, switch off noisy or ringing wires,
                       check/verify channel assignment (hard+software) including the multiple TDC channel assignment due to delay line readout






b) Delay Line Multiplexer

     Setup for the characterization of the digital delay line chips:  http://dilbert.physics.wm.edu/elog/DAQ/18
 
      Graduate student John Leckey measured the delay and time jitter of 43 chips and is currently analysing the data.
      It seems that the overal spread in the delay is in the order of 120ps. The error of a single measurement is +-5ps.  Apparently there is systematic shift of ~50ps
      between the outputs #1,2,3 and #4,5,6 of each chip. Not clear what is causing it but this shift is uncritical anyway (TDC resolution 120ps).
 
3) Flatness Scanner

   next time, preview: http://dilbert.physics.wm.edu/elog/Construction/428

4) Wire Tension Measurement

Yesterday graduate student  Siyuan Yang started on the wire tension measurement project.  We are currently setting up a simple test setup where a wire will be strung with a known weight (tension) and
soldered under tension to copper cladded PCBs. Currently we are using a stack of  small 3/4" x 3/4" x 1/4" thick Neodymium block magnets (3170 Gauss surface field). We would like to replace it with a
stronger and larger magnet from KJ Magnets .

Methode A) A function generator (FG) is hooked up to the wire crossing partially a magnetic field. The function generator sends an AC current (sinus) through the wire. A CCD camera is used to track the resulting
                    vibration/deflection of the wire. Tune the FG frequency to the resonance frequency of the tensioned wire.
 
Methode B) Send a short pulse through the wire crossing partially a magnetic field. Measure induced time dependent voltage, extract resonance frequencies (inc. higher orders) and determine
                    wire tension.       


That's all folks !