Lab 8 Measurements.
Yuri reported on the measurements done in Lab 8 before and after the panel machining on the Gerber. The goal was to discover if the panel bowing was somehow effected by the Gerber machining. All the panels machined for the P3 and P3' (PNPI kit) prototype have been measured before and after the Gerber machining using the "sagitta" method, i.e. tensioning a string between the panel extremities and measuring the sagitta between the string and the panel itself. The measurements and documentation have been prepared and performed by the TD QA/QC group. Yuri Plotted the results of the maximum sagitta before and after machining, finding very little difference. The maximum value for the sagitta typically assumes values between 40 and 80 mils over the lenght of the panel. One panel only has a sagitta as bad as 140 mils. The difference in measurements before and after the Gerber machining is a nice gaussian, centerd at 0 and with a sigma of 5 (or less) mils. The conclusion is that there is no extra bowing introduced by the Gerber machining and that half of the panels (7 out of 15) have a maximum sagitta of 50 mils or more.

Lessons Learned in the P2' Assembly. 
Oleg summarized the experience on the assembly of the P2' chamber, discussing suggestion for improvements when problems where encountered. He subdivided his presentation according to 8 assembly steps:

Panel Drilling and Cutting (AXXIOM)    OK
Strip Milling (GERBER) 

• A strip groove depth of 4 mils is not enough, producing many skips (regions where the copper was not cut through). Running the cutter again at a deeper setting effectively produced a groove whoich was 7-10 mils deep and 20 mils wide. Probably 7 mils is optimal, and that is the depth adopted for the P3 panels and the subsequent production.

• The machine was leaking, failing to produce the proper epoxy mixture (50/50 mix of resin and hardener). In almost all the panels there was a bad area of approximately 20-30 '' where the glue took almost a month to harden. It is clear that we need to pay more attention to the operation of the machine. Since then we fixed the machine and put the vendors on a standing service contract.

Chamber Dry Assemly

• A dry assembly was performed before sending the panel through production. We observed a bowing of approximately  70 mils on the top panel (the last of a 7 panels stack) which is a favctor of 4 gratern than what is acceptable for this chambers. A bowing panel exerts a force of approximately 1 lb/mil. According to a model developed by Nelson, 6 panels can exert on the bottom panel a force of approximately 200 lb through the buttoms.  For forces of this magnitude, the buttoms surface area is so small that local deformation in the panel can take place. As a result the buttoms were re-designed. The buttoms for P2' are a simple piece of plastic with a wider area, while for P3 and the Aging Test Chamber a better designed will be used with the intent of adopting it for the final production.

• The first problem observed is that the wire spools are not long enough to allow the completion of winding on a P2 panel. We will specify our wire procurement in order to have a lenght of 2.8 km of wire in a spool, to avoid changing the spool during winding.
• There is an indication that the wire diameter changes by 3-4% both from direct microscope measurements on the wire and tension measurements on the anode panels. This is slightly more than the specs to the company (2% from SYLVANIA).
• Whenever a skip or a "double wind" happens (typically once/twice per panel) and we stop and "back-track" the winding machine, we loose tension for approximately 10-15 winds. This information will enter our procedures, to be sure that we always back-track properly whenever the winding stops.
• The adjustment of the combs for a preoper wire positioning is rather crude at the moment. Evgeni designed a setup which should make the adjustment much more reliable in the future. This new setup will be tried on P3 first.
• The winding speed can be increased. During the winding of the last of three anode panels, the speed was increased by 10%, with no major difference in the final tension measurements.

• During the process of gluing the wires to the anode bars, in some instances the wires were not in contact with the pads, making the subsequent soldering risky (i.e. in 3 instances, the solder flowed directly to the pad without catching the wire). In the last panel the wires were pushed to be in contact with the pads during the gluing. The problem is worsend by the panel bowing and can be solved by a proper design of the combs. No action has been taken at this moment.

Automatic Soldering Machine

• We observed more than 10 bad soldering joints with FR-4 burning. The major reason is that the distance between the panel and the focus of the automatic soldering machine varies up to 30-40 mils, and this is the main reason for unstable soldering. The Purdue group will investigate a control to utomatically adjust the distance between the panels and the focus of the soldering machine.
• In the longer anode bars, not all pads where properly covered by gold plating. This may be due to the company not owning gold-baths long enough to accomodate our longest anode bars. Other vendors will be investigated before moving to the full production.

• Tools need to be designed for a faster bending and soldering of the caps and resistors.

First Results from Wire Position Measurements
Sergei presented the first results from the CCD camers used in the Tension measurement machine to determine the position of the wires.  After proper illumination and focusing, the image is composed on a grid of 512*480 pixels, where each pixel is 7.6 microns. Sergei presented a number of differential distributions (some of which are still confusing to me! ). The end point is that when all the information was summarized in a plot showing the distribution of  distances between wires, he obtained a gaussian centered at 416.5 pixels (3.165 mm) and a sigma of 4.32 pixels (0.032 mm) with almost all the outliers within +/- 5% (+/- 160 microns) from the average. The goal was an interwire distance of 3.161 mm, with a spread in distances of  +/- 200 microns.  Sergei also presented a measurement of the wire diameters. The wire diameters appears to vary between 45 and 90 microns, which is clearly absurd. Probably  the effect of proper focusing is preventing a direct diameter mesurement.
 

Two Week Schedule
Giorgio Presented the following 2-weeks schedule:

P2'                                                                Resp.
==
Panel Electrical Meas.                                      O.Prokofiev, B.Jensen
Panel HV Test                                                  O.Prokofiev, B.Jensen
New Buttons Installation                                   O.Prokofiev
Panel Cleaning                                                  D.Northacker, B.Jensen
Chamber Cart Assembly                                  G.Smith, E. Borissov

P3 - Lab 8
==========
Panel on Cordex                                              Y.Pishalinikov, G.Kobliska
Panels to MP9                                                 G.Smith
Panel Bowing Meas.                                         Y.Pishalinikov, G.Kobliska

P3 - MP9
========
Vertical Tension Machine                                 Purdue
Soldering Machine Mod.                                  E.Borissov,B.Koecher
Winder Prot. Guard                                         E.Borissov
Gluing Machine Hook-up                                 J.Zimmermann
Panel Cleaning P3                                            D. Northacker.
Parts Procurement                                           G.Apollinari,A.Korytov