FE cards tests 7 July 2016 D.Autiero 1
Cryogenic ASISC were tested in February and production for entire 6x6x6 (700 chips) had been purchased after the acceptance tests in March. The FE cards were designed to host the chips + filtering capacitors + spark protection components (diode pairs extensively studied in the past) + front end components not put on the anode: polarizing resistor of strips and decoupling capacitor (discussions in March-April related to the connectivity and grounding system). The FE card layout has strict requirements to fit in the rails and in order to be able to be plugged correctly on the cold FT. The final layout of the FE cards was corresponding exactly to the one of the fake FE cards used to test the chimneys, a pre-batch of 5cards was produced in May. In situ test for one chimney with pre-batch of 5 cards 22-23/6 2016 Before in situ test cards had been pre-tested at IPNL both from the: mechanical point of view (see presentation at the 3x1x1 meeting of 2/6/2016 and next slide) cards exactly corresponding to the layout of the fake FE cards and fixation system had been re-measured and cross-checked with Sebastien electrical point of view channel by channel with charge injection from a card emulating the cold feedthrough and full chain of 2 m of flat cables on the blades 2
FE card PCB layout 3
Cryogenic ASICS Filtering capacitors for LV lines Decoupling capacitors Spark protection diodes Polarizing resistor 4
The final FE card has exactly the same mechanical layout of the fake FE no problems expected at the level of the insertion Meeting of 2 June 2016 The two layouts and all patterns of holes for the connectors correspond perfectly. When superimposing the two cards the connectors can be inserted with their pins easily crossing the two cards in a row and there is no difference on all the card borders. 5
Fixation system on the blades, example with a fake FE card Fake FE card Blade Thickness 1.6 mm Thickness 1.6 mm Groove for the Bridge M4 Screws integrated in the bridge and Connector pins Thickness 2 mm not adding any extra thickness 0.8 mm depth 6
Mechanical tests Isolated cold FT (check clearances) Horizontal chimney Full insertion test of cards in the vertical chimney of the top-cap 7
Lessons learned: 1) As already observed with the fake FE cards (which are identical), the cards need some light scraping of the two edges with abrasive paper in order to slide smoothly in the rails 2) As already observed with the fake FE cards the upper holes of the blade for the bridge fixation are in a position too close to the edge interfering with the flat cable connector occupancy, they have to be further drilled to make them slightly elliptical Clash of occupancy between the flat cable connector and the cable holder fixed by the screws 3) The horizontal chimney is not representative of the chimneys mounted on the top-cap: fake FE cards which were sliding without problems on the top-cap chimneys were getting stuck on the horizontal chimney 8
4) The cable loop on the flat cable connector make a little bump of about 2 mm which can differ from connector to connector and interfere with the insertion of the nearby slide Cable loop on the flat cable connector for the safety clamp may clash with the edge of the bridge during the insertion of the nearby blade and prevent full insertion Clearances are very tight , need to make sure that the cable loops of the safety clamps are all flat enough 9
After preparation for these little tuning problems insertion of the 5 blades in the top-cap chimneys was very smooth and reproducible 1) Alone I could insert the blades without problems 2) There is a very good mechanical feeling of the last step of plugging/unplugging in the cold FT connector 3) The handle at the top of the blade is a good stopper for overpressure in the insertion and matches the mechanical feeling of plugging in the cold FT connector 10
Pulsing tests: We used the cabling set up by Cosimo to inject the charge in the strips (tests the full chain) and we connected to our own pulsing system ( Cosimo’s system was not yet operative) Chimney not closed: proper warm flange not yet available All channels pulsed simultaneously can check is there are any disconnections in the chain Checked systematically all the 320 channels of the 5 cards. System working well and showing that channels were responding on all cards from the charge injected in the strips Strange feature: pulsing always present in the same way even when connecting to connectors of other chimneys (then found that the PCB of the FT with the pulser connectors has an unforeseen internal connection short-circuiting the 3 connectors. Feature to be fixed for the full functionality of the pulsing system but not limiting the test performed on the FE cards Found a couple of channels with no signal and observed that the signal was coming back by lifting slightly the blade and disappearing by increasing the insertion pressure (the opposite of what one would have expected) 11
Environmental noise: Observed some environmental noise of pickup of the 2m flat cables inside the chimney (pure pickup with the electronics off/disconnected , power supplies off, blade disconnected from cold flange) +-20 mV noise repeated at typical periods of 5 us and 100 us Same noise pattern observed synchronously just connecting the scope probe to the metallic structure of the top-cap Connection to top-cap Pickup of 2m flat cables the entire top-cap was oscillating at these frequencies because of bad grounding connections). Spent a few hours with Francois Duval to try to put the top-cap at a better grounding and to switch off a few equipment suspected create EMI. Much better situation is expected on the cryostat where a real grounding scheme is being put in place (subject on next tests after July 15th) 12
FE cards tests completed at cold at IPNL after the insertion test at CERN Tested at various temperatures in between room temperature and LN2 temperature; charge injection via the cold flange connector with an ad hoc card, full 2m flat cables connections setup used as in the chimneys As expected no differences observed for these tests on the ASICS mounted on the final FE cards with respect to the behavior of the ASICs measured during the characterization tests of the on the single chip test cards as performed in February Production of additional FE cards needed for 3x1x1 being launched given the completion of the tests 13
Conclusions: • The FE cards passed the mechanical and electrical checks • The additional FE cards for the 3x1x1 are being produced • Additional tests on the environmental noise foreseen after July 15th, after the implementation of the grounding on the cryostat 14
Recommend
More recommend
