PS & Services Stefan Rummel PXD SVD Meeting Prague - PowerPoint PPT Presentation

LMU München - Excellence Cluster Universe PS & Services Stefan Rummel PXD SVD Meeting Prague 21.01.15-23.01.15

Overview • Noise measurements • Noise of Step Down converter • Load regulation and optimization • Services: • Kapton development Stefan Rummel 2

Noise measurements – New campaign at MPI Current EMCM probes module Spectrum analyzer Stefan Rummel 3

Noise measurements – New campaign at MPI • A year ago the noise emission of PS units of PXD system was measured (Demonstrator PS) • PS unit loaded by a set of resistances. • Noise emission spectra totally characterized • A new noise test campaign was carried out first week of December in order to continue the characterization of the PXD noise (Preproduction PS) • The main goal of this test was to characterize: • System noise level of PXD system. • PS units + power cables + EMCM • The effects of ASIC operation on noise emission level. Stefan Rummel 4

Noise measurements – New campaign at MPI • The noise current in the power cables has been measured under four different conditions : • PS standby (Power DC-DC converters off) • PS on - Voltage set to 0V (DC-DC converters on) • PS on - Voltage set to nominal values: Analogue ASICs OFF • PS on - Voltage set to nominal values: Analogue ASICs ON • Frequency range : • 10 kHz up to 50 MHz • BW : 1kHz. • Common Mode from 3 DC-DC converters has been measured as well as noise level in other conductors (CM+DM) Stefan Rummel 5

3. PXD system noise emission test » All OFF - Background noise – CM DC-DC1 • CM : (DCD_AVDD+Refin +Source+ ANG_GND) CM DC-DC1 70 all off STAND-BY ON with V=0 60db = 1 mA 60 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON 50 40 40db = 100µA 30 dB  A 20db = 10µA 20 10 0db = 1 µA 0 (rms) -10 -20 4 5 6 7 10 10 10 10 Frequency H.-G. Moser, B2IM, December 2014 6

PXD system noise emission test CM DC-DC1 CM DC-DC1 70 70 all off all off STAND-BY STAND-BY ON with V=0 ON with V=0 60 60 ON with V=Nominal and ANALOG OFF ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON ON with V=nominal and Analog ON On – V=0 50 50 standby 40 40 30 30 dB  A dB  A 20 20 10 10 0 0 -10 -10 DC-DC 1 -20 -20 4 5 6 7 4 5 6 7 10 10 10 10 10 10 10 10 Frequency Frequency CM DC-DC1 CM DC-DC1 70 70 all off all off STAND-BY STAND-BY ON with V=0 60 ON with V=0 60 ON with V=Nominal and ANALOG OFF ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON ON with V=nominal and Analog ON 50 50 40 40 30 30 dB  A dB  A 20 20 10 10 0 0 On – Vnom – Ang on On – Vnom – Ang off -10 -10 -20 -20 4 5 6 7 10 10 10 10 4 5 6 7 10 10 10 10 Frequency Frequency New noise components show up H.-G. Moser, B2IM, December 2014 7

PXD system noise emission test CM DC-DC2 70 all off STAND-BY ON with V=0 60 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON 50 40 30 dB  A 20 10 0 -10 -20 4 5 6 7 10 10 10 10 Frequency » CM DC-DC 2 Again, new noise components show up, even with V=0 (but DC-DC converters on) H.-G. Moser, B2IM, December 2014 8

PXD system noise emission test DM DHPCORE (DHPVDD) 70 all off STAND-BY ON with V=0 60 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON (3 chips) 50 40 30 dB  A 20 10 0 -10 -20 4 5 6 7 10 10 10 10 Frequency » CM+DM (DC-DC 3) : DHP_VDD Again, new noise components show up, even with V=0 (but DC-DC converters on) H.-G. Moser, B2IM, December 2014 9

Triggering effects » The noise current in the power cables was measured with and without trigger • Without: only ADC in DCD active, no data processing in DHP • With: in addition: Data processing in DHP, data transfer » Frequency range • 100Hz – 20 kHz » Four lines were measured : • 2 digital lines: DHP_VDD & DCD_VDDD (DHP and DCD supply) • 2 analogue lines: DCD_VDDA & SOURCE (DCD supply & sensor) The largest effect is expected on DHP_VDD (activity due to trigger) H.-G. Moser, B2IM, December 2014 10

Triggering effects: Digital lines Trigger effect on Digital signal:DHPCORE (DHPVDD) Trigger effect on Digital signal: DCDVDDD) 80 80 trigger off trigger on trigger on trigger off 70 70 60 60 50 50 40 40 dB  A dB  A 30 30 20 20 10 10 0 0 -10 -10 -20 -20 2 3 4 10 10 10 2 3 4 10 10 10 Frequency Frequency DHP_VDD _VDD DCD_VDDD Digital activity Trigger Off - green Trigger On - green of DHP Trigger On - blue Trigger Off - blue clearly visible H.-G. Moser, B2IM, December 2014 11

Triggering effects: Analogue lines Trigger effect on Analog signal: DCD V DDA) Trigger effect on Source signal 80 80 trigger on trigger off trigger off trigger on 70 70 60 60 50 50 40 40 dB  A dB  A 30 30 20 20 10 10 0 0 -10 -10 -20 -20 2 3 4 2 3 4 10 10 10 10 10 10 Frequency Frequency DC_DVDDA SOURCE No cross talk on Trigger on - green Trigger on - green Trigger Off - blue Trigger Off - blue analogue supplies! H.-G. Moser, B2IM, December 2014 12

Noise measurements – new noise components • Noise measurement revealed an increased noise of “Preproduction” units vs. “Demonstrator” • Main differences: • More powerful DC/DC converter • More DC/DC converter in “Preproduction” units to feed the bipolar regulators • Additional step converter for analog and digital supply voltages  More noise which distributes in the unit  New DC/DC converter card for Main production comprise additional filtering at input and output of DC/DC converters  Form this measurements one cannot judge the performance of the module  Either sensitivity analysis of the module required  Fernando is working on that  Or comparative measurement with different PS’s  Felix showed promising measurements Stefan Rummel 13

New noise sources – Step Down converter • Noise sources in converter: • Ripple due to current through inductor @ switching freq. • Switching noise – HF, due to fast switching of FET’s and resulting ringing Strategy for mitigation: • Filtering • Controlling the slope of the edge • Active circuitry • Ferrite beads, Resistors in gate drive circuitry Stefan Rummel 14

Noise measurements – of Step Down converter • Mitigation using Ferrite beads and resistors in Gate drive improve Peak to Peak noise by factor 2  Given design can be significantly improved Stefan Rummel 15

Load regulation • Current generation of DHP is very sensitive to deviation of supply voltage  Load regulation might be an issue  High sense wire resistance is known to deteriorate load regulation Design goals of sense wire input stage: • High input resistance • Voltage limitation in case of broken sense wire Stefan Rummel 16

Load regulation – Input stage design • Two variants in Preproduction Unit: • Regulator with active sense amplifier • Error amplifier in differential configuration Stefan Rummel 17

Load regulation • Investigation of two scenarios: • Long power cable + intermediate cable  Rsense = 3W • Long power cable + intermediate cable + 24 W to simulate module / Kapton resistance Stefan Rummel 18

Load regulation • Preproduction in modified configuration significantly improves output DC impedance to 2.5mW significantly better than commercial unit • Baseline for main production: Active sense amplifier with 51k protection resistor Stefan Rummel 19

Main Production • Lessons learned from recent test efforts: • Improving noise of Step Down converter • Improve filtering of DC/DC converter cards  Added CMC filter and converter inputs  Improved DM on outputs • Load regulation:  Active sense input for analog and digital channels  Higher input resistivity for circuitry connected to sense wires in the OVP card Stefan Rummel 20

Services - Kapton • First Kapton layout has been finished • Documentation of interfaces available • Bond scheme • Assignment of connectors • Current load on bond wires and connector pins  PXD9 EOM – Kapton interface correct  Current load has been conservatively chosen • Minor modifications will be done within this week  Start of production within Jan. 2015 Stefan Rummel 21

Summary Noise Measurements: • New noise components visible compared to prev. measurement • Additional currents visible while triggering and analog section in operation • No crosstalk between analog and digital voltage visible • Most significant change in hardware are the step down converter and some additional low power converter • Measures improve the noise of the step down converter have been demonstrated • In addition filtering of DC/DC converter cards improved • Load regulation in presence of sense resistance can be improve from 30mW to 2.5mW for the preproduction PS’s and finally towards .5mW for the Main production  Recent test effort gave valuable input for the main production • First Kapton Design finished, after fixing some minor issues production will be started Stefan Rummel 22