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

ps services
SMART_READER_LITE
LIVE PREVIEW

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

LMU Mnchen - 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


slide-1
SLIDE 1

LMU München - Excellence Cluster Universe

PS & Services

Stefan Rummel

PXD SVD Meeting Prague 21.01.15-23.01.15

slide-2
SLIDE 2

Overview

Stefan Rummel

  • Noise measurements
  • Noise of Step Down converter
  • Load regulation and optimization
  • Services:
  • Kapton development

2

slide-3
SLIDE 3

3 Stefan Rummel

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

slide-4
SLIDE 4

4 Stefan Rummel

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.
slide-5
SLIDE 5

5 Stefan Rummel

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)

slide-6
SLIDE 6

H.-G. Moser, B2IM, December 2014

» All OFF - Background noise – CM DC-DC1

  • CM : (DCD_AVDD+Refin +Source+ ANG_GND)
  • 3. PXD system noise emission test

10

4

10

5

10

6

10

7

  • 20
  • 10

10 20 30 40 50 60 70 CM DC-DC1 Frequency dBA all off STAND-BY ON with V=0 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON

6 60db = 1 mA 40db = 100µA 20db = 10µA 0db = 1 µA (rms)

slide-7
SLIDE 7

H.-G. Moser, B2IM, December 2014

10

4

10

5

10

6

10

7
  • 20
  • 10

10 20 30 40 50 60 70 CM DC-DC1 Frequency dBA all off STAND-BY ON with V=0 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON

PXD system noise emission test

10

4

10

5

10

6

10

7
  • 20
  • 10

10 20 30 40 50 60 70 CM DC-DC1 Frequency dBA all off STAND-BY ON with V=0 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON 10

4

10

5

10

6

10

7
  • 20
  • 10

10 20 30 40 50 60 70 CM DC-DC1 Frequency dBA all off STAND-BY ON with V=0 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON 10

4

10

5

10

6

10

7
  • 20
  • 10

10 20 30 40 50 60 70 CM DC-DC1 Frequency dBA all off STAND-BY ON with V=0 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON

standby On – V=0 On – Vnom – Ang off On – Vnom – Ang on DC-DC 1

New noise components show up

7

slide-8
SLIDE 8

H.-G. Moser, B2IM, December 2014

PXD system noise emission test

» CM DC-DC 2

10

4

10

5

10

6

10

7

  • 20
  • 10

10 20 30 40 50 60 70 CM DC-DC2 Frequency dBA all off STAND-BY ON with V=0 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON

Again, new noise components show up, even with V=0 (but DC-DC converters on)

8

slide-9
SLIDE 9

H.-G. Moser, B2IM, December 2014

» CM+DM (DC-DC 3) : DHP_VDD

PXD system noise emission test

10

4

10

5

10

6

10

7

  • 20
  • 10

10 20 30 40 50 60 70 DM DHPCORE (DHPVDD) Frequency dBA all off STAND-BY ON with V=0 ON with V=Nominal and ANALOG OFF ON with V=nominal and Analog ON (3 chips)

9

Again, new noise components show up, even with V=0 (but DC-DC converters on)

slide-10
SLIDE 10

H.-G. Moser, B2IM, December 2014

» 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)

Triggering effects

10

slide-11
SLIDE 11

H.-G. Moser, B2IM, December 2014

DHP_VDD _VDD Trigger On - green Trigger Off - blue

Triggering effects: Digital lines

10

2

10

3

10

4

  • 20
  • 10

10 20 30 40 50 60 70 80 Trigger effect on Digital signal:DHPCORE (DHPVDD) Frequency dBA trigger off trigger on 10

2

10

3

10

4

  • 20
  • 10

10 20 30 40 50 60 70 80 Trigger effect on Digital signal: DCDVDDD) Frequency dBA trigger on trigger off

DCD_VDDD Trigger Off - green Trigger On - blue

11

Digital activity

  • f DHP

clearly visible

slide-12
SLIDE 12

H.-G. Moser, B2IM, December 2014

Triggering effects: Analogue lines

10

2

10

3

10

4

  • 20
  • 10

10 20 30 40 50 60 70 80 Trigger effect on Analog signal: DCDVDDA) Frequency dBA trigger on trigger off

10

2

10

3

10

4

  • 20
  • 10

10 20 30 40 50 60 70 80 Trigger effect on Source signal Frequency dBA trigger off trigger on

DC_DVDDA Trigger on - green Trigger Off - blue SOURCE Trigger on - green Trigger Off - blue 12

No cross talk on analogue supplies!

slide-13
SLIDE 13

13 Stefan Rummel

  • 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

Noise measurements – new noise components

slide-14
SLIDE 14

14 Stefan Rummel

  • 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

New noise sources – Step Down converter

slide-15
SLIDE 15

15 Stefan Rummel

  • Mitigation using Ferrite beads and

resistors in Gate drive improve Peak to Peak noise by factor 2  Given design can be significantly improved

Noise measurements – of Step Down converter

slide-16
SLIDE 16

16 Stefan Rummel

  • 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

Load regulation

slide-17
SLIDE 17

17 Stefan Rummel

  • Two variants in Preproduction

Unit:

  • Regulator with active

sense amplifier

  • Error amplifier in

differential configuration

Load regulation – Input stage design

slide-18
SLIDE 18

18 Stefan Rummel

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

Load regulation

slide-19
SLIDE 19

19 Stefan Rummel

  • 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

Load regulation

slide-20
SLIDE 20

Main Production

20 Stefan Rummel

  • 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

slide-21
SLIDE 21

Services - Kapton

21 Stefan Rummel

  • 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

slide-22
SLIDE 22

Summary

Stefan Rummel

Noise Measurements:

  • New noise components visible compared to prev. measurement
  • Additional currents visible while triggering and analog section in
  • peration
  • 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

22