1. HECII-Workshop Munich, 12./13. February 2009 Measurement Strategy Workshop on Technology Selection for the SLHC Hadronic Endcap Calorimeter, Munich 2009 Olaf Reimann for the MPI-HEC-Group

2. HECII-Workshop Munich, 12./13. February 2009 Want to give a short and rough overview about our measurement setup and motivations. Goals and Selection Criteria What to measure? –Dynamic Gain and Gummel-Plot –S-Parameter Basics Measurement Setup Outline

3. HECII-Workshop Munich, 12./13. February 2009 Design Goals Important for device selection

4. HECII-Workshop Munich, 12./13. February 2009 Selection Criteria Main device for a technology selectio is the transistor. (Passive devices are less critical)

5. HECII-Workshop Munich, 12./13. February 2009 Definitions Dynamic common emitter gain: Static common emitter gain: Damage Factor: In many papers we find different values for the current gain. Therefore we use the definitions: Different types of transistors have different principal parameters. To compare the behavior under radiation or temperature change we need a normalization: In spite of their names both values are defined for DC.

6. HECII-Workshop Munich, 12./13. February 2009 Dynamic Gain (and Gummel-Plots)

7. HECII-Workshop Munich, 12./13. February 2009 Gummel Plot A lot of papers are based on Gummel-plots. -> Needed for comparison Very useful for model-parameter extraction -> Needed for simulation Calculation of B and  are based on I B and I C -> Needed for consistency check with S-parameters Examples: Gummel-plot from Cressler and MPI-group (All Gummel-plots for U CB =0V)

8. HECII-Workshop Munich, 12./13. February 2009 S-Parameter Basics

9. HECII-Workshop Munich, 12./13. February 2009 Why S-Parameters? Detector CableAmplifier Typical detector system: Example: Simple bipolar transistor amplifier (common emitter) Current-voltage conversation at cable impedance and/or R I (Depending on pulse width) Amplification: A lot of parameters are important ( , r BE, r CE ) -> Two-port description -> S-parameters

10. HECII-Workshop Munich, 12./13. February 2009 Definition Measurement of reflected and incident voltages Measurement is done at an reference impedance (50 Ohm) Measured in frequency domain (Value, Phase) -> Full description of the DUT (in linear regime) Broad frequency range (up to several GHz) Transformable to other parameters (h, Z, Y, …)

11. HECII-Workshop Munich, 12./13. February 2009 Network Analyzer (Principle) Basic architecture of a two-port vector network analyzer. (from Wikipedia under GNU-License)

12. HECII-Workshop Munich, 12./13. February 2009 Measurement Setup (Rez)

13. HECII-Workshop Munich, 12./13. February 2009 DC/S-Parameter Setup

14. HECII-Workshop Munich, 12./13. February 2009 Long vs. Short Cables Example: S-parameters (IHP-MOSFET) measured with 25cm and ~40m cable S11 @ 25cm cable: light blue S11 @ 40m cable: brown S22 @ 25cm cable: blue S22 @ 40m cable: violet S21 @ 25cm cable: blue S21 @ 40m cable: violet

15. HECII-Workshop Munich, 12./13. February 2009 Realization Testing up to 37 devices (transistors) Measuring DC-values and S-parameters Measurement during neutron irradiation

16. HECII-Workshop Munich, 12./13. February 2009 Target Region Deuterium target (neutron source) Up to 11 boards with devices under test Cabling for RF-DC-signals Up to four devices per board located in the beam Neutron flux density spectrum at different positions:

17. HECII-Workshop Munich, 12./13. February 2009 Example: Bipolar Transistor Interesting Region Parameter |S21| in dB Parameter |S11|in dB (“Input-Impedance”) Currents (in mA) and voltages (in V) npn-bipolartransistor from IHP SiGe-technology 0.42x0.84µm 2 structure size 2 elementary cells in parallel Included ESD-protection Positioned in slot 1 All plots over neutron fluence (cm -2 )

18. HECII-Workshop Munich, 12./13. February 2009 Definition of an „Overall“-Parameter Transfer-function for a worst case scenario (all S-parameters are integrated to one value, better for comparison): with „Worst Case“: (Called „gain“ in Agnes slides)

19. HECII-Workshop Munich, 12./13. February 2009 Summary Identification of some interesting (critical) parameters (Gain, dynamic range, …) DC-characteristics and Gummel-plots for “compability” with literature S-parameters for full characterization in frequency-domain -> Simulations for all kinds of signals are possible „40m“-measurement is not a problem -> „live“-measurement under radiation A lot of data are collected (but a lack of manpower prevents us from a deeper data analysis)

20. HECII-Workshop Munich, 12./13. February 2009 Basic Design Considerations (Discussion)

21. HECII-Workshop Munich, 12./13. February 2009 Discussion about measurement technique –Validity Is the S-parameter measurement an accepted technique? Is our measurement setup accepted (Online / 35m)? Better to use “Gummel”-measurement? –Comparison S-parameter / pulse measurement Are they comparable? –Do we have enough data to decide between technologies? Principal decision: FET via bipolar –Noise Is noise a strong selection criteria? – Temperature dependence (Warm / cold behavior) Better to use a principle stable technology (Factor 2 limit)? Use of a different bias scheme (current source) for bipolars? Design Considerations

22. HECII-Workshop Munich, 12./13. February 2009 HEC-II ”Proto”-Amplifiers (Discussion)

23. HECII-Workshop Munich, 12./13. February 2009 Building a „Proto“-Amplifier is useful? –Some selected devices with external (discrete) passive elements –It‘s possible to test different circuit schemes Reduce cost because it‘s not necessary to use wafers Saving time Very easy to do some changes Very easy to characterize (in circuit measurement) –Measure under radiation Impulse response Noise Dynamic range In combination with detector and cabling … –Reduce influence of gain variations Proto-Amplifier