RD53A injection capacitor measurement results & RD53A pixel hit - - PowerPoint PPT Presentation

RD53A injection capacitor measurement results & RD53A pixel hit delay measurement results

2/02/2018 – Magne Lauritzen

Part 1 Injection capacitor measurement

Goal

• Measure capacitance of the injection

capacitors that are present on every analog front end.

Method

• Each RD53A has two banks of 100 parallel

injection capacitors for the purpose of measuring the capacitance (SX and DX).

• We cyclically charge and discharge the banks,

and deduce the capacitance from the frequency, voltage, measured current.

• The method is detailed in sec. 10.9 of the

RD53A manual.

Method, cont.

• We measured the inj. cap. values of both

banks for multiple charge-discharge cycle frequencies for one chip.

• For all other chips, we measured at a single

frequency.

Measured injector capacitor value in the SX bank of a RD53A chip as function of

• frequency. Parasitic capacitance has been subtracted.

There is normally a small variation in the injection capacitor value when measured via the DX or SX bank. The SX bank appear to have slightly greater capacitance. The measurements show decent agreement with the nominal value of 8.5fF. Chip ID Bank Average = 8.21fF Average Δ(SX-DX) = 0.04fF

Summary

• The measured test charge injection capacitor

values (~8.2fF) are slightly less than the nominal value (8.5fF).

• There is a slight variation in the capacitor

values (~0.04fF) depending on which side of the chip they are located (left vs right)

• The statistics are low; results might not be

representative of all chips.

Part 2

Analog front end trigger output delay as function of injected charge

Goal

• From injecting a charge into an analog front

end, to a digital signal being output, there is a delay.

• We wished to measure how this delay

changed as a function of injected charge.

Method

• Set injection voltage (Vcal_high – Vcal_med).
• With oscilloscope, trigger on injection signal

(cal_edge).

• Measure the delay until HitOr rising edge.
• Repeat measurement x10 000 to build statistics.
• Increase injection voltage and repeat.

Method, cont.

• # of injected electrons given by:
• The delay should decrease with increasing Ne.

This is due to an effect called time walk (see next slide)

Time walk

300e- 600e- 4000e- 30000e-

Further tests

• Measure how the absolute delay changes as a

function of the comparator bias current

• Measure the delay on other front end flavors

(differential, synchronous)