The temperature dependence of the Calibration DAC of the Master 0 chip of VAL K3-165 have been measured in the range between -15 and 60 degrees. Master 0 is the chip closest to the temperature sensor (PT1000). A parametrisation of the current results can be used to correct measurements with chips (modules) working at different temperatures. For example, the KEK test beam results.
The module is kept inside the Valencia test beam box without cover plates. The whole setup is in a climate chamber. The temperature of the hybrid (close to M0) is read off the PT1000. A PicoProbe is placed on the CAL0 pad of the chip. The signal is externally amplified (gain=10) using a LeCroy 612A amplifier. For each DAC setting and temperature, as in the figure below, pulses (yellow curve) are histogrammed on the oscilloscope (blue curve) with about 4 Million entries. The resulting histogram is fitted with the sum of two Gaussians. The difference of the centers is used as the measured DAC output voltage.
The measured DAC output voltages are listed in the table below for a number of temperatures and DAC settings.
| Climate chamber temperature | 40 | 25 | 0 | -20 | -30 |
| Hybrid temperature | 54 | 40 | 16 | -5 | -14 |
| 16 fC nominal | 160.9 | 161.9 | 165.0 | 168.5 | 168.9 |
| 12 fC nominal | 123.2 | 123.9 | 125.9 | 127.5 | 128.6 |
| 8 fC nominal | 82.7 | 82.8 | 84.4 | 85.7 | 86.1 |
| 4 fC nominal | 41.6 | 41.9 | 42.3 | 43.2 | 43.2 |
| 2 fC nominal | 22.6 | 23.3 | 24.1 | 22.0 | 20.7 |
The same DAC setting at different temperatures gives different pulse heigths. The lower the temperature the higher the pulse. Plots of voltage versus temperature for different DAC settings: 16fC nominal, 12fC nominal, 8fC nominal, 4fC nominal, 2fC nominal.
The slope of the ouput voltage versus nominal voltage clearly decreases with temperature.
The slope, as fitted from the previous plot, shows an approximately linear dependence on temperature:
Therefore, the output voltage can be parametrised as:
V_DAC = (DACresponse + beta * T) V_nominal
The measured temperature dependence (beta) for the CalDAC is -0.00082 1/K.
The DACresponse at 0 degrees is 1.038.
We used exactly the same setup as for the Non-Irrad module. But Vdd=5.1V, because of the known Vdd problem after irradiation on this batch of chips. The results show the same behaviour.
Using the same parametrization:
The measured temperature dependence (beta) for the CalDAC is -0.00069 1/K.
The DACresponse at 0 degrees is 0.981.
Vdd=5.1V is only needed in order to run the 12 chips. With a bit of luck and pacience the first 3 chips could be readout with Vdd=4V.
An extra measurement was done. 3 chips of the irradiated module (M0, S1, S2) were measured with the climate chamber set to 0ºC. For each of them, the DAC response curve was obtained, for Vdd=4V and Vdd=5.1V.
Results are summarized in this table:
| Chip | Vdd | Temp. | DACresponse | Reference to Non-irrad M0 |
| Non-irrad M0 | 4V | 16º | 1.024 | 1 |
| Irrad M0 | 4V | 13º | 0.987 | 0.964 |
| Irrad S1 | 4V | 13º | 1.020 | 0.996 |
| Irrad S2 | 4V | 12º | 0.990 | 0.967 |
The difference between Vdd=5.1V and Vdd=4V represents a bigger response for the second. The slope difference is 0.017.
The climate chamber was at 0ºC for all the measurements, but from Vdd=5.1V to Vdd=4V the difference in power consumption, makes that the hybrid was 7ºC colder in the second case. This temperature difference affects also the DAc response. Although it doesn't explain all the slope difference.
The setup is similar to that used for the measurement of the CalDAC. But now two DC probes are placed on the pads VTHP and VTHN. The voltage difference between the two can inmediately be read off a voltmeter.
The measured voltages are listed in the table below.
| Climate chamber temperature | 40 | 20 | 0 | -20 | -30 |
| Hybrid temperature | 60 | 42 | 22 | 1 | -10 |
| 0 mV nominal | -0.1 | -0.1 | -0.1 | -0.1 | -0.1 |
| 40 mV nominal | 36.8 | 38.3 | 39.9 | 41.4 | 42.2 |
| 80 mV nominal | 73.4 | 76.5 | 79.7 | 82.7 | 84.3 |
| 120 mV nominal | 110.3 | 114.8 | 119.6 | 124.1 | 126.6 |
| 160 mV nominal | 149.6 | 155.5 | 161.8 | 167.9 | 171.2 |
| 200 mV nominal | 186.1 | 193.6 | 201.5 | 209.2 | 213.2 |
| 240 mV nominal | 222.5 | 231.4 | 240.9 | 250.3 | 255.0 |
| 280 mV nominal | 259.2 | 269.5 | 280.4 | 291.5 | 297.0 |
| 320 mV nominal | 296.3 | 308.2 | 320.5 | 332.8 | 339.3 |
| 360 mV nominal | 332.9 | 346.3 | 359.9 | 373.9 | 381.3 |
| 400 mV nominal | 369.2 | 383.6 | 399.0 | 414.7 | 423.0 |
| 440 mV nominal | 405.7 | 422.3 | 438.8 | 455.5 | 464.8 |
| 480 mV nominal | 444.4 | 462.0 | 480.7 | 498.8 | 509.3 |
| 520 mV nominal | 480.9 | 500.0 | 520.0 | 539.6 | 551.0 |
| 560 mV nominal | 517.2 | 537.8 | 559.1 | 580.3 | 592.5 |
| 600 mV nominal | 554.4 | 575.2 | 598.3 | 621.1 | 634.4 |
| 640 mV nominal | 588.3 | 611.1 | 635.2 | 659.1 | 674.1 |
Again, the same DAC setting at different temperatures gives different voltage. The lower the temperature the higher the voltage. The slope of the ouput voltage versus nominal voltage decreases with temperature.
And again, the slope shows an approximately linear dependence on temperature. But the beta parameter is quite larger: beta = -0.00187. The DACresponse at 0ºC is 1.036.
We used exactly the same setup as for the Non-Irrad module. Vdd=4V in this case. The results show the same behaviour.
Using the same parametrization:
The measured temperature dependence (beta) for the ThrDAC is -0.00177 1/K.
The DACresponse at 0 degrees is 0.98.
The temperature coefficient is roughly the same as in the non-irradiated chip. But again, the DAC response at 0ºC is 94.6% of the non-irradiated chip.
- The Calibration is done in situ, at the same temperature. Obtaining a relationship between Threshold DAC settings and Calibration DAC settings at this temperature. Barrel and forward modules obtain different relationships as they run at different temperatures.
- During the testbeam data taking, the Threshold DAC is used at the same temperature. The previous relationship gives a Threshold in terms of Calibration DAC settings.
- We have seen that the Calibration DAC depends on the temperature, thus, the temperature difference between barrel and forward modules will produce different results.
- There is no effect due to the temperature variation of the Threshold DAC at the testbeam (as the calibration is done at the same temperature).
- The temperature difference between barrel and forward in the KEK testbeam was 35ºC. Applying a correction of 1 + beta * 35 = 0.971 to the charges of the forward modules, improves the agreement between the unirradiated modules.
| MODULE | VAL165 | CG170 | KEK11 | KEK22 |
| Median Charge (275V) | 3.81 | 3.79 | 3.63 | 3.45 |
| Corr. Median Charge (275V) | 3.70 | 3.68 | X | X |