4.3 Validation and Testing
119
Table 4.2 MCH production test statistics
Test station
1
2
Devices tested
21 586
22 406
Devices passed
15 720
17 147
Devices failed
5872
5264
Fail category
Analogue current
913 (4.23%)
721 (3.22%)
Link synchronization
301 (1.39%)
288 (1.29%)
Analogue pedestal
2186 (10.13%)
2119 (9.46%)
RMS noise
2819 (13.06%)
2375 (10.60%)
Gain
2336 (10.82%)
2045 (9.13%)
Rise time
1973 (9.14%)
1654 (7.38%)
Crosstalk
221 (1.02%)
146 (0.65%)
Ring oscillator
246 (1.14%)
291 (1.30%)
I 2 C
216 (1.00%)
275 (1.23%)
JTAG
473 (2.19%)
364 (1.62%)
Memory test
1245 (5.77%)
1403 (6.26%)
Scanchain
1551 (7.19%)
1269 (5.66%)
Bit errors
546 (2.53%)
505 (2.25%)
operate with a ring oscillator frequency down to 80 MHz, and so is not affected, since
the spread in values barely goes below 100 MHz.
4.3.3 Test Results for Final SAMPA Production
Table 4.2 presents the current results from the two test stations for the ongoing production testing of the SAMPA for the MCH as of winter 2019.
8 About 40 000 devices
have been tested, with a current yield of 74.7 %. About 6.5 % of the devices have been
scheduled for a retest due to inconsistent results. Previously, about 13 000 devices
were tested for the TPC, but error statistics are not available for that batch. However,
the experience showed that retesting devices that failed testing could bring the yield
up by about 7 %-points. Primarily analogue based tests might pass on second testing,
while digital test would not. The secondary testing thus might bring the final yield
close to 80 %. Comparing the results with those for the testing of v2, we see they are
close to the same considering the lower statistics for the v2 testing.
8 Test of second batch was performed by Anders Oskarsson and David Silvermyr, Lund University,
Sweden.
119
Table 4.2 MCH production test statistics
Test station
1
2
Devices tested
21 586
22 406
Devices passed
15 720
17 147
Devices failed
5872
5264
Fail category
Analogue current
913 (4.23%)
721 (3.22%)
Link synchronization
301 (1.39%)
288 (1.29%)
Analogue pedestal
2186 (10.13%)
2119 (9.46%)
RMS noise
2819 (13.06%)
2375 (10.60%)
Gain
2336 (10.82%)
2045 (9.13%)
Rise time
1973 (9.14%)
1654 (7.38%)
Crosstalk
221 (1.02%)
146 (0.65%)
Ring oscillator
246 (1.14%)
291 (1.30%)
I 2 C
216 (1.00%)
275 (1.23%)
JTAG
473 (2.19%)
364 (1.62%)
Memory test
1245 (5.77%)
1403 (6.26%)
Scanchain
1551 (7.19%)
1269 (5.66%)
Bit errors
546 (2.53%)
505 (2.25%)
operate with a ring oscillator frequency down to 80 MHz, and so is not affected, since
the spread in values barely goes below 100 MHz.
4.3.3 Test Results for Final SAMPA Production
Table 4.2 presents the current results from the two test stations for the ongoing production testing of the SAMPA for the MCH as of winter 2019.
8 About 40 000 devices
have been tested, with a current yield of 74.7 %. About 6.5 % of the devices have been
scheduled for a retest due to inconsistent results. Previously, about 13 000 devices
were tested for the TPC, but error statistics are not available for that batch. However,
the experience showed that retesting devices that failed testing could bring the yield
up by about 7 %-points. Primarily analogue based tests might pass on second testing,
while digital test would not. The secondary testing thus might bring the final yield
close to 80 %. Comparing the results with those for the testing of v2, we see they are
close to the same considering the lower statistics for the v2 testing.
8 Test of second batch was performed by Anders Oskarsson and David Silvermyr, Lund University,
Sweden.