1. Bug in analog part
1.1. huge noise at high temp
1.1.1. Shmoo temperature, clk frequency, ADC input voltage
1.1.1.1. Temperature threshold is about 130 deg.C
1.2. wrong reference voltage
1.2.1. not possible as adata is correct 50%
1.3. High impedance node
1.4. Noisy supply voltage
1.4.1. Voltage drop due to current consumption. Possible effect if the current limiter of OUT is not activated ?
1.4.1.1. Monitor Vdec and Idd
1.5. Different layout
1.5.1. Parasitic thyristor
1.5.1.1. -> Layout review, compared with 90371
1.5.1.2. -> ADC adata versus input voltage controlled on pin INPUT
1.5.1.2.1. cancelled
1.5.2. Leakage current
1.5.2.1. Probably not as adata is correct 50% time
1.5.3. Parasitic capacitive coupling with a digital signal
2. Bug in digital
2.1. Timings
2.1.1. Master clock too fast
2.1.1.1. Decrease clock frequency
2.1.1.1.1. No very significant effect (ndp) -> DOUBTFUL, to be repeated
2.1.1.2. Use the 12MHz frequency
2.1.1.2.1. All adata are correct at main-clock frequencies below about 20MHz, see below
2.1.2. Bus access conflict due to another block
2.1.2.1. SENT
2.1.2.1.1. -> use a RAM program, with ADC-self-test sdata
2.1.3. Patch issue: bad interrupt management ?
2.1.3.1. remove patch, use RAM pg
2.1.3.1.1. Patch effect very unlikely
2.1.4. Change in digital, compared to 90371 ? (overflow management,...)
2.1.4.1. define patch to go to 90371 overflow mode
2.1.4.1.1. See below
2.1.4.2. or write the port Saturate=0 in test mode
2.1.4.2.1. Both options tried with Ram program: in both cases, the ADC self-test diagnostic 16 and 17 pass
2.1.4.2.2. Peter sees that it has an effect on SOME failing device, at least in a test mode.
2.1.4.2.3. BUG IN ADC_OVF FLAG
2.2. ADC digital failure
2.2.1. Reset due to DMA error
2.2.1.1. Play with NV_DIAG_DMA_ERR and NV_DIAG_CPU_ERR
2.2.1.1.1. cancelled
3. Bug in interface A-D
3.1. Different schematic
3.1.1. Comparison -> no difference at top, same ADC-top-analog used
3.1.1.1. Difference in inputs of one comparator ?
3.1.1.1.1. not confirmed: no difference observed
3.2. Different bias current
3.2.1. Play with Ibias trimming value
3.2.1.1. No effect (ndp)
3.3. Option different than in 90371
3.3.1. Explore all ADC options
3.3.1.1. No effect (ndp) -> DOUBTFUL, to be repeated
3.4. Try the different analog diagnostics ?
3.4.1. cancelled
3.5. Store of the nvram at 160 degC
3.5.1. out of scope
3.6. Try RAM-probe mode in SENT (DAL)
4. Bug in SW / System
4.1. adata RAM rewritten
4.1.1. -> use RAM pg
4.1.1.1. Pass
4.2. regular reset when ADC self-test is done with negative inputs
4.2.1. -> use RAM pg
4.2.1.1. Ram pg passes, but there is no reset
4.3. Bad lot
4.3.1. Ask PSC as he succeeded the ADC self-test at high temp (conditions ?)
4.3.1.1. Test pass but at 125 deg.C...
4.3.1.1.1. redo at 160 deg.C (PSC, planned)
4.3.2. The tested part is failing
4.3.2.1. Try on a AAA part the ADC self-test
4.3.2.2. Fab waiver does not touch ADC or digital