Imprecise pulse width capture using Atmel SAM D51
We are using an ATSAMD51P19A and the Timer/Counter for Control Applications (TCC) to detect the pulse width from the assertion of a pin to the detection of a signal on another pin, and are having trouble achieving our desired resolution of +/- 10 nsec.
The below block diagram shows our Atmel START configuration to capture the pulse width.
Using a delay generator, we know the time delta from the assertion of the pin PA23 to the rising edge of the pin PA22 is 720 nsec +/- 5 nsec. Given our configuration, we expect that we should capture a pulse width of 720 nsec +/- 10 nsec. However, we are seeing 720 nsec +/- 50 nsec. Our application requires +/- 10 nsec, so the +/- 50 nsec is undesirable.
We have tried many configurations and have been unable to achieve an accuracy of less than +/- 50 nsec. Any suggestions or feedback on how to achieve a higher resolution is greatly appreciated.
We resolved our own issue.
The root cause of our problem was an oscillation on the MCU_VDDCORE/MCC_VSW supply rails of the ATSAMD51P19A, which was caused by the the low capacitor value connected to the MCU_VDDCORE/MCU_VSW circuit.
By adding a larger capacitor to the MCU_VDDCORE/MCU_VSW circuit, we removed the oscillation, which in turn solved our timing issue. Our variance went from +/- 50 nsec to +/- 10 nsec.
- How can i print a floating-point (VFP) number in ARM?
- Targeting ARM architecture with .NET compiler
- Building for ARM? Cannot do cross-compilation with `go install` when GOBIN is set
- How to create a QEMU ARM machine with custom peripherals and memory maps?
- ARM64 Buffer Overflow-Cannot overwrite $pc
- How can I easily install arm-elf-gcc on OS X?
- What registers to save in the ARM C calling convention?
- undefined reference to ruserok error
- perf on arm cortex-a7 produces no callstacks
- How to install cross compiler (on ubuntu 12.04 LTS) for microprocessor SA1100?
- glibc build with flags _FILE_OFFSET_BITS=64 and _TIME_BITS=64 fails
- ARM Intrinsic: Insert complex zero after each complex float sample
- Arm bootloader and kernel visualisation error
- A hand-coded branch to a bounce bed was not taken but taken with debugger attached
- How do I build an Arm version of Docker File for Qnap Arm based Server
- About ARMv8-A `SVC #0x0f0f` to machine code
- ARM Cortex-A8: Whats the difference between VFP and NEON
- How to assign an integer value to a Macro in trace32 cmm script?
- Why an illegal instruction exception is reported when I execute BC.cond instructions such as bc.eq or bc.ne on Armv9.0-a cores?
- Are there ARM intrinsics for add-with-carry in C?
- What is the difference between FIQ and IRQ interrupt system?
- Trace32 CMM script : understanding the shift operator
- How to extract required bit fields from 32 bit variable in Trace32
- Prefetch instructions on ARM
- Encoding and decoding ARM instructions to/ from binary
- Optimizing a for loop with lookup-table using ARM Neon instructions
- What are the differences between C/C++ bare-metal compilation and compilation for a specific OS (Linux)?
- Compiler Optimization Causing Incorrect Results
- Is there an ARM Neon Gather Instruction?
- Why are so many LDR and STR instructions generated with this simple code?