jenda
Member level 3
Hi all
I have to record a signal that occurs randomly. It is important to record the whole signal and not to loose the first few samples (pretrigger). Normally this is done by continuous sampling and writing the data to a ring buffer.
The challenge is that my circuit mustn’t consume more then 0.5-1mW while waiting for a signal because it is powered by a small battery. The sampling rate should be 100ksps and the resolution 12-14bit. Continuous sampling and storing data to a buffer consumes too much energy.
I thought about delaying the input signal by using a bunch of allpass filters. But I didn’t find much on that. I don’t want to build the filters in a discrete way..
Some integrated analog delay line circuits I found are SAD-512 and MN3004. But I couldn’t find the power requirements and the MN3004 works only in the audio band.
Another idea I thought about is to use a CPU that wakes up very quickly (~5us). An analog trigger is waking up the CPU. If this is possible, I only would loose the first sample of the signal.
I also thought about some smart ADC with integrated data buffer. The circuits I found are called <sensor conditioners>, ADC with some CPU and EEPROM. They are mainly used to calibrate sensor and ADC. They quite meet my requirements expect power consumption.
Thanx, j
I have to record a signal that occurs randomly. It is important to record the whole signal and not to loose the first few samples (pretrigger). Normally this is done by continuous sampling and writing the data to a ring buffer.
The challenge is that my circuit mustn’t consume more then 0.5-1mW while waiting for a signal because it is powered by a small battery. The sampling rate should be 100ksps and the resolution 12-14bit. Continuous sampling and storing data to a buffer consumes too much energy.
I thought about delaying the input signal by using a bunch of allpass filters. But I didn’t find much on that. I don’t want to build the filters in a discrete way..
Some integrated analog delay line circuits I found are SAD-512 and MN3004. But I couldn’t find the power requirements and the MN3004 works only in the audio band.
Another idea I thought about is to use a CPU that wakes up very quickly (~5us). An analog trigger is waking up the CPU. If this is possible, I only would loose the first sample of the signal.
I also thought about some smart ADC with integrated data buffer. The circuits I found are called <sensor conditioners>, ADC with some CPU and EEPROM. They are mainly used to calibrate sensor and ADC. They quite meet my requirements expect power consumption.
Thanx, j
