Accuracy; Temperature - Measurement Computing PCI-DAS4020/12 User Manual

PCI-DAS4020/12 User's Guide
Pre-trigger mode
Data transfer
A/D Conversion time
Sample rate
Throughput

Accuracy

Absolute accuracy
Typical accuracy
Accuracy components:
Gain error
Offset error
Integral linearity error
Differential linearity error
Board error is a combination of gain, offset, integral linearity, and differential linearity error. The overall
absolute worst-case error of the board may be calculated by summing these component errors. Worst case
error is realized only in the unlikely event that each of the component errors are both at their maximum
level, and causing error in the same direction. Though this is very uncommon, it is still possible, and the
calculated worst case error of the PCI-DAS4020/12 board is ±7.0 LSB.
Each PCI-DAS4020/12 board is tested at the factory to make sure that its actual worst case error is less
than ±5.0 LSB. Allowing for a 10% guard-band, the absolute worst-case error of a board is ±5.5 LSB.
Typical accuracy can be calculated from the various component typical errors in a similar fashion. This
typical maximum error calculation for the PCI-DAS4020/12 board yields ±3.4 LSB. However, this again
assumes that each of the errors is forcing an error in the same direction. Empirical evidence supports a
conservative typical error budget of ±3.0 LSB.
Calibration

Temperature

Gain drift
Offset drift
Overall board temperature drift
Overall, worst-case temperature drift is calculated by adding the drifts corresponding to the gain and
offset drifts. This worst case number is unlikely to occur, as it requires both gain and offset drifts to be at
their maximum levels, and to be affecting the measurement in the same direction. However, the overall
board D/A drift specifications have been calculated in this fashion.
Unlimited number of pre-trigger samples, 16 Meg post-trigger samples.
Compatible with both digital and analog trigger/gate options. Data acquisition
initiated via TRIG1. Post-trigger phase initiated via TRIG2.
Via dual 32 K x 24 sample FIFO, SRAM based, with Bus-Master DMA and
scatter-gather, interrupt, or software polled.
40 ns
20 MHz max, 1 kHz min
Single channel: 20 MHz continuous
Two channels (0 and 1 or 2 and 3): 20 MHz continuous
Four channels: 10 MHz continuous
In background mode, the maximum throughput may be impacted by bus or
interrupt activity.
Table 6-3. Analog input accuracy specifications
±5.5 LSB worst case error (either range)
±3.0 LSB error (either range)
±2.0 LSB max, ±1.0 LSB typ
±1.5 LSB max, ±1.0 LSB typ
±2.5 LSB max, ±1.0 LSB typ
±1.0 LSB max , ±0.4 LSB typ (No missing codes guaranteed)
Table 6-4. Calibration specifications
Auto-calibration. Calibration factors for each range stored on board in non-volatile
RAM
Table 6-5. Temperature specifications
6-2
±5 V range: ±0.10 LSB/°C max
±1 V range: ±0.13 LSB/°C max
±0.11 LSB/°C max, all ranges
±5 V range: ±0.21 LSB/°C max
±1 V range: ±0.24 LSB/°C max
Specifications