EL125x - Functioning

General features

The EL1252, EL1254 provides not only the input level, but also the last time stamp at which a channel underwent a rising or falling edge. The status byte for each channel provides information about changes in the latch processes.

EL125x - Functioning 1:

Calibration

The time delay between the physically real edge of the signal voltage at the terminal input and the time stamp in the EtherCAT slave controller (ESC) is kept below 1 µs by the optimized electronics, but is not indefinitely short. If the user has a particular need for precision, calibration must be carried out, and the precise time delay measured in accordance with the environmental conditions.

The terminal provides the following information for each channel (see Fig. TwinCAT tree, EL1252):

EL125x - Functioning 2:
TwinCAT tree, EL1252

Status Byte

EL1252

0

0->1

1

1->0

ContinuousMode (default)

0 (b#00000000)

0 (b#00000000)

0 (b#00000000)

0 (b#00000000)

SingleEventMode

0 (b#00000000)

1 (b#00000001)

1 (b#00000001)

2 (b#00000010)

 

 

 

 

 

EL1254

0

0->1

1

1->0

ContinuousMode (default)

0 (b#00000000)

4 (b#00000100)

4 (b#00000100)

0 (b#00000000)

SingleEventMode

0 (b#00000000)

5 (b#00000101)

4 (b#00000100)

0 (b#00000000)

The times of the edges are made available in the form of 64-bit times, i.e. as 8 bytes of process data based on the Distributed Clock of the terminal.
They are constructed as follows: LatchPosXXXY (XXX: POS/NEG, rising or falling edge; Y: channel, 1 or 2).

EL125x - Functioning 3:

Temporal consistency of the input data

The input data of the EL1252, EL1254 can be functionally divided into two groups. This is also visible from their storage location in the ESC registers:
- Input channels (Channel x Input) and Distributed Clocks status bytes
- Distributed Clocks latch times.

The two data blocks are updated at different “speeds” - whilst the input data is read in and made available upon the arrival of the Ethernet frame at the EtherCAT slave, the DC latch times can still be updated during the processing of the datagram on account of the external latch event. Therefore, the latch times can still change up until a few ns before being read into the Ethernet frame - this ensures that it is really the latest possible status of the inputs that is mapped into the frame.
Example: Low/0V is present on channel 1 and the collecting Ethernet frame arrives at the EL1252/EL1254. During frame processing, but before the DC latch data is read in, a rising edge occurs that leads to LatchPos1 being updated - this updated time is then written to the collecting datagram. The two pieces of information, “rising edge” vs “signal level”, are thus inconsistent in this case.
Recommendation: In order to precisely analyze the input channels, especially in the case of short pulses at the inputs, exclusively the 64-bit time stamp LatchNeg/LatchPos should be evaluated.