Clarification on how Holder block works

[gabrielenava]

I did the following test: a ramp signal (starting ramp value: 10) is sent into a subsystem, containing an HOLDER block. Every 2 seconds, the subsystem containing the HOLDER is enabled/disabled for a period of 1s each. See pictures attached.

Inside the subsystem:

These are the output: (Simulation time: 30s)

• Signal sent to the ENABLE block:

• Ramp signal (no HOLDER)

• Ramp signal passing through the HOLDER block

Now we know that the output of HOLDER block will always be the first value sent to the block. Maybe this may be an issue for some applications.

[francesco-romano]

Thanks @gabrielenava
I think that it may be worthwhile to move the implementation of the Holder block to c++ (the implementation is trivial).

Currently blocked by #46

[S-Dafarra]

This is not a problem of Holder block, but rather it is an intended behavior of the Enable Simulink block.
If you select reset from the drop menu in the options of the Enable block, the internal state of the Holder block is reset when enabled, thus changing the value of the Holder with the new one.

[francesco-romano]

Thanks @S-Dafarra.
I knew about that option, but I don't know if @gabrielenava tried it. @gabrielenava can you try if the holder block still exhibits the same problem with this option changed?

[gabrielenava]

No it doesn't. I didn't know about this option, changing it solves the issue.

[francesco-romano]

Great. For the time being this can remain as simple matlab code if it is compatible with code generation