Although the version of Simulink is updating every year, I didn’t see any improvement on the most frequently used block ‘Scope’. Sometimes I found it is very inconvenient to use. There are 8 problems/suggestions that I summarized according to my experience:

1. I need a cursor.
Just like a real oscilloscope, I need cursors to read current time tick, to measure X or Y difference. I don’t want to zoom in many times and make subtraction.

2. I want to zoom out X, Y-axes independently.
There are ‘Zoom’, ‘Zoom X-axis’ and ‘Zoom Y-axis’ buttons. But I do not understand why there is no ‘Zoom out’,'Zoom out X-axis’,'Zoom out Y-axis’ buttons. Sometimes, I really need to zoom out X, Y-axes independently.

3. I DO NOT want ‘Autoscale’ button to change my saved axes settings.
‘Autoscale’ button is useful, but not always. After many times of ‘Zoom in’ and ‘Zoom out’, I get the ‘best’ view of the waveform finally. So I ‘save current axes settings’. Afterwards, I want to take a look at the overall waveform so I click the ‘Autoscale’ button. Then, I hope to return to my ‘best’ view so I click the ‘restore saved axes settings’ button. But unfortunately, it does not bring me back!

4. I want to view the waveform from t1 to t2 exactly.
Sometimes I only want to display part of the waveform exactly, e.g. from t=0.15 s to t=0.23 s. However, it seems that the ‘Time range’ parameter does not support such setting. It only allows a partial view starting from t=0.

5. I want to save my waveform after the simulation.
Why I have to remember to check the ‘save data to workspace’ box before the simulation? Many times after a long time simulation, I found I forgot to check this box and I have no choice but to run it again! Why not just provide me a menu to save the data to workspace at any time I prefer? Also, ‘Copy figure’ menu (like the one in ‘Figure’ window) is often desired.

6. I want to organize many scope windows into a single window.
For large scale simulation, many waveforms need to be monitored. But too many scope task buttons on my OS taskbar with undistinguishable title make my feel confused. To find the desired scope window, I have to check them one by one. I hope I can put these scope windows into a single window and use a tabbar to label them, just like the M-editor. It seems that the ‘dock scope’ button can provide this function, but after docking, I cannot use the ‘Zoom’ button!

7. I need a more accurate time tick.
It seems that the time ticks are not accurate enough. After many times of ‘Zoom in’, the neighboring time ticks become the same (e.g. ’0.0001,0.0001,0.0001,0.0001′), although the simulation time step is actually smaller than 1e-4. If I cannot get a cursor, at least I hope I can read the time tick accurately by myself.

8. I hope ‘Autoscale’ function can reserve some margin for the waveform.
Boolean signals are frequently used. But after autosacling, I find the yellow waveform completely overlaps with the figure frame, making it very hard to distinguish between 0 and 1. In this situation, I have to set Y-min, Y-max to -1 and 2 respectively so as to view clearly. This is time-consuming, so I hope the ‘Autoscale’ function can always reserve some margin for the waveform.

A mini-GUIDE would be nice for mask GUI designer.

In R2007b your really limited based on the requirements you have provided. The easiest way to hide IP in that release is to use RTW to Generate an S-function. This does not support inlined code generation, so you have to use an RTW target that supports calling non-inlined S-functions. Even in that situation, I have seen some users generate code, then use a C-language obfuscation tool to provided the source to be included with the final generated model. This would still be needed today as our R2009b IP protection solution is for simulation only.

It has two changes that are especially relevant to my work right now, but unfortunately at this stage I won’t be able to switch versions.

First, I hear it fixes my favorite bug: 529072

That one was killing me on this project, so it’s good to hear it’ll be gone for the next one.

Second, is the IP protection feature. We’re struggling with that right now at the end of our project (I know, should have thought about it sooner!).

I’m hoping you can suggest something that I can do with 2007b. Basically I have a block within a large model that I need to protect. The block with all the IP is currently an Embedded MATLAB code block. We want to deliver the block as part of the model, but the customer must not be able to see the code (they’ll look at the raw mdl text if they have to, trust me). Can you suggest any non-invasive ways of doing that? The end result is built using rtw and is run in a “realtime”-windows environment (don’t ask! :) ).

Great blog, and thanks for any reply you can give. At the very least, the hoops I’m describing here will be an effective sales tool for 2009b. :)

Cheers