GUI data display

Hello Danilo,

Here is what I understood from your post. Let me know if it is correct.

-          you have limited memory on the target – let’s say an array with 200 words plus few additional words.

-          you are sampling 10 variables 2000 times a seconds.

-          you want your user to see the data of these 10 variables in 10 graphs with 200 points but in a continuous mode – all points in the arrays are continuously sampled. You want him to see clear sin functions instead of random dots.

Considering the slow transfer rate I can try to define a scheme that will try to satisfy these, but it will assume other tradeoffs. Can you let me know if these tradeoffs are acceptable, before I go ahead and experiment with the idea?

-          Each graph cannot be updated every 200 ms. We will try to cycle them as fast as we can, but assuming the fastest refresh rate it 200 ms that means that each graph will be updated every two seconds or so. The data in the graph will be continuous, but there will be big chunks of data, never displayed to the user. The graph for variable one will display samples from 0.000s to 0.200s but will never display variable one samples from 0.200s to 2.000s.

-          All graphs will not be able to show sampled data at exactly the same time. Since all samples cannot be stored in the target memory, we first store samples of variable one for the first 0.2 seconds, transfer it to the host, store variable two for the next 0.2 seconds, transfer it at the host, etc. At the end: graph one with show continuous data from time 0.000s to 0.200s, graph two will show data from time 0.200s to 0.400s , graph three will show data at time 0.400s to 0.600s, etc. graph ten will show data at time 1.800 to 2.000s. then we start updating graph one again at time 2.000 seconds to 2.200 seconds.

 Will solution with these tradeoffs be acceptable to you?

 Dobrin

Hi Dobrin, 

First at all thanks for your prompt reply. You perfectly got my point. I think the second solution is better in my case. To be honest  my main problem is that I will have low frequency signals (about 10Hz) and I need to display at least 2 periods of the signal, but at my sampling rate I can barely display half period. 

What I've got so far is:

AdcaResults[resultsIndex] = AdcaResultRegs.ADCRESULT0;
AdcaResults1[resultsIndex] = AdcaResultRegs.ADCRESULT1;
AdcaResults2[resultsIndex] = AdcaResultRegs.ADCRESULT2;

resultsIndex++;

if(RESULTS_BUFFER_SIZE <= resultsIndex)
{
resultsIndex = 0;
}

Buffer size=80;

Actually I think your idea is brillant since I guess I can store the data somehow on the host computer based on some code on the javascript file in the GUI and keep displaying it until the next actualization.  If you can go ahead with the idea it would be great. I will try on my side too.

Best regards,

Danilo

Hi Dobrin,

Thanks for your previous reply. I had to open the files with the notepad to see your code lines because I couldn't open them with my CCS-GUI installation (v5.5), but they were really helpful.

 Now I have another question and probably it was implicitly solved  in your files but it would be wonderful if you can give me further explanation.

 I want to be able to read/write variables in the target without binding them to widgets. I'm trying to do something more complicated but just to give you an example of what I'm trying to do

Let's say I want to read the variable called  "Dc_volt" (uint16 type)  in the target and multiply by two and write the result in another variable called  "Dc_current" in the target as well.  What I'm doing so far is I  bound "Dc_volt" to one text widget  (value field) and "Dc_current" to another text widget and use the preprocessing function to call a function read_writex2 which looks more or less like:

function read_writex2() {

wc= dijit.byId("widget_161");
z3=dijit.byId("widget_165").get("value");

z4=z3x2;

wc.set("value",z4);

}

So as far as I understand  the function read_writex2() is call from the widget and then executed. I was wondering if it is possible to do something like this but directly from the GUI without the bindings to widgets. I mean each refreshing time the GUI calls and executes read_writex2() and has access to the variables Dc_volt and Dc_current from the server rather than from the bind to the widgets. I know I can uncheck the visibility field on the widgets and have some "dummy" widgets just to read/write and execute functions but it doesn't seem to be efficient or at least a good coding practice. So hopefully you can help me about this with some advice/ explanation about how to do this or some references. Thanks in advance for your help.

Best regards,

Danilo