Is absolutely necessary an external oscilloscope to display waveforms of PWMs ??

Hi Mileva!

It is not absolutely necessary. But at least in my opinion any other device or softwares do not provide the accuracy of real-time display of any process better than oscilloscope.

Regards,

Igor

In orden to show all the grafics i have to go to "tools==> Graph ==> single time"  and set the Graph properties in real time simualtion, to show for example "AdcBuf"

I haven't an oscilloscope in my home, is important to know so that i can show all my signals like PWMs signals..

Yes, you can make use of the graph tool available in the CCS.

Regards,

Gautam

The graph tool can plot data from memory. It cannot plot signal generated by PWM peripheral.

Regards,
Maciej 

Hi,

Can't you feedback your PWM outputs to you CAPture inputs? This way you're able to measure your PWM pulse width and, if triggered by PWM interrupts, you'll be able to do some interrupt based logging by add the captured results to an array creating your history. Than you're able to watch the array in your watch window or just export the array to your excel workspace (select al, CTRL+C, Paste in excel) and generate a Graph here.

You're not able to capture your HiRes resolution, but at least your SYSCLK resolution (10ns@100MHz).

Best regards,

Tjarco

Tjarco Boerkoel said:

Captured results to an array - creating a history.  Then - export the array to Excel

Applaud Tjarco for this post - Excel (& similar) are so powerful - so often identify problems - lead to better (& faster) solutions.

Tjarco's suggestion of feeding back key outputs - to CAPture inputs - inspired! 

While always best/safest to monitor the actual, "real-world" conditions (as suggested above) you may also consider regularly "logging" your key PWM variable - over the time/range of interest.  (but do this after you've confirmed that the variable selected very well reflects actual PWM duty (or other item of interest)) 

Here's "proof" - Tjarco's idea - harvested via transfer - ARM's SRAM to Excel: (BLDC motor - 2300RPM Target - under load)

Scale: Hor mS, Vert. RPM (top), PWM variable (bottom)  Data (2 of 12 different - 32 bit values) captured @ 1mS intervals

Hi Guys,  i really apreciate your help, and i would like to know the following question,  if i'll conect pin 3 (ADC-A0) with  pin17(EPWM-1A-GPIO00) by a cable conector(wire), i can show the AdcBuf, and that will be the signal of my PWM indeed?, am i right?, thanks in advance..

"I use the microcontroller TMS320F28027"

Sorry - but as I'm here...

PWM may not be properly monitored via "direct connect" to an ADC.  Recall PWM is a sequence of width modulated - full-scale swing (i.e. 0 -> 3V3) pulses.  And "usually" in the frequency range "above human hearing" for many applications.  (although 100Hz or so should work for simpler applications - but "filtering" of such low frequencies will prove a challenge.  {i.e. l  o  n  g    time constants}

What may work is the creation of an external R-C network (and the R-C's output then fed to the ADC) - which will yield a DC Voltage directly reflective of the PWM's duty cycle.  R-C values depend upon the PWM frequency and the response rate you desire.  (cb1 departs - ARM land...)

Hi Mileva!

There are a reasonable low-cost and convenient hybrid solutions with the USB interface and the display on the PC monitor like this (and with a spectrum analyzer function also)

http://www.velleman.co.uk/contents/en-uk/p382_pcsu1000.html

Regards,

Igor

I bought one of this about 2 years back, very useful, absolutely taken care of my oscilloscope requirements.

Hi Mileva!

Mileva K. said:

My exercises consist of generating signals PWMs of 10KZ, 50KHz and a signal PWM with period of 500 miliseconds.

1 Objectively. This is a acceptable option for your current task.

2 Subjectively. AD resolution 8 bit is not enough (better 12).

3 Rerspectively. 10 MHz is not enough (better 50-100). But now it is difficult to determine your future needs. 

You need to decide himself. Based on your financial possibilities.

Regards,

Good luck,

Igor

It is a very hard job to fully & properly anticipate both your present and future needs.

And - as past US Defense Secty. stated, "you encounter known (unknowns) and unknown (unknowns.)"

Unless you can "steal" that scope - believe you will be far better served by choosing a PC/Hybrid Scope package.  The "front-end" of PC/Hybrid will vastly exceed that of the local shop's offering - and the front-end is an especially crucial part of any scope capability.  (proven by Tek/Agilent emphasis)

Most here will guarantee that your needs will expand - beyond your "current" yellow highlights - and the PC/Hybrid solution best promises to meet your present and future requirements...