0
\$\begingroup\$

I have a question that reads:

A full-wave controlled bridge rectifier is considered, shown below. Vs = 230V. Load is represented by \$i_o = 10\$ A. Frequency is 50 Hz and the delay angle is 45\$^{\circ}\$.

(a) Sketch \$V_{o(avg)}\$, \$V_{SCR1}\$ and \$i_s\$

(b) Derive the formula for calculating the average value and RMS value of the output voltage

enter image description here

Do I assume that this FWB rectifier has an RL load or an R load?

I am assuming that there is a purely resistive load:

So, they look roughly like this

(a)

enter image description here

(apologies for paint)

(b)

$$V_{o(avg)} = \frac{1}{\pi} \int_\alpha^\pi V_m \sin{\omega t} \text{ d} \omega t = \frac{V_m}{\pi}\left[-\cos{\omega t}\right]_\alpha ^\pi = \boxed{(1+\sin{\alpha}) \frac{V_m}{\pi}}$$

and RMS:

$$V_{rms} = \sqrt{\frac{2}{\pi}\int_\alpha^\pi V_m^2 \sin^2{\omega t}\text{ d}\omega t }$$

Then, I think I just use \$\sin^2{\omega t}=\frac{1-\cos{2\omega t}}{2}\$ to contiune the integration. Is this right, or am I making things too complicated?

\$\endgroup\$
8
  • \$\begingroup\$ Think about it a bit. The average voltage is going to be zero because you are modifying positive and negative half cycles equally. \$\endgroup\$
    – Andy aka
    May 1, 2017 at 18:42
  • \$\begingroup\$ @Andyaka Are you sure? My answer for the average voltage agrees with the vast literature surrounding this topic... It would only be 0 if \$\alpha = 180^{\circ}\$... The average value of a rectified sinusoid is not 0... \$\endgroup\$
    – smollma
    May 1, 2017 at 18:46
  • 1
    \$\begingroup\$ Yeah I had a brain reset!!! \$\endgroup\$
    – Andy aka
    May 1, 2017 at 18:57
  • \$\begingroup\$ @Imsavk you can provide an answer to your own question \$\endgroup\$
    – Voltage Spike
    May 1, 2017 at 20:37
  • \$\begingroup\$ And you can mark it as answered \$\endgroup\$
    – Voltage Spike
    May 1, 2017 at 21:44

1 Answer 1

0
\$\begingroup\$

Following the assumption that there is a purely resistive load: by the definition of the root mean square, you can easily get to:

$$V_{rms} = \sqrt{\frac{1}{\pi} \int_\alpha^\pi V_m^2 \sin^2{\omega t} \text{ d}(\omega t)}$$

without any trouble, for any signal. The rest is trig identities and calculus.

$$V_{rms} = V_m\sqrt{\frac{1}{\pi}\left( \int_\alpha^\pi \frac{1}{2} -\frac{\cos{\omega t}}{2} \text{ d}(\omega t)\right)}$$

$$V_{rms} = V_m\sqrt{\frac{1}{\pi} \left(\left[\frac{\omega t}{2}\right]_\alpha^\pi - \left[\frac{\sin{2\omega t}}{4}\right]_\alpha^\pi\right)}$$

$$V_{rms} = V_m\sqrt{\frac{1}{\pi} \left(\frac{\pi}{2} - \frac{\alpha}{2} +\frac{\sin{\alpha}}{4}\right)}$$

$$\boxed{V_{rms} = V_m \sqrt{\frac{1}{2}-\frac{\alpha}{2\pi}+\frac{\sin{2\alpha}}{4\pi}}}$$

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Mark your own answer as accepted instead of changing the title as "solved". It fits better the site policies. \$\endgroup\$ May 1, 2017 at 21:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.