4
\$\begingroup\$

I have problem with the model of given power limiter (below are the images). Problem is in the red rounded resistor on the last image - that resistor overheats for some reason (I've put 3 ohm one instead of 2.2 ohm).

What should I do?

enter image description here

enter image description here

\$\endgroup\$
  • 5
    \$\begingroup\$ Please don't ever do that again: posting 3.5 Mpixel images, and zipped as well! Scale them first, then upload (unzipped!) to an image server and notify us. We'll insert the images in the question. But not the way they are now. \$\endgroup\$ – stevenvh Jun 22 '12 at 15:26
  • 6
    \$\begingroup\$ @stevenvh, uploading higher quality is okay, the images should re-scale automatically with even mildly decent browser. If someones browser is not supported they need to take that to meta not demand smaller images from users. On the note of zipping, I would have left that to the user to correct, that can hide viruses and such. \$\endgroup\$ – Kortuk Jun 22 '12 at 16:23
  • 1
    \$\begingroup\$ @Kortuk - I have a fast Internet connection, and I have no complaints about the SE servers either, but sometimes I can tell it's a way too large image as I see it being built up on my screen. Most of the time it's not needed. As a matter of fact I found the third image here (I removed the second) easier to read when scaled down to 40 %, that's 84 % less pixels. The second was the 3.5 Mpixel, and contained no useful information. \$\endgroup\$ – stevenvh Jun 22 '12 at 16:52
  • 3
    \$\begingroup\$ @stevenvh, I realize your points, my issue is primarily how aggressive people can be about things as simple as higher resolution pictures. I was mostly responding to the "please dont ever do that again" part. The zip is an issue, but low rep users cannot post pictures themselves and he probably thought it helpful. I would really prefer every user posted too high of resolution images then too low. \$\endgroup\$ – Kortuk Jun 22 '12 at 17:41
  • 3
    \$\begingroup\$ @stevenvh, I may be communicating poorly. Your comment was informative, but I am asking we a bit kinder, and as far as issues go I think too high of res is better. A hint to the OP about uploading might be effective, good meta post, but even with a hint he still cannot upload the file. It just happens. On the note of votes you have 3 comment upvotes and 1 comment downvote(you need 60k rep to see that :P). Votes on comments are great, I share feedback with you because you are very good about listening to it and it is a good use of my time. You were not being chastised, I was just sharing. \$\endgroup\$ – Kortuk Jun 22 '12 at 18:40
6
\$\begingroup\$

What power does the parts list specify for R4? The output current is up to 2 A, then a 2.2 Ω resistor will dissipate 8.8 W. A 3 Ω resistor will dissipate even a lot more: 12 W. Even at 1 W a small resistor will get hot.

Now the thyristor only conducts for short periods, in order not to have a higher voltage on the smoothing capacitors than necessary. And here Wouter brings up an important point. If the current is pulsed it will cause a lot more dissipation. At a constant 2 A current the 3 Ω resistor dissipates 12 W. If we pulse 20 A at a 10 % duty cycle we get the same average current, but when we calculate the dissipation:

\$\mathrm{P = 10\% \times I^2 R = 0.1 \times 20^2 \times 3 \text{ W} = 120 \text{ W } !! }\$

Note that the circuit was designed to reduce power dissipation in the voltage regulator's power transistor.

\$\endgroup\$
  • 1
    \$\begingroup\$ I think it is even worse. Power is I^2*R. In this circuit the current can be active for only a small portion of the time, which will cause a higher power dissipation than a continuous curtrent. This circuit is a poor-mans attempt at SMPS. Maybe replace the R with an inductor and add a diode from ground? \$\endgroup\$ – Wouter van Ooijen Jun 22 '12 at 17:09
  • 1
    \$\begingroup\$ @stevenvh Agreed. Most likely a 10W cement resistor is needed here (these old designs are certainly not efficient...) \$\endgroup\$ – Adam Lawrence Jun 22 '12 at 17:29
  • \$\begingroup\$ Hello, I'm very grateful for your answer and for everyone other's comment on this matter. I have concluded that it's mostly design flow, so I would appreciate if someone could link me to some better design. Thanks guys, you're awesome! \$\endgroup\$ – user10468 Jun 23 '12 at 19:23

Your Answer

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