I was doing some transient response testing on a 30 amp, 14 volt DC power supply I just built. I found that when a load was applied (a 20 amp load) the voltage dropped down to almost zero for about 50 uS then popped back up to 14V, then upon removing the load the voltage spiked up to about 18 for about 2 uS, then the LM317 panicked and shut down for about a millisecond, then came back on. So THAT'S what they mean by "the cap will improve transient response".
Anyway, just for fun, I took a big can (100,000 uF, 25V) cap and connected it across the output, then I had to go take a leak. While I was leaking I heard a loud bang which scared the beans out of me. When I came back to the power supply, the room was full of very familiar smelling smoke and shreds of paper everywhere.
I'll bet you can guess the dumb, DUMB DUMB!!! mistake I made!
Darn... that cap cost $25 USD, now it's splattered all over my workroom.
Hutkikz:
Ahh the memories of when I first learned that.
When I was in high school (1971-1975), I used to hang around the machine shop (they wouldn't let me take shop because it was for the "dumb kids").
Anyway, the shop teacher used to shut off all the breakers at the end of the day (no power to machines, lights, outlets... anything).
So, being the prankster I was (am?), I took a handful of radial lead electrolytic caps of 1000 uF if I recall and plugged a bunch of them into all the wall outlets.
Big surprise, lots of bangs, smoke and paper confetti the next morning!
That's where I learned what aluminum electrolytics smell like when they explode.
My very first job was repairing televisions on a factory production line. It was always very obvious when someone new was hand inserting the power supply caps. As soon as the sets were powered on the production line, the noise and smell was fantastic. You quite often got 5 or 6 sets going off before someone shut the line down.
Hi,
I had that happen when demnstrating to Amateur Radio class, power supplies,
AC == Transformer== Rectifier==Smoothing Cap oops, didn't check peak voltage.
While drawing on blackboard, remember them, RB Cap "bang" shot to the ceiling, where it collapsed like a concertina.
Lots of aluminium and paper and smoke.
Next time I taught it, I carefully calculated peak voltage so it would do it again.
I didn't say why I had a small perspex shield in place.
Still scared the chalk out of my hand.
Its surprising what equations and circuits are remembered when things go bang?
Tom.... :o :o :o
PS, I wasn't around when this went BANG.
krupski:
My cap looked very similar... except the aluminum can blew completely off.
That supply had a cover over it, the blue plastic condom on the can was squashed up against inside of the cover.
Under the blue condom the can is squished.
The power supply still worked, it was the smoke that alerted the customer that something was possibly wrong.
Its part of a 800V DC bus for a microwave power inverter.
Tom...
I can post the diagram of it for you if you're interested.
Please do. I have a 20 volt 16 amp transformer I've had laying about that I've wanted to build into a decent analog supply for awhile. I've just been too lazy to design something. Have a bunch of old hermetic packaged '723s that I had considered for the core but never got around to it...
avr_fred:
Please do. I have a 20 volt 16 amp transformer I've had laying about that I've wanted to build into a decent analog supply for awhile. I've just been too lazy to design something. Have a bunch of old hermetic packaged '723s that I had considered for the core but never got around to it...
Sure... here it is (outputs 13.9VDC):
(click pic for full size)
Sorry it's a kinda crummy quality image, but I used an image editor to draw it....
avr_fred:
Please do. I have a 20 volt 16 amp transformer......
20 volts may be a bit high. As it is, my supply uses an 18V RMS transformer which gives about 25V peak and about 20V under full load. The NPN transistors have to dump 20-13 = 7 volts at 40 amps which is a whopping 280 watts!
I may end up using a different transformer to minimize the drop while still having enough overhead to maintain regulation.
I get away with the horrible power waste because the supply is for an amateur radio (ham radio) transmitter and the full load is only applied intermittently (either CW or SSB - both of which load the supply with probably a 30 to 40% duty cycle - full continuous 40 amps would probably overheat it).
I don't have a good power resistor array to test the supply under full load... all I have is an array of eight 4 ohm power resistors wired in parallel which only gives me a 26 amp load. The supply gets quite warm, but handles the 26 amps continuously with no problems.
I heard a loud bang which scared the beans out of me. When I came back to the power supply, the room was full of very familiar smelling smoke and shreds of paper everywhere.
