No I have not used heat sinks as yet. My usual practise is to power up for a about a minute, no load, just to check the voltage. then I would add heat sinks and a load. so no shorting tabs. the rebuild will be finished this evening. I shall post later.
Ok so I did a rebuild and now it is working fine. I still have no idea what was wrong as I checked my circuit countless times. Thanks for your help guys 🙂
An educated guess
Then if nothing else proves to be amiss, possibly a highg-frequency oscillation caused by some facet of physical layout and/or wiring variation.
You should have an ~0.1 uF cap connected from the input pin to ground, with as short length of leads as possible. Still tecnically the regulator should "fold back" at max Iout and hold that current to a reasonable degree. But ultrasonic oscillations, that typically go well into RF ranges of frequency can be destructive, as you would be operating the IC outside its specified parameters. And the one that blew may have been slightly weak but OK sans oscillations.
Do you have an oscilloscope or frequency counter, which could help trobleshoot the problem? With a scope you would turn the vertical gain way up, while turning the scope's frequency knob up in steps. It can be quite hard to observe, but if there.s any such oscillation present, you don't want it to be. That's why the bypass cap is recommended, close to the input pin to ensure stability.
As an enfineer I worked for when starting out explained, a three terminal regulator is a essentially an amplifier with very high current gain, it wants to amplify any rogue oscillation present. And evidently they don't handle that all too well.
Good Luck & Best Regards, "DAAB" -
Then if nothing else proves to be amiss, possibly a highg-frequency oscillation caused by some facet of physical layout and/or wiring variation.
You should have an ~0.1 uF cap connected from the input pin to ground, with as short length of leads as possible. Still tecnically the regulator should "fold back" at max Iout and hold that current to a reasonable degree. But ultrasonic oscillations, that typically go well into RF ranges of frequency can be destructive, as you would be operating the IC outside its specified parameters. And the one that blew may have been slightly weak but OK sans oscillations.
Do you have an oscilloscope or frequency counter, which could help trobleshoot the problem? With a scope you would turn the vertical gain way up, while turning the scope's frequency knob up in steps. It can be quite hard to observe, but if there.s any such oscillation present, you don't want it to be. That's why the bypass cap is recommended, close to the input pin to ensure stability.
As an enfineer I worked for when starting out explained, a three terminal regulator is a essentially an amplifier with very high current gain, it wants to amplify any rogue oscillation present. And evidently they don't handle that all too well.
Good Luck & Best Regards, "DAAB" -
If you're after a variable dual supply, try this:
Simple Dual-voltage Tracking Power Supply for 3 to 15 volts joebrown.org.uk
I've recently built one myself; works like a charm.
Simple Dual-voltage Tracking Power Supply for 3 to 15 volts joebrown.org.uk
An externally hosted image should be here but it was not working when we last tested it.
I've recently built one myself; works like a charm.
i LIKE IT. QUESTION: Since this PSU could be built using the three 3-terminal regulators independently of each other, what is the function of the op amp?
(my electronics background includes a lot of industrial and military avionics gear.
Often they will be designed such as I mention, independent devices for redundancy and reliability. I am now retired and starting to study how "the other side" does it).
Thanks.
(my electronics background includes a lot of industrial and military avionics gear.
Often they will be designed such as I mention, independent devices for redundancy and reliability. I am now retired and starting to study how "the other side" does it).
Thanks.
The op amp is basically the adjustment pot for the 337. This way you only need one simple pot (with only two wires even), and R2 and R3 make sure that the negative voltage tracks the positive as exactly as the resistors are matched to each other. I used 1% metal film without special matching, as that's about the same level of precision one would achieve using two resistors for each regulator in an ordinary configuration. Op amp can be pretty much anything I guess. OP07 works for me, as it's cheap, ubiquitous and has lower drift and offset than a 741 for example. Just make sure to use a vreg with 8V, as with a 5V one the negative rail will go down to about -3V only. I used a 78L05 that was on hand and put a white LED for an additional 2.5V in the ground lead of the reg. Works like a charm. Anything higher like 12V will also work, but since the negative voltage is taken from before the reg, you should watch out to not exceed the max of the op amp. It's already 30V as mentioned in the schematic and 36V is the safe limit of many standard opamps.
Preamp, thanks for the reply. About 2 seconds after I posted that question I realized that your schematic was t tracking type and therefore the reason (in very general terms) for the opamp stage. Your elaboration made my understanding of it less vague.
Most of the three terminal requlator circuits I have built use the LM 309 and LM 337K Steel chips, the 337K type making a 5 amp PSU easy as pie. But a few months back I called the local electonics outlet fo a price, It was about $60 USD, ouch!
So I ordered a few LM350's on line for $1.50 each. I plan to put three of these with their inputs connected in parallel to a beefy transformer salvaged from an APC 650 VA UPS. Then I'll have three 13.5 - 13.8 VDC regulated outputs at 3 A max, splitting the 13.xx V loads between them. Savings of ~ $55 and some redundancy as well.
Thanks again, "DAAB"
Most of the three terminal requlator circuits I have built use the LM 309 and LM 337K Steel chips, the 337K type making a 5 amp PSU easy as pie. But a few months back I called the local electonics outlet fo a price, It was about $60 USD, ouch!
So I ordered a few LM350's on line for $1.50 each. I plan to put three of these with their inputs connected in parallel to a beefy transformer salvaged from an APC 650 VA UPS. Then I'll have three 13.5 - 13.8 VDC regulated outputs at 3 A max, splitting the 13.xx V loads between them. Savings of ~ $55 and some redundancy as well.
Thanks again, "DAAB"
- Status
- Not open for further replies.