lm317 kit - a closer look

[@stephendredge6077]

The transistor turns on and clamps the output to about 2 volts if the Pot wiper goes open circuit or noisy.Without it bad contact on the pot wiper will allow the output to go to maximum. Potentially damaging the device connected to it.

[@andyfraser5876]

That's a very good point. When I built this kit a few years ago, I couldn't see the point of the transistor, so I left it out and bridged the bottom of the pot to 0V. I've never had a problem with it, but I now see why the transistor is there. It's very unusual for Chinese kits and modules to have any surplus components, especially a 'just in case' one, as this is.

[@carloca71]

Very good point indeed, I remember that old style "adjustable" power supplies with a rotary switch and a bunch of zener diodes, the went full value for a short time while operating the switch who knows what could happens, I bet there are stuff done that way until noways, for peace of mind always disconnect the circuit when adjusting the voltage specially when using that cheap power adapters or (in)famous power supply kits :-)

[@pileofstuff]

That does seem like a reasonable theory.

[@jerril42]

This is better, speed rounds are okay, but I enjoy emphasis on how it works. Great breakdown, loads of good comments. Thanks for the quality time.

[@phillipbartlett1819]

Keep up the explanations of the circuits. I know about 95% of everything electronics but you always teach me something new that makes me even more curious about what I don't know. Thanks

[@jp-um2fr]

I can't say I agree with throwing resistors at the LM317 and hoping it might work is a good idea. The formula is quite simple. Don't just grab a 5K pot, measure it first as most are quite a way off. All you need is the pot full resistance value and the max volts out you have from your supply ( - 1.25V.) The lM317 is OK for 1.5A but the LM338 is good for 5A. Two of those in parallel and you have quite a good bench power supply. Just remember a heatsink the size of a loaf and the biggest PC fan you can find. You can go posh and add say a TIP35C or more.

[@TheEmbeddedHobbyist]

An important point is that the difference between the input voltage and the output voltage is dropped across the regulator. So heat is a big problem, 15v input output at 5V 1A. You have 5 watts on load but 10 watts on the regulator. So if the input voltage was 37V you might end up needing two in series to share the heat generated. The diode between the output and input could be there as some regulators don't like voltage on the output and nothing on the input.

[@MACYNET323]

Can you suggest a kit or already made small circuit with 2 or 4 LM317?

[@TheEmbeddedHobbyist]

@@MACYNET323 Sorry I don't look at many kits, butan AliExpress search might find you something.

[@fredflintstone1]

Nice breakdown of the kit design :-)

[@TheEmbeddedHobbyist]

Just like to add, with just one resistor you can make it a constant current supply. Select resistor value so that it drops 1.25V at required current. Connect resistor between output and adjust pin, connect load between adjust pin and 0V if you see what I mean. Great vid again.

[@pileofstuff]

I was hoping t try that in the future for driving higher power LEDs.

[@TheEmbeddedHobbyist]

@@pileofstuff I think switching regulators are best suited to thing this that. Linear power supplies are better for low noise applications switch mode supplies are better for high currents. If you wanted to drive a 3V LED at 1 amp from a 12V supply you would have 3 watts for the LED and 9 watts for the regulator. With 12 watts taken from the power source! Using a switch mode regulator with 75% efficiency it would be 3 watts for the LED and 3/.75 4 watts from the power source. I think that’s right, it’s 23:43 so time for bed in the UK so not the best time for thinking. :)

[@pileofstuff]

@@TheEmbeddedHobbyist Where's the fun in simply going with the "best" off the get-go? I'm mostly just tinkering with alternate possibilities.

[@TheEmbeddedHobbyist]

@@pileofstuff i think I've been doing this for so long that for power IMHO:- Linear = low noise+heat switching = low heat +noise But tinker away but remember if the data sheet said 100A current limit for Iout, it means the pin might take that but there is no heatsink that would fit in a box you could bolt it to for the chip to take it. So happy tinkering and making great vid's i can watch in the morning.

[@englishrupe01]

@@pileofstuff Lol......that's right: we have to get to the finish having suffered, don't we? At least that's how it always seems to work for me...lol!

[@helmuthschultes9243]

Suspect the transistor base/pot circuit is not traced correct. As is with wiper at ground the base will always be biased off until collector base breakdown exists. For the transistor involved I think that is above max rating of LM317 and so should not happen.

[@fuzzy1dk]

The transistor is there do clamp the output voltage to a low value if the wiper on the pot goes open circuit, which it often does when it gets worn and dirty. Without the transistor the output goes to max voltage if the wiper on the pot goes open circuit

[@wrex7044]

The ”fuse” is a Polyswitch.

[@Mr_Meowingtons]

What's the point of the LM 317 having that driver transistor?? I though that was crazy to.. At first maybe I thought it was for some voltage sensing but it still doesn't make any sense

[@MACYNET323]

Sir can you explain if is possible add another lm317 on this kit?

[@pileofstuff]

Depends on where yo want to add it and for what purpose.

[@MACYNET323]

@@pileofstuffI Need 3A for XML Led on linear driver.Would be interest see a circuit with 2 or 4 lm317 on parallel

[@grahamrdyer6322]

I suppose they might have a use but I tend to use the MUCH more efficient buck or boost regulators, I just like switch mode, I would only use those regulators for RF.

[@pileofstuff]

I like them for audio as well. Or when I want to keep a very low part count.

[@zx8401ztv]

I also don't know why the transistor is there, the pot and typically 240 ohm resistor is all that's needed, plus a smattering of caps.. I have used a line of npn transistors to switch in different resistors using a 4017 decade counter and a push button. That was for the program voltages on an eprom programmer, donkeys years ago lol. And of cause you can jack up the fixed regulators in the same way.

[@jyvben1520]

see the Stephen Dredge explanation, these comments maybe split apart, how does youtube organize these ? i don't know. at this moment both comments are sequential

[@zx8401ztv]

Jyv Ben I don't know what youtube was thinking by tampering with the comment order, it used to be ok untill some pillock messed about :-(. I often just type the persons handle on my rambling messages just in case the dredded youtube plays musical chairs :-D. Re stephen, i would just wire the pot different.

[@jyvben1520]

@@zx8401ztv if i reply then the name is already there (chromebook/chrome) easy

[@zx8401ztv]

Jyv Ben, ahh that must be the reason that my comments on streams would be to no one, unless i typed there name in. I'm still using win xp pro 64 for the last ten years and most browsers don't support it. No its not full of viruses, a decent firewall and virus protection is updated regular. People are not carefull enough on line, can't fix stupid people. Sorry waffle mode :-(

[@jyvben1520]

@@zx8401ztv there name, typo, their name ;-) win xp pro : wow ancient, even windows 7/8 is now unsupported (free)

[@DrexProjects]

There is a transistor in this circuit but in a different spot. It says it is for Increasing LM317 Current. makingcircuits.com/blog/increasing-lm317-current-with-a-transistor/ So I guess that's not what it is for. But this link may explain it makingcircuits.com/blog/high-current-lm317-power-supply-circuit/

[@jensschroder8214]

There is also the LM337. It is a LM317 for negative regulation. A online calculator for the LM317 is heer: www.netzmafia.de/skripten/hardware/LM317/LM317.html

[@helmuthschultes9243]

Possibly the wiper goes to base

[@Chris_the_Muso]

The only thing I can think of for the transistor is that if the adjustment voltage goes very high (ie broken or intermittent wiper in the pot), the transistor will turn on and limit the output voltage to a safe level - essentially the turn on voltage of the transistor plus the 1.25 volts, plus maybe a bee's dick more in windage. So.... maybe 1.85 volts plus a bit, probably under 2V. I looked at a lot of suggested schematics trying to figure this out, but absolutely *none* of them had this transistor. It's not like the Chinese to add in superfluous components just for the sake of it though, so I'm guessing they had problems with the pots in early versions of this kit, and maybe this particular circuit is the result of that (add in a 5c transistor so that they can continue to use a nasty 20c pot essentially saving them about 60 cents on every kit while keeping the pitchfork brigade off their backs).

[@pileofstuff]

That's a solid theory.

[@boonedockjourneyman7979]

Keep looking for what others miss. Also, take a look at the "all about circuits" website. At least three KZbinrs are ripping them off verbatim for content. The problem is that the site gets things about half right. You could help a lot of people by taking a look, the bad guys are obvious. Sort out the crap these frauds are repeating. I have several students who have gotten sucked into the KZbin "learning electronics" thing only to be taught by frauds ripping off half right websites. Hey, these guys even run Patreons based on dead rip offs of classic websites.