LM317 2N3055 Variable Circuit Problems

[heavyfletch2nd]

Hi all I am a noob learning and having trouble with building a new power supply i got all the parts needed and built it . i have had it running but it constantly runs with adjust lm317 and 10k pot not working am i missing somthing simple in this circuit that i just dont understand . could anyone explain the components their size and pins and the gnd points . The transformer i am using puts out 16v ac off load after connecting to my built circuit i get 22v dc at my output and the adjust is not working at all mabe a point of a volt only

 

[(*steve*)]

Where did you get that circuit from?

I think you'll find the correct circuit in the LM317 datasheet.

 

[davenn]

Where did you get that circuit from?

I think you'll find the correct circuit in the LM317 datasheet.

uh huh

gotta be one of the worse drawn ones I have seen. It's been drawn in a very unconventional way
took me a few moments to check the pins connections and realise that the 317 was actually wired ok
just drawn badly

 

[heavyfletch2nd]

Thanks for the replies it is appreciated hope I get people with knowledge interested in helping I would like to add extra features into this circuit , right then I think I got it ,I built the circuit In the bread board again with spare parts ,i swapped around the collector and emitter wires on the four 2n3055's and connected the tab of lm317 to the collector ,it is adjusting but very sensitive on the 10k pot only uses half the turn , it only goes down from 21v to 5.5v off load not the 1.5v like it says in the drawing description ,I tested amperage up to 13.6a by connecting a cupple of large resistors on the output with multi metre hooked up ,it melted a crocodile lead in 5 seconds it is powerfull , how can I get the 5.5v down to 1.5v on the lowest setting and can I get the 10k pot to be less sensitive using the full turn rather than the half turn , and I dont understand why 16v ac turned into 21v dc after going through all the components I expected a voltage drop not gain . Thanks for replying I have plans for this so would appreciate the experts help as I go along

 

[davenn]

can I get the 10k pot to be less sensitive using the full turn rather than the half turn

try a 5k pot

and I dont understand why 16v ac turned into 21v dc after going through all the components I expected a voltage drop not gain

Why ?

Rectified 16V AC ( just the bridge rectifier) = 16V - voltage drop across 2 of the diodes ( the 2 conducting for each cycle of AC ) 0.7 across each diode = 1.4V total

16 - 1.4 = 14.6 VDC with lots of AC ripple

we add a capacitor to filter out the ripple and this brings up the average DC voltage

14.6 / 0.707 = 20.65 VDC close enough to your 21V

the other way is Vout = Vin/0.707 - 2 x 0.7

= 16 / .707 = 22.63V

22.63 - 2 x 0.7 ( Vdrop across 2 x diodes)
= 22.62 - 1.4 = 21.23 again close enough to your measured 21 VDC



Dave

 

[davenn]

have a read of this tutorial ......

http://www.bristolwatch.com/ele/basic_ac_rectification.htm

 

[heavyfletch2nd]

Thanks dave . Looking at that circuit diagram I coppied and built why did i have to change the collector and emitter wiring around so the metal plate i have the 2n3055 on now live + out and how and where would I controll the current if its possible . Allso I tried adding green led for power on , the "red red violet gold" resistor gets way too hot so tried others up and down the range until the led was barley lit and all resistors I try for the led will burn the finger.

 

[davenn]

Thanks dave . Looking at that circuit diagram I copied and built, why did i have to change the collector and emitter wiring around so the metal plate i have the 2n3055 on now live + out

I think you must have mis-identified the pins of the 2N3055 the emitter and base shouldn't be connected to the heatsink, as they are the 2 pins
the metal case of the 2N3055 is the collector and if no insulators are used, the heatsink will be at the input voltage, not the output voltage
If you use heatsink pads and screw isolators between the transistors and the heatsink, then you can have the heatsink completely neutral.

Note that the 2 pins, base and emitter, are closer to one end of the case than the other .... this is how you can identify which is which
the output is from the tied emitters, via the emitter resistors, not the collectors


The LM317 appears to be connected OK just drawn badly

this is the standard way ( just one current pass transistor in this case)

Allso I tried adding green led for power on , the "red red violet gold" resistor gets way too hot so tried others up and down the range until the led was barley lit and all resistors I try for the led will burn the finger

Red Red Violet isn't even a resistor value .... maybe you read it wrong ?

Have you resistor and green LED across the input to the LM317 where the voltage is constant
for a 22V supply, use a 4.7k Ω yellow violet red



Dave

 

[davenn]

where would I controll the current if its possible

lets get the PSU working properly before worrying about current limiting
it's not going to be so easy with high current capability

 

[heavyfletch2nd]

I got all the pins correct dave it still not working when I wired it to bread board last night changed a few wires around like I wrote in above paragraph I must have been reeding adj of the lm317 alone not through the 2n3055s . Do you have one of each of these components or similar you could put into a bread board and put together and build that circuit just to see what I mean nothing seems to go through 2n3055 . I took it all apart rebuilt 3 times I just cant get it working . Thanks for all this dave ★★★

 

[heavyfletch2nd]

Why is the pic of lm317 showing pins different to the circuit pic you added ,vin on left side of lm317 vout right side of lm317 and adj middle can you use different pins as adj? I just fried a 10k pot after i put together that circuit with vin on left vout on right and adj middle now I am even more confused!!!! why do the drawings show it that way if its not that way ?

 

[(*steve*)]

On a circuit diagram the components are shown in a way which makes the function of the circuit clear. Part of this involves following certain conventions, one of these being that in general inputs are shown in the left and outputs on the right. The actual physical placement of the pins is often different and you must take this into account when you build the circuit.

 

[davenn]

Why is the pic of lm317 showing pins different to the circuit pic you added ,vin on left side of lm317 vout right side of lm317 and adj middle can you use different pins as adj?

because as I was explaining, the way your original circuit was drawn by whoever did it was totally non standard layout
when drawing circuit diagrams, input is always on the left and output on the right

you still physically wire the regulator as per the regulator pic I put about the circuit layout ....
personally, I think, the person who designed the LM317 should be strung up for designing the pin layout the way he did

Do you have one of each of these components or similar you could put into a bread board and put together and build that circuit just to see what I mean nothing seems to go through 2n3055

I have built PSU's in the past like that, I know they work

Why is the pic of lm317 showing pins different to the circuit pic you added ,vin on left side of lm317 vout right side of lm317 and adj middle can you use different pins as adj? I just fried a 10k pot after i put together that circuit with vin on left vout on right and adj middle now I am even more confused!!!! why do the drawings show it that way if its not that way ?

see my comment above .... ALSO NOTE --- I have labelled the pin numbers on that circuit diagram

I took it all apart rebuilt 3 times I just cant get it working . Thanks for all this dave

for a start just build the PSU without any pass transistors, when that is working and you can adj the voltage correctly, then add just 1 transistor and get that working, then work from there .... if you cant get just the regulator working without the pass transistors, then you still have a wiring problem


Dave

 

[heavyfletch2nd]

Thanks guys it is imbarracing but I did say before I started I was a noob at this ,I searched for a simple drawing with high current found that one looked easy bought the part list and cracked on with it and failed , I didnt know about power in on left and out on the right in a drawing when I wired up what I could see how it was drawn on the one dave posted lol funny ,now I learned my first real lesson 'im smiling writing this' , but going back to the main drawing I need to build has it been drawn in a way that would suggest the pin lay out . Dont give up on me yet I may be asking silly q's and i wouldn't blame you for having a chuckle at me "I am" lol, im a mechanic not a electrical engineer the only drawings I have used are flat pack cupboard or bed they require no electricity , there is a hel of a lot more to consider here than I imagined , im going back to the shed have a go at wiring daves posted drawing again with that bit more knowledge I just learned ill get back on here in half hour or so let you know when I set my shed alite :-D

 

[(*steve*)]

I have wondered why this isn't an option for variable regulators? You could parallel the transistor and its associated resistors at the red points as far as you see fit.

You would probably want a resistor from pin 2 of the 317 to ground as well. (and the capacitor values would need revision)

Placing the 2N3055's on the same heatsink as the 317 would offer some degree of thermal protection and the base current to the 2N3055 (or multiple 2N3055s) is limited, so the output current would be limited to maybe 20A (max output current from the 317 x current gain of the 2N3055 at high currents)

for the newbies, 4R7 is 4.7Ω, and 0R47 is 0.47Ω. Placing R or k, or M, or μ, etc in place of the decimal point is a great way of avoiding 4.7Ω being read as 47Ω if the decimal place disappears. The disappearance trick is mostly a factor in manual typesetting and photocopying -- neither are likely on the internet, but it also helps with preventing errors in manual transcription.

 

[(*steve*)]

Having parsed the original circuit (and my brain hurts) it appears to be topologically similar to mine above with the exception of base resistors (R3) which are not strictly necessary and the change in the feedback loop position.

@heavyfletch2nd, if you need help with your circuit, people can easily help you with one that looks like the one above, but may not be able to with the one you posted -- even though they are almost identical in function. The reason is, the layout of the circuit diagram is very important in conveying functional information to the reader.

If you don't have that skill then any old layout will work just as well. But with experience, a poorly drawn diagram is like getting into a car where the wheel turns clockwise to turn left, and the accelerator is on the left. Sure, you'll be able to drive the car, but you will only be able to do it with a lot of mental effort, and you won't be able to talk to a passenger or deal with an emergency,

 

[heavyfletch2nd]

Back again guys. In the first pic in the bread board this is adjusting without npn 2n3055 connected just a small 12v fan.
in the second pic with 2n3055 connected there is no volts out of emitter.
if you look closley at the third pic of the actual build you will see i have mounted the out put resisors "gold colour" on the same heat sink as 2n3055's , i didnt realise the casing of these resistors were going to be live, ooops dumb ass should have checked first.
now i think all 2n3055's are dead.

 

[(*steve*)]

The emitter resistors should not need to drop a large voltage, and thus should not dissipate so much heat that a heatsinked resistor is needed. If you add base resistors, balancing will be even better.

Yes I'm also surprised the resistors were not isolated from their case, but this is something you need to check before mounting then on a common heat sink.

You should not need a heat sink at all for testing the output voltage as nothing should even get warm without a load.

I recommend you use a current limited power supply for testing this as you can set the current limit low enough that nothing can be damaged.

 

[heavyfletch2nd]

Hi steve . Yes I should have tested them first ,another stupid mistake I know but i am learning 'the hard way' lol ,I have cut off that one side of the heat sink with the out put resistors on it I will bolt that in separately once its powering correctly , this psu is going to be my only psu I dont have a variable current psu untill I have finished this one and added the current limiting to it , what I have found tho is an old 10v 1.5 a adapter wich I am now using on the bread board to test these circuit's before jumping in with both feet first , i know 1.5a is not really that low for testing but its better than using the 15a transformer i bought to make this psu and its all I have left that works wich is another reason I am building a high current psu I ruin all the wall adapters and old laptop chargers I use charging batterys with fast chargers for the rc planes and cars I have ,but I do agree if I had one I most certainly would use it , I built this with too much confidence plugged it in and expected it to just work , I cant do anything until the new 2n3055's arrive on friday now, but I am now pretty sure that I did wire this thing up correctly the first time following the un conventional diagram I originally posted , i think the problem lies with thise 4 out put resistors being on the heat sink LIVE and using the large 15a transformer on the first test run ,again noob mistake, now I blown it all up I can re build knowing what I did wrong the first time and NOT make the same mistakes with the extra knowledge kindly given from you and dave you both really gave me a few things to concider when reeding and building circuits correctly thanks for putting in the time to post to me thanks to you too dave if you're still following I will get back on here on friday when the new 2n3055's arrive unless we get more suggestions on the thread I will reply to.

 

[(*steve*)]

Another newbie mistake is carving up the heat sink when a better solution is to electrically isolate the resistors from the heat sink.

And unless you put each resistor on its own heat sink, you may have to electrical isolate them anyway.

With a 10V input voltage your max output voltage will about 7 volts, so don't be surprised if things don't work as expected above this.