LM317 2N3055 Variable Circuit Problems

[heavyfletch2nd]

good point dam it why didn't I think I could kick myself , they're all conducting via outer casing so wouldn't be any point in having four separat resistors lol thanks , I'll tig it back on and get some insulators I may aswell get some for the 2n3055's too as its all coming apart for round two on friday ding ding , its a extra pice of mind if its all insulated thanks for the kickstrat there steeve lol:-D

 

[heavyfletch2nd]

I cant even get this ↓
to work properly when I copy it into this↓↓
even all colour wires are kept the same so I could'nt make a mistake , I only get 0.8 of a volt adjust , tried new 317 and new 10k pot , i could bang my head on a wall !! Give me another kick start steeve please

 

[Mongrel Shark]

I just started reading this thread when I saw this.

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

Usually when I do that I need new 3055's.
Not saying thats your problem, but something to consider in future.

Technically, if everything is right, your circuit should do that current just fine. But turning it up to 11 before fine tuning is complete is normally a good way to fry stuff in my experience. On the other hand its also a good way to find bits that could use improving. You can get the smoke to show you where the bad bits are... If the resistors all turn cherry red you need bigger ones etc etc. You probably want a fan for those transistors.


As the to the last post. Looks like you are trying to put DC through a series capacitor (blue cylinder?). I don't think that works very well.. I could be wrong, are there other connections on the other side of your perf board?

You have a blue cylinder in the bottom pic that’s not in the breadboard. So clearly you did something new there. Also the breadboard has a bonus yellow output wire. Would it help if that was connected to a load on the perf board? I find my circuits work best with both input and output connected at the same time.
Hint, If something is made to have a load, and you run it with no load. It may not work right.

 

[heavyfletch2nd]

a new board without a capasitor I had in there , "the light blue shiny one" I changed that for a normal looking cap 16v 10uf , put in a new lm317, " I only have 2 left out of ten" now I got adjust in my prototype board and not frying the lm317 , I have connected jumper leads to a new 2n3055 it is emitting and adjusting , at last im getting somewhere lol that was a new cap 40v 10uf

 

[Mongrel Shark]

lol. If you don’t fry at least 4 transistors a session. Your not trying hard enough.

I did that my first few power transistor projects too. I save my dead 3055's. I have over 100 now.

Now I just get everything in a 100 pack or bigger.

 

[heavyfletch2nd]

Its definitely a learning curve "brain bashing" lol hopfully I'll get this circuit in some sort of built state tomorrow mabe even use it on a six amp fast battery charger for a test run see how warm things get but I must sleep its 3:17am uk , is it nice n sunny where you are opposite side of the world

 

[heavyfletch2nd]

Right then ( all new components ) it was all working fine on the 10v 1.5a wall adapter just a bit low on the volts as expected , I put In the big transformer connected it all up connected multi meter to ouptput with no load and it started at 21.6v a minuet later the volts started dropping and adjust started failing slowley , i quickly connected a load 12v 1.5a small fan it then stablised at 8v with half the adjust working , it was like watching the lm317 slowley dying until the load was connected , is it ok to put 21.6 v through pin 3 input on the lm317 it is being fed from the main live rail ,
i connected a battery charger just to see how it would react and got 8v again , it shows 11v off load but starts dropping slowley . But this is definatley progress compared to the start of the project i am pretty pleased with what i have learned so far, its been very interesting/addictive five hours in the shed today lol

 

[ModemHead]

It looks like from your earlier picture that you are using a 2200Ω (red-red-red) resistor for the top of the divider. This resistor value is too large to provide the necessary minimum load current, so the LM317 will not work when the supply is unloaded.

The LM317 needs a minimum load in the neighborhood of 5mA. Since the regulator tries to maintain its 1.25V reference voltage across the resistor at all times, this means the resistor should be no more than about 250Ω. (1.25V/.005A = 250Ω) The power dissipated by the resistor will only be about 6mW (0.005A * 1.25V = .00625W) so a power resistor is not required.

As a side note, driving an NPN pass transistor with a constant voltage will result in poor regulation and a large dropout voltage. The recommended way to boost a 3-terminal regulator is with a PNP device driven by the input current to the regulator. Examples of this technique is in most 3-terminal regulator data sheets.

Also, the current gain of a 2N3055 can drop into single digits at high currents, so it requires a *lot* of base drive, all of which has to be provided by the LM317, which is probably going to get taxed considerably here. Good thing you have a large heat sink for it.

 

[heavyfletch2nd]

Thanks for the reply.. i think im close to victory here its getting fustraiting , this is a lot harder than I expected to be when deciding to build a high current adjustable psu , I got it working "sort of " just not properly ,.yet, are you saying I should try it without the resistor in it and it should work because the power requirments of the base of 4x 2n3055's , what would I put in its place if I do that?, I have two 12v fans running off the lm317t pin 3 's live feed is that ok to do? , and when I watched the voltage dropping on the meter what was that doing with no load? , Was that the lm317t dying due to that resistor ? This circuit is working in the bread board using a wall adapter rated 10v 1.5a but i am only running bread board with one 2n3055 . Thanks for replying I need all the advice I can get being a first timer :-D

 

[Bluejets]

Those Turnigy multi-chargers are not designed to be used like a regular power supply.

Very confusing to see just what is connected to what on that breadboard #22 top

 

[heavyfletch2nd]

?? Im powering the the charger with the "not quite built correctly yet" power supply.

 

[Bluejets]

Can't follow what you are trying to do.
Those Turnigy chargers take an input from 11 to 17v and have their own microcontroller to give appropriate output to whatever battery type one is charging.
Why you want to feed it with a voltage regulated supply is beyond me.

 

[davenn]

It looks like from your earlier picture that you are using a 2200Ω (red-red-red) resistor for the top of the divider. This resistor value is too large to provide the necessary minimum load current, so the LM317 will not work when the supply is unloaded.

agreed, way too big
@heavyfletch2nd why are u using that value when the data sheet specifically shows a much smaller value ... 240 Ohms ... ?
This is going to totally screw up the regulator operation

As a side note, driving an NPN pass transistor with a constant voltage will result in poor regulation and a large dropout voltage. The recommended way to boost a 3-terminal regulator is with a PNP device driven by the input current to the regulator. Examples of this technique is in most 3-terminal regulator data sheets.

With that I also agree, normally 2N2955 devices and often a smaller driver transistor from the input to the reg.
but depending on your project, the above will work ok



dave

 

[ModemHead]

Sorry for being unclear. My first point was that the 2200 resistor should be more like 270, 220 or 180 ohms for the LM317 to regulate voltage properly with no load on the supply. This resistor is necessary for the LM317 to function, it cannot be left out. It is OK to run fans from the unregulated rail.

Second point was that for very high current, the LM317 may struggle to provide enough drive current for the 2N3055s. The resolution for that is to consider a better design, as recommended in the datasheet.

 

[heavyfletch2nd]

Thank you dave I get a bit confused with the ohms when people write it down as 220 or 2200 , my mistake I just do not know enuf yet to know these , so let me get this right and ask probably another silly question "and you can all chuckle if you like its fine im on L plates here lol", so I'm currently using resistance in the thousands on the lm317 and should be using in the hundreds ?? ( If I dont see a k it is not thousand ) I thought (red red red gold )was 2 hundred and 22 ohms 5% , where do the extra ( 0 ) come from on my red red red i am using at the moment , yes go on laugh lol thanks for replying dave . "And bluejets" I need high current to run two of these chargers at the same time each charger pulls up to 6amps on fast charge max setting hence the high current psu , they dont work in the shed on wall adapters and they cook laptop psu's, they only work properly when im out in the car, I can run both off the car battery just fine on max at the same time on the field , I also wanted a adjustable power supply for my shed not just for these chargers, to run other other things that aren't 12v and to test/build other projects in the future , and just to learn this kind of stuff as its really interesting addictive and satisfying to learn and build my own things, every one starts somewhere this is where im starting with this adjustable psu . Thank you all for putting in the time to post to me I hope you all follow this closley so I can learn from you all and mabe add to this circuit if / when I get it running properly please keep replying I have a amp meter and a volt meter to fit to this when I finally get my head around my learning curves . Again I thank all of you , remember I am a real noob not done this before L plate first timer lol

 

[duke37]

red is 2
red is 2
red is two 0
so 2200 or 2k2 or 2.2k

 

[heavyfletch2nd]

Ok thanks the multiplyer in this case is red=2×0 im picking it up slowley, i just downloaded a resistor chart to look at What colour is 250-260-270 like i am supposed to be using? , red blue red then like chart shows ? and what is the functioning difference up and down these ranges . Is 2200 adding too much resistance or not enough resistance just so I know if im starting process this properly in my head . By the time I get this working I will probably be down to the skull from all the head scratching lol

 

[duke37]

You are more likely to find a 270Ω resistor since it is a preferred value.
red 2
violet 7
brown one 0
so 270.

I have not been following the thread so will not comment on the value.

 

[heavyfletch2nd]

I just found one a red purple brown . I put it in got good volts out of the output 19.7 i thought brill its working but then the 10k pot smoked ,

 

[heavyfletch2nd]

And again the 317 and 10k pot fried just like overloading ??????