LM317 2N3055 Variable Circuit Problems

[heavyfletch2nd]

Hi steve , it is a big transformer I don't have any specs for could it be that? I have 4x 4700uf caps on the input rail, I have 2 spare caps the same they came im a pack of six should I add these spare ones to help with smoothing , could I add diods to all the pins of lm317 , or can we reduce current to the vin on lm317 as 317 is only opening transistors or all of the above ??, just a few thoughts I was thinking of last night while I was in bed , can you see my point on "if I can controll led's or a tail light for a car I should be able to controll transistors?

 

[(*steve*)]

Do you have a diode (or 4 arranged as a bridge rectifier) between the transformer and the caps?

 

[heavyfletch2nd]

No I am using a solid ac to dc 4 leg chip.

 

[heavyfletch2nd]

Done the transistor check on the ones I replaced
all pretty much the same reedings on all 4 , these are not fried then?

 

[Bluejets]

Google is your friend.

 

[(*steve*)]

No I am using a solid ac to dc 4 leg chip.

That is probably a bridge rectifier.

It's still quite possible the problems are external to your circuit. Frankly it may be time to think about using a 9V battery or something equally benign as a power source to begin with.

I may try breadboarding the circuit myself and send you my pictures and results. I have an advantage over you in that I have power supplies where I can set the current limit to a very low value which makes it really hard to break stuff

 

[heavyfletch2nd]

Thanks steve that would be great , just to see if I am the broken pice of the circuit lol , I have been reading about capasitors today and reading the lm317 data sheet I downloaded texas instruments sheet best one I have seen with all the different applications , the recommended circuit for high current looks much different to the one I'm trying to put together "like someone mentioned earlier in the thread" , it also recomends a cap of 0.1uf or higher as close as possible to the vin pin if the 317 is 6" or more away from the main input smoothing caps for stability , i have my 317 very far from the main input smoothing caps at the end of wires is this going to make a lot of difference to the problem im having , also are my main input caps the right sort of cap ? They are 4700uf but I have red about different caps today high and low freq how can you tell the difference , or is the high freq like a 4700uf and the low a 1uf is that how it works or a different kind of cap all together , there are many with different colours , I twistet my brain a bit today with ohms law, kirchofs law and capasitors it is trarting to make more sense to me now but still hard if I didnt have the trusty calculator , I thought about putting a good diod on the vout and a resistor on the vin, I worked it out to be 13.33333333 ohms on calculator I think 20 watts , lol thats a lot of three's I am really unsure I got that right but I'm trying lol , how far away was I ? "Just out of interest"

 

[hevans1944]

Three significant figures is all you need for most electronic calculations, So, 13.3 ohms which (depending on the circuit) would be rounded up or down to the nearest "standard" value, which depends on allowable tolerance in resistance value. Google that. I have no idea if the value you "calculated" is correct or even appropriate.

Datasheets provide a lot of useful information for circuit design, especially if they have an Applications section. Print them out in their entirely and place them in a three-ring binder along with your journal of circuit notes... you DO keep a journal of circuit notes, riiiight?

I wouldn't worry about operating frequency for capacitors yet. That is mostly a concern for radio frequency work and high-speed logic. It has nothing to do with the value of capacitance and everything to do with how the capacitor is constructed. Your 4700 μF polarized electrolytic capacitors are correct for low-voltage power supply filtering. They should be mounted close to the LM317 input terminal and ground... a few inches of 14 AWG connecting the two. The distance from the capacitors to the diode bridge rectifier and power transformer affects only the power losses in the connecting wires. Use 14 AWG for that and you shouldn't have to worry about it until your current requirements get up around 20 to 30 amperes.

Join the amateur (not ammeter) radio club in your area and make a nucience of yourself there. The hams won't mind too much. Most of them love indoctrinating newbies.

Also, unless you are physically handicapped, work on improving your punctuation, spelling, and grammar. Makes it a lot easier for us pedants to read and understand what you are saying.

 

[(*steve*)]

This circuit should look familiar:

 

[(*steve*)]

I'm using a 220 ohm resistor and that pot is almost certainly a lower power unit than yours.

The current flowing through that is 5mA and the max power is 25mW (so I'm not afraid of smoke

I can wind the output voltage down to 1.3V -- no smoke. My input voltage is 12V, but let's wind that up to 31V. Still no smoke, but the current drawn is maybe 6mA.

They do recommend a 120Ω resistor to allow the maximum "minimum current to maintain regulation".

with my 1k pot, the maximum output voltage is 7V and there is no smoke.

I'd say your problem is outside what you are showing us.

 

[ModemHead]

all pretty much the same reedings on all 4 , these are not fried then?

At least the junctions are not shorted, but the reading should be closer to 0.5V. The 1.0+ readings would be a rather peculiar failure mode for a transistor, so I have to admit I'm a bit suspicious of your multimeter.

the recommended circuit for high current looks much different to the one I'm trying to put together "like someone mentioned earlier in the thread"

Yeah that was me probably. The circuit you're working from is not the best, but it should not be burning up pots.

i have my 317 very far from the main input smoothing caps at the end of wires is this going to make a lot of difference to the problem im having

You are right, long leads are not recommended for various reasons. It is not outside the realm of possibility that the long leads contribute to instability of the regulator, leading to uncontrolled oscillation and self-destruction.

 

[heavyfletch2nd]

Thanks for that Steve , what would happen if you changed your pot to a 10k , put in a cap on v out and turned up your current very high "with transistors fitted"? , I do get this working like you have in the bread board , I can now get it working in the prototype board with only one transistor and no cap on the v out pin , I can't turn down the current like you can when moving over to the big transformer , adding two transistors is not working for me , when the other lm317's arrive I will use the wall adapter to power the prototype board and try two transistors with that see what happens , thanks for putting in the time

 

[heavyfletch2nd]

Hi modem head , I did not have this lm317 data sheet from texas instruments when I first started , the ones I downloaded were just bar graphs and numbers wich I did not understand , I am surprised how much I have learned in just one month , I feel like I have come a long way from when I first started this thread , if you have red the whole thread you will see what I mean lol ;-)

 

[(*steve*)]

I just wanted to show you it works

Even though I wound down the current limit, the circuit didn't try to exceed it, so I could have turned the current limit off without any change in behavior.

Place some of your output capacitors at the input of the 317 and don't have any output capacitor to begin with.

And get the basic circuit operating (smoke free) before you move up to the big brother with the 2N3055s.

 

[heavyfletch2nd]

Ok steve I will leave out the big transformer for now , did you read my comments on thinking about limiting the current to the v in on lm317 and usung diods on the lm317 pins , "it was just a theory" , the pics are for modemhead , looks like you are right I pulled out a old metre hooked them up at the same time check it out lol.

 

[ModemHead]

the pics are for modemhead , looks like you are right I pulled out a old metre hooked them up at the same time check it out lol.

For ranges where the multimeter has to *output* current to test a diode or resistor, it is not a good idea to parallel connect them that way. But yeah, maybe you should measure a fresh 1.5V battery and see if that multimeter is not way out of calibration or something.

 

[heavyfletch2nd]

Modemhead I just tried that, I get the same reading off both meters 1.26v off a 1.2v rechargable aa battery , but it seems way out of cal when testing a single diod on the diod setting . Uh?

 

[ModemHead]

It's predictable that two different multimeters will read diode junctions *slightly* differently since the voltage depends on the current, and different multimeters use different amounts of test current. Most DMMs are in the 0.5mA to 1.0mA range though, and give similar numbers.

In any case, before wasting time on a red herring, let's chalk it up to cheap multimeters (no offense intended) since it sounds like the basic DC cal is close enough, and the transistors are not base-emitter shorted.

 

[ModemHead]

Maybe this will help your further understanding. Math is involved, but it is very simple:

Voltage = Current times resistance, ie. V = I * R. All but the most pedantic people will refer to this as "Ohm's Law".

With a little algebra, this equation can also be arranged as:

I = V / R
-or-
R = V / I

This means if you know any two of the variables (voltage, current, resistance) you can calculate and predict the remaining one. Science rules.

----------------------

OK, now let's consider the LM317 design goals, there are two main ones:

1) Vary the amount of current allowed to pass from Vin (pin 3) to Vout (pin 2) in order to maintain exactly 1.25V between the Vout and ADJ (pin 1). If the voltage is too low, allow more current, if the voltage is too high, pinch off some current.

2) Allow the least possible amount of current in or out of the ADJ terminal. So little in fact, as to allow us to disregard it completely in our calculations.

----------------------

Now look at the basic circuit for the device.

R1 is your 270 ohm resistor, and R2 is your 10K pot, (technically a rheostat in this case.) Call the voltage across R1 "V1", and call the current through it "I1". Similarly with R2. Now, if the LM317 is doing it's thing, we know that V1 will be 1.25V, and the current in or out of the ADJ pin is too small to care about, so I2 is the same as I1.

Since we know R1 and V1, we can say that I1 = V1/R1 = 1.25/270 = 0.0046, or 4.6mA. And as long as that LM317 is operating within its limitations, this current stays exactly the same! It's what we call a "constant current source".

Now look at the output voltage Vout. Kirchoff tells us that it is the sum of V1 and V2, but that should be fairly intuitive I hope. We know what V1 is, it is always 1.25V, and V2 = I2 * R2. And we know what I2 is, it is always 4.6mA. So that allows us to know what the output voltage can be for any given value of R2. Consider the following conditions at various percentages of rotation for R2:

0%, R2 = 0 ohms: Vout = 1.25 + 0.0046 * 0 = 1.25V

20%, R2 = 2000 ohms: Vout = 1.25 + 0.0046 * 2000 = 10.45V

50%, R2 = 5000 ohms: Vout = 1.25 + 0.0046 * 5000 = 24.25V

100%, R2 = 10000 ohms: Vout = 1.25 + 0.0046 * 10000 = 47.25V

--------------------------

I think you might see a problem above, the voltage out seems to be independent of the voltage input. Not really, as the LM317 cannot output more than the input voltage minus something called the "dropout voltage", which is the minimum voltage across Vout and Vin that needs to be maintained in order for the device to keep doing its thing. This dropout voltage changes with temperature and loading, but at room temp with minimal load it's roughly 1.5V. One of the graphs in the data sheet describes this.

Now all this should explain at least a couple things: 1) In normal operation the pot shouldn't be overloaded because it only has about 5mA going through it, and 2) Why you reach max voltage at low settings on the pot. In fact, 10K is a bit large for your project, it probably should be more like 5K to make more of the rotation range useable.

Hope this helps. It is Sunday and I have a little time to spare.

 

[heavyfletch2nd]

Modemhead I wish I knew you posted this earlier I didn't get a notification , I think its about 6:00pm where you are so hope your still around , thanks for that I do kind of understand this math I am only a month into this tho, it is starting to sink in its just knowing when to do ÷ or × on what I want to know , what do = and * stand for so I know , I had a tidy up of the bench earlier and I found a lm317 I didnt use in a packet so I have taken this project apart , disconnected the fans , disconnected the power supply , fitted new 317 with a cap of 1uf 50v , connected a 9v pp3 battery "as steve suggested" i replaced the 270ohm 1 watt with a 220 ohm 1/2 watt , put jumper leads to connect everything up lay all out on the bench and it seems fine??? No smoke not even a tiny spark in the pot , all 4 transistors are working , it is adjusting properly with a small fan connected to the output ??!!??!! So here is the million dollar question , why will the pot and lm317 go wrong when connected to a 10v 1.5a wall adapter or the main large 16v 15a transformer ???, the circuit is obviously working correctly with low current , I ran it for half hour or so turning the pot up and down waching the fan speed and slow it is fine , I know if I connect it to either the wall adapter or transformer it will pop like every other time "9 times" , look at the pics the jumper leads are basically how it all goes back to together in the box , no shorts when its back together the only thing that will be different is the supply.