» Electronics »Linear voltage regulator with regulation on LM317 and PNP transistor

Linear voltage regulator with adjustable LM317 and PNP transistor

Hello!
In this article I will talk about another linear voltage regulator, which I assembled relatively recently. It is built on the popular LM317 chip and a bipolar PNP transistor. The finished module is as follows:

Related video:


In the past article I talked about a similar linear voltage regulator on TL431 and NPN transistors.

This circuit, in contrast to the aforementioned, contains slightly fewer parts, and is able to withstand higher currents, thanks to a more powerful transistor.

Main characteristics:
• Input voltage up to 30V (in my version, because the capacitor at the input to 35V)
• Output voltage 3-25V (depending on the current, the higher the current, the lower the maximum output voltage)
• Current up to 9A (with a TIP36C transistor with an input voltage of 18V and an output of 12V, but generally depends on the selected transistor and power dissipation)
• Stabilization of the output voltage when changing the input
• Stabilization of the output voltage when the load current changes
• Lack of protection against short circuit
• Lack of current protection

The module is assembled as follows:


Explanations according to the scheme:
The LM317 microcircuit purchased on AliExpress (most likely not the original one) has 3 outputs. The findings are indicated in the diagram and the picture in the lower right corner.

The chip controls a powerful bipolar PNP transistor VT1. I used TIP36C for this purpose. The main characteristics of the transistor: voltage - 100V, collector current - 25A (in fact, 8-9A, because the transistor is not original and was bought by Ali Express), a static current transfer coefficient of 10.

It is very important to monitor the power dissipated by the transistor so that it does not exceed 50-55 watts (for a transistor in a TO-247 package or similar in size, and for transistors in a TO-220 case - no more than 25-30 Watts). You can calculate by the formula:

P = (U output -U input) * I collector

For example, the input voltage is 18 V, we set the output voltage to 12 V, the current we have is 9 A:
P = (18V-12V) * 9A = 54 Watts

Resistors R1, R2, R3 set the voltage that our circuit will stabilize. Resistor R1 is taken as standard at 240 ohms (any power). Resistor R2 is variable, it is better to take in the region of 2-3k ohms. Initially, I set it to 4.7k Ohm, as a result, somewhere in the middle of the knob's rotation range, the voltage reaches its maximum value and does not change further.I soldered a 3.9k Ohm resistor parallel to the potentiometer, the adjustment became smoother and the entire range of knob rotation began to be used. Resistor R3 is optional, serves to slightly move the lower and upper boundaries of the adjustment range towards the increase. General rule: the greater the total resistance of resistors R2 and R3, the higher the output voltage. This is confirmed by the formula from Datashita:
Linear voltage regulator with adjustable LM317 and PNP transistor

Resistor R4 is used to slightly limit the current to the input of the LM317 chip. Resistance 10 Ohm. LM317 as much as possible can pass through itself about 1A (up to 1.5A, if the original). At first glance, the power of the resistor R4 should be:

P = I ^ 2 * R = 1 * 1 * 10 = 10 Watts

But since the current also passes through the base of the transistor VT1, bypassing the resistor, you can take the resistor R4 and 5 watts.

The above components form the core of the circuit; everything else is additional elements to improve stability and provide some protections.

Capacitor C2 (ceramic 1-10 microfarads) - is soldered in parallel with a variable resistor and improves stability of regulation. To protect the LM317 microcircuit when the capacitor C2 is discharged, a D2 diode is placed. They, together with the D1 diode, protect the microcircuit and the transistor from reverse current. Diode D3 serves to protect the circuit from EMF self-induction when powered by electric motors. Capacitors C4 (electrolytic 35V 470-1000 uF) and C5 (ceramic 1-10 uF) form an input filter, and capacitors C1 (electrolytic 35V 1000-3300 uF) and C3 (ceramic 1-10 uF) form an output filter. Resistor R5 at 10k Ohm (any power) creates a small load for the stability of the circuit at idle and helps to discharge capacitors faster in case of power failure.

Build process:
At first, everything was assembled by hinged installation and tested.

Then I soldered the circuit on the breadboard in the form of a module.


Added a small radiator.

With such a radiator, the circuit can work for a long time only at low currents. In order for the circuit to work for a long time at full power, you need a more massive radiator.

LM317 and transistor can be mounted on a radiator without insulating gaskets, as According to the scheme, these conclusions (LM317 output and transistor collector) are connected.

I tested the finished module and checked the characteristics.

In general, I liked the circuit: quite simple and you can get a decent current. What is missing is protection against short-circuit and current. Well, it’s over. The efficiency is not high and it gives off a lot of heat. But this is a feature of all such linear circuits, which personally does not really bother me.

Thank you all for your attention! I hope the article was useful to you.
7.4
8
8.7

Add a comment

    • smilesmilesxaxaokdontknowyahoonea
      bossscratchfoolyesyes-yesaggressivesecret
      sorrydancedance2dance3pardonhelpdrinks
      stopfriendsgoodgoodgoodwhistleswoontongue
      smokeclappingcraydeclarederisivedon-t_mentiondownload
      heatirefullaugh1mdameetingmoskingnegative
      not_ipopcornpunishreadscarescaressearch
      tauntthank_youthisto_clueumnikacuteagree
      badbeeeblack_eyeblum3blushboastboredom
      censoredpleasantrysecret2threatenvictoryyusun_bespectacled
      shokrespektlolprevedwelcomekrutoyya_za
      ya_dobryihelperne_huliganne_othodifludbanclose
13 comments
Yes, of course, the diode is drawn on the contrary, sorry sorry
Protection against reverse polarity when connected (it was a transceiver feeder, like Kenwood 100-watt, from the onboard long-distance network). And I lied something with R1 - I put pieces of nichrome with a diameter of 1.2 mm there, but not 2 Ohms, maybe 0.2. It has long been. But for your product this is unnecessary, remove both.
Guest Sergey
I put the reed switch on the off network.
Author
And what function does the VD1 diode perform? And is he turned in that direction? In my opinion, in the version, as in the short circuit diagram, it will be ...
You can try protection in this version, I once did.
R4 wire. The diode is parallel to R7.
Author
Personally, I am completely satisfied with the stability of the circuit, including the above drawdowns. I didn’t know about the use of the built-in protection in the MC, interesting!
Quote: EandV
... an increase in the load current by 1.3 A floats +/- 50 mV. ...

This is how many percent of the output voltage, do not give a damn about such a drawdown?
About the "ring".A transistor with an OE inverts the phase, I came across this a couple of times, you are tormented to pick up capacitors.


... • Lack of protection against short circuit
• Lack of current protection ...
You can use the built-in in the MS, by selecting the resistor R4 in the region up to 1 Ohm (BE voltage is released on it) from about 1.5 A, the protection current of the MS, to the desired or acceptable value. Practically tested.
Author
Useful information, thanks!
From the foregoing, I can conclude that a fellow amateur radio artist is exaggerating, and stabilization is quite suitable for circuits of this level.
Quote: EandV
at a current of about 3A, the output voltage floats +/- 20 mV when the input changes. With a sharp increase in the load current by 1.3 A, +/- 50 mV floats.
This is a normal reaction to external influences. When the load current increases, it is unlikely that there will be "+/-", the usual reaction to increasing the load is a drawdown.
“Floating” is when the output voltage changes with constant input voltage and load.
The circuit seems to be linear, there should not be any ripples and interference.
Why's that? It is not connected in any way. Even the LM317 voltage regulator itself needs output capacitance for frequency compensation. And the entire stabilizer is essentially a transistor with an OE, in the collector circuit of which a load is included, and the LM317 is the source of its base current. The transistor is taken low-frequency, with a small h21e, so in this case there should not be big problems with stability, but this does not mean that everything will be smooth when using faster transistors.
Author
If you look in the video, at a current of about 3A, the output voltage floats +/- 20 mV when the input changes. With a sharp increase in the load current by 1.3 A, +/- 50 mV floats. All the same, I did not position the scheme as a laboratory technician, therefore, as for me, it was quite normal. If you put a larger capacitor at the output, it may be even better.
And what is ringing there? The circuit seems to be linear, there should not be any ripples and interference. Or am I confusing something?
Guest Alex
Tell me more how this circuit rings and voltage floats ...
Quote: EandV
If the output of the diode bridge is 30V, such a 35V capacitor is likely to explode.
Explode, perhaps, and will not explode, it’s like he’s lucky.)) But for sure it is degrading.
When choosing a filtering capacitor at the rectifier output, it is necessary to take into account the permissible changes in the mains voltage, the nature of the load, the temperature at the place of use, and the parameters of the capacitor itself (there are other characteristics besides the capacitance, permissible voltage, and ESR). In order not to bother with this, take a margin of voltage of 50% - and you will be happy. )))
Author
Quote: Ivan_Pokhmelev
For the input capacitor, such a margin is small. With an input voltage of 30 V, the capacitor should be at least 40 V, and preferably at 50.

I agree, with 30V I got excited. I had about 16V at the output of the diode bridge and about 21V after 2x capacitors of 10,000 uF at 35V, only after that the above module was connected.
If the output of the diode bridge is 30V, such a 35V capacitor is likely to explode.
Quote: Ivan_Pokhmelev
In this case, do not forget to isolate the radiator from the PSU case.

Definitely. There will be an output voltage on the radiator.
Input voltage up to 30V (in my version, because the capacitor at the input to 35V)
For the input capacitor, such a margin is small. With an input voltage of 30 V, the capacitor should be at least 40 V, and preferably at 50.
LM317 and transistor can be mounted on a radiator without insulating gaskets, as According to the scheme, these conclusions (LM317 output and transistor collector) are connected.
In this case, do not forget to isolate the radiator from the PSU case.

We advise you to read:

Hand it for the smartphone ...