Alternative Transistors for LM317 High Current Voltage Regulator Circuit

[Prithwiraj Bose]

Hi,
I am referring the below circuit diagram to make a High Current Linear Voltage Regulator circuit. This circuit is given in LM317's datasheet.

I need the regulator circuit to handle no more than 5 AMP Output Current. I think TIP73 a little overestimate for this current. Moreover, I do not have this transistor handy. For 5 AMP output, what would be the alternatives for TIP73 and 2N2905 transistors in this circuit?
Can I replace them with 2N3055/TIP3055 and TIP42/MJE2955 respectively? Would 2N3055/TIP3055 cause any interference to the LM317 to regulate the output voltage?

 

[Alec_t]

A good way to search for alternative transistors is to use the parametric search facility on the web-sites of major distributors such as Digikey, Farnell, etc.

 

[73's de Edd]

Sir Prithwiraj Bose . . . . . . . . . . ( Did 'ya know Amar ?)

After looking at Tejas Instrumentations specs sheet and that new numbers . . . still quite bucolic alpha spec . . .such as would be permissible with a linear pass design supply.
It's using a smaller TO-220 casing, which your '3055's metal or TO-3 plastic casing would be improving upon.
I see no immediate fault in using the "old, true and time tested industry standard" 2N3055 and matching the added count to your power transformers current capability .
For any additional units, balance their matching with inserted, individual series emitter resistors.

Referencing:




73's de Edd

 

[Audioguru]

He wants only 5A, not 45A. A single power transistor will be fine for only 5A.

 

[duke37]

He wants only 5A, not 45A. A single power transistor will be fine for only 5A.

It depends on the supplied voltage, it is necesary to keep well inside the safe operating area. More pass transistors there are will increase reliabilty until there are too many. Aircraft have gone from eight engines down to two to increase reliability and cost.

 

[Prithwiraj Bose]

Thank you all for your responses. I have got it working fine with a single 2N3055 for 5 amp. I have written a blog post to to share a working circuit and PCB layout of LM317 power supply

 

[Audioguru]

The datasheet for a 2N3055 shows that at 4A its minimum current gain is 20 and is less if its Vce is less than 4V. With a current gain of 15 then with a C to E current of 5A its base current must be at least 5A/15= 333mA. Then the voltage feeding its 470 ohm base resistor must not be less than 157V! That is why the high current LM317 circuits in the datasheets show output transistors (LM195) with extrermely high current gain.

Your circuit works only with your luck of getting a 2N3055 transistor with current gain much higher than most, a high input voltage and a low output voltage in addition to testing to only 2.5A.
You should design circuits with worst case spec's so that everycircuit works properly.

 

[Prithwiraj Bose]

The datasheet for a 2N3055 shows that at 4A its minimum current gain is 20 and is less if its Vce is less than 4V. With a current gain of 15 then with a C to E current of 5A its base current must be at least 5A/15= 333mA. Then the voltage feeding its 470 ohm base resistor must not be less than 157V! That is why the high current LM317 circuits in the datasheets show output transistors (LM195) with extrermely high current gain.

Your circuit works only with your luck of getting a 2N3055 transistor with current gain much higher than most, a high input voltage and a low output voltage in addition to testing to only 2.5A.
You should design circuits with worst case spec's so that everycircuit works properly.

Thanks for pointing out the theory. I would definitely like to test the scenario with practical input voltage and 3V test output and try to draw 3+ Amps from the circuit. It is definitely worth learning. Thanks for pointing out the need to use a high gain transistor. So the transistor must have high gain at low voltage and at the same time high Current limit tolerance, right?

 

[Chemelec]

Whichever Transistor you use, It will need a GOOD HEAT-SINK on it.

 

[Audioguru]

The problem is worse when the output voltage is close to the input voltage because then the 470 ohm base resistor has a small voltage across it and a low current in it.

The datasheet of an LM317 shows LM195 output transistors that have extremely high current gain. The datasheet does not show a very old TIP73 ordinary transistor with very low current gain of about 20 times.
The current gain of an LM195 is 110000 times. It is an entire amplifier.

 

[(*steve*)]

The problem is worse when the output voltage is close to the input voltage because then the 470 ohm base resistor has a small voltage across it and a low current in it.

On his web site, it is not a 470Ω resistor, it is a 4k7 resistor. 4R7 would be better for his R1.

And I have no idea what his R7 is supposed to do. If it is too limit the maximum voltage, or to compensate for the lack of availability of the correct value pot, then it will do this, but in a way that essentially renders half of the travel of the pot useless. Another thought is that the pot has an audio taper, and that this is an afternoon to linearise it.

Also, for a 5A output, the transformer should be rated for at least 7A.

Edit: his PCB has traces all the same size. If any of these are carrying his 5A, it would make sense to have wider traces.

I suspect it has not been tested at full power (max current, max voltage) or that the circuit on his blog is different from what he tested.

 

[TedA]

I have written a blog post to to share a working circuit and PCB layout of LM317 power supply

Here's the schematic from that post:

The TIP2955 / 2N3055 transistor combination is reasonable to use. This PNP is larger than required, but costs little more than a smaller part.

Agreed that 4.7k is a bit large for R1. A much lower value would allow more output current at higher output voltages. You do want some resistance here to limit the fault base current in the PNP. 4.7 ohms is likely enough for that. 10 to 100 ohms might be OK.

Also agree that a 5A AC rated transformer winding will be beyond its rating with 5A DC load. The value of C1 is also a bit small for a 5A current. 10,000uF would be a more usual value. C1 requires an adequate ripple current rating. There is a tradeoff between ripple voltage and ripple current. A larger C1 value reduces the ripple voltage, but increases the ripple current through the transformer, rectifier and C1 itself.

Even at only 5A output, the circuit might benefit from paralleled output transistors. This makes the heatsinking easier.

Ted

 

[icesteve]

May I ask one question here please, which may also be of interest to others on this subject: does this current-boost/bypass arrangement in any way compromise the short circuit capability of the LM317?

Thanks,
Steve

 

[Audioguru]

May I ask one question here please, which may also be of interest to others on this subject: does this current-boost/bypass arrangement in any way compromise the short circuit capability of the LM317?

Yes, this simple circuit does not detect and does not control excessive current. Additional parts are needed as shown on the datasheet of the LM317.

 

[duke37]

How does it do this? The current limit of the LM317 is set internally. The pass transistors are at risk though.

 

[Audioguru]

How does it do this? The current limit of the LM317 is set internally. The pass transistors are at risk though.

Like any power supply that has current limiting or current regulation, it uses a low value resistor in series with the voltage pass transistor to sense the current then it amplifies the resulting voltage and compares it to a reference voltage. The comparator drives the ADJ pin on the LM317 so that the output voltage is reduced enough to limit or to regulate the output current in the load.

 

[duke37]

I agree but the circuit shown in #10 has no current sense but there is no risk to the LM317 which has its own current limit.so shorting the output does not compromise the short circuit capability of the LM317.

A circuit as described by Audioguru (#16) would be advised to protect the pass tranistors and the voltage source.