Do I need a Pass Transistor?

[Overclocked]

I am using a LM317T in constant Voltage mode (well adjustable- but constant). Data Sheet says it can supply a minimum of 1.5A, and can supply a maximum of 3.4A (typical rating is 2.2A). I might not be pulling that much current now, but there is always the option of having it there. The transformer is rated for 12.25V RMS and can supply 4A. 3.4A is fine with me, but would it be a good practice to still use a pass transistor to supply the remaining current?

The way I figure it, using the right resistor, I can have the Vreg regulate 2A of current and have the pass transistor pass the other 2 A to share the load equally.

Also, What value of capacitance per amp do I need for a linear PSU? Ive heard Values of 2,000uF per amp All the way up to 80,000 uF Per amp (this was from a article in Nuts and volts)! 80,000 uf Is Way too much, but 2,000uF to 3,000uF Per Amp seems just fine. True or not?

 

[audioguru]

Look again at the datasheet. Some are better than others:
1) If there is no more than 15V across it, some LM317 regulators can supply 1.5A, some can supply 2.2A and some can supply 3.4A. Which one do you have??
2) If there is 40V across it, some LM317 regulators cannot supply more than only 150mA. Others can supply more.
The LM317 reduces its max output current depending on how many volts are across it, to protect itself.

Have you looked at heatsinks for it? It can't dissipate more than about only 18W when it has a huge heatsink.

If you add a pass transistor then a second transistor is needed for current-limiting. Or the pass transistor will smoke and burn if the output is shorted.

The main filter capacitor's value in a full-wave rectified supply depends on how much ripple you can tolerate. Here is the graph.

 

[Overclocked]

Look again at the datasheet. Some are better than others:
1) If there is no more than 15V across it, some LM317 regulators can supply 1.5A, some can supply 2.2A and some can supply 3.4A. Which one do you have??
2) If there is 40V across it, some LM317 regulators cannot supply more than only 150mA. Others can supply more.
The LM317 reduces its max output current depending on how many volts are across it, to protect itself.

Have you looked at heatsinks for it? It can't dissipate more than about only 18W when it has a huge heatsink.

If you add a pass transistor then a second transistor is needed for current-limiting. Or the pass transistor will smoke and burn if the output is shorted.

The main filter capacitor's value in a full-wave rectified supply depends on how much ripple you can tolerate. Here is the graph.

LM317T in a TO-220 Package and its heatsinked.

If Im looking at the right section...

Current Limit (Vin-Vout) <= 15V K,T,S Packages 1.5A (min) 2.2A (typ) 3.4A (max)

So with that said, If my xformer can is giving 12.25vRMS (or somewhere around there...) I will be getting 17.33V before the rectifier. Assuming the rectifier has a 1.4V voltage drop, then the max the regulator will see (after the capacitors) is 15.93V. So If I have the reg set at max voltage, it will supply 14.68V (15.93-1.25), and the difference would be 1.25, thats Under 15V. At min voltage (1.25V is the lowest it can go- which is fine) the difference is 14.68V Still under 15V.

So at all expected Voltages, it should still provide more than 1.5A. I probably wont need more than 1.5A, but the option is always there.

 

[audioguru]

The peak voltage from the transformer is 17.33 as you say. But when loaded the bridge rectifier is working very hard and will have a loss of 2V.
Then the LM317 has a typical dropout voltage of 2.25v at 1.5A and it could be more.
So the max output voltage is only 13V at up to 1.5A and less at higher current.

 

[Overclocked]

The peak voltage from the transformer is 17.33 as you say. But when loaded the bridge rectifier is working very hard and will have a loss of 2V.
Then the LM317 has a typical dropout voltage of 2.25v at 1.5A and it could be more.
So the max output voltage is only 13V at up to 1.5A and less at higher current.

Alright thanks! So I'll design for the transistor to start conducting after 1.5 A.

 

[audioguru]

A second transistor is needed for current-limiting so that the pass transistor or the transformer doesn't smoke if the output is shorted.

 

[Overclocked]

A second transistor is needed for current-limiting so that the pass transistor or the transformer doesn't smoke if the output is shorted.

Yea, I have that circuit all set also. The only problem is the resistor required current limiting has a very low ohms value, and is in series with the load, so it passes all the current.

Cant a Fuse be used instead?

 

[mneary]

A fuse can certainly be used, but the resistor (properly rated) will recover after everything cools down. Fuses have to be replaced.

Since no schematic has been posted, we're assuming the typical current boost configuration....

 

[Hero999]

Also there's the question of ripple and mains voltage tolerance.

What are you using the power supply for?

My bipolar power supply I built for experimenting with has a centre tapped 15VAC transformer and outputs up to 15.7V but it probably won't supply 1.5A at this voltage if the mains supply is low without some ripple on the output, although I've. However this doesn't bother me as I don't intend to pull huge currents at this voltage, the full current beyond 13.8V.