needed 317/337 double regulation

[jaudio]

I need a schematics for LM317/337 double regulated power supply

Thanks in advance

 

[Eva]

Have you checked the datasheets from National Semiconductor? They contain plenty of useful application examples.

 

[jaudio]

Have you checked the datasheets from National Semiconductor? They contain plenty of useful application examples.

Yes I have it in hand. I guess I was kinda looking for tips.

This is what I have on paper

24,000uf+.1uf--[317]--.1uf+12,000uf___ .33ohm ___ 3300uf+.1uf--[317]--.1uf+1000uf___ .33ohm.___

each 317 has a pot(100uf cap in parallel) at adj pin to ground

 

[Eva]

With such amounts of capacitance, remember to add diodes to protect the regulators from reverse biasing as the datasheet shows

 

[AndrewT]

Hi,

24,000uf+.1uf--[317]--.1uf+12,000uf___ .33ohm ___ 3300uf+.1uf--[317]--.1uf+1000uf___ .33ohm.___

That last 0.33ohm is going to ruin your regulation. Is that what you need?

 

[jaudio]

Hi AndrewT

I definitely don’t want to ruin the regulation.

I have a power supply in a separate chassis and need to get 40volt down to 18volt and only using 317/337. I will use 3 regulators per rail to get the voltage to 18volts. The 18volt will feed a preamp with 48,000uf per rail. I didn’t think the last .33ohm resistor would be a problem. Should I leave it or remove it?

Thanks

 

[richie00boy]

Sorry but that is a very inelegant solution. Why not use a single regulator, or simple emitter follower fed from zener reference?

 

[Eva]

It's quite obvious that he is looking for something exotic while he has little or no knowledge on how to use a LM317/LM337 (two things that usually happen together, as the more electronics you understand, the more you laugh at exotic).

If he just wanted something functional, he would calculate the required capacitances depending on load current and on the maximum allowed input voltage ripple in order to obtain the desired output ripple (and the resulting values would be much smaller than the ones he mentioned).

I understand exotic trends quite well, we as humans are very prone to fill the knowledge gap with imagination when we don't have anything better to use.

 

[richie00boy]

I was trying to put it in a less upfront way.

 

[jaudio]

Good, you know more about electronics.

richie00boy
Here’s the thing. I need to build the power supply with the parts I have but Thanks

Eva
I don’t need the exotic, just the functional. I planed to do the capacitance calculations but I have the equation on paper somewhere. This is not my only hobby but it is the only one, as much as I love it, which is not going to put money in my pockets, so tend not to keep formulas and equations in my head.

Your first post was helpful, thank you. I really don’t know how to take your second.

 

[Eva]

How much current are you going to draw from the regulators and which is the estimated input voltage?

 

[richie00boy]

He seems to have 40V input, and feeding a pre-amp I doubt he will need much current. There is absolutely no need to use 3 regulators. You can make do with one if you are prepared to buy a zener diode.

I would not use such massive capacitance because you will suffer in the high frequency domain because of it and you will not gain any worthwhile improvement in lower end ripple etc even going down to 10x smaller caps.

As Eva has already pointed out you WILL need a reverse voltage protection diode.

 

[peranders]

The 18volt will feed a preamp with 48,000uf per rail.

That's more than a high power power amp. How you got any specification of what you are going to achieve? It looks like massive overkill.

 

[RetroAudio]

where's the fun??

For me the fun in this hobby not only comes from doing stuff but learning what i'm doing and why probably accounts for most. If you're into overkill ask yourself what you're trying to kill.

 

[peranders]

It can be practical to know what is "normal" so you'll know when you are up to overkill.

In this case you have to remember the high frequency properties also and 48000 uF will create good such.

 

[Eva]

LM317 datasheet states a typical 80dB figure for 50/60Hz ripple rejection when using a 10uF capacitor between Adj. pin and GND.

That means that 1V p-p of intput ripple will produce only 100uV p-p of output ripple. That seems enough for me, altough it may be not for you.

On the other hand, to calculate input capacitor ripple due to 50/60Hz rectification, we can just assume that the capacitor is sourcing all the current during 7ms out of the 10ms of each 50Hz pulse.

Then, for a 100mA current consumption (assuming it's a preamp and not a power plant) and a bare 680uF capacitor we would have:

V = I * t / C = .1 * .007 / .00068 = 1.03V p-p input ripple (and thus 100uV p-p output ripple)

So a 680uF input capacitor, a 10uF adjust pin capacitor, and a 10uF output capacitor as the datasheet recommends would suffice to make a functional regulator with a relatively quiet output.

 

[jaudio]

This is the finished power supply. It is capacitance is 1kuf__resistor__12kuf+.1uf+.01uf/lm317/3.4kuf+.1uf snubber----1kuf+.1uf+.01uf/lm317/.1kuf+.01+.1kuf_snubber
and 100uf on adj pin.

I have jump point at bridge and bewteen the lm317 sets,so the supply is flexable.

Thanks to all who helped ,especially juma and rudi

 

[jaudio]

again