LM317/LT337 questions

[Mike]

I have an old broken variable power supply that I want to rebuild and get working again. It has a transformer that outputs 84VCT (according to my measurements) and I'm having troubles finding regulators that can handle the voltage (because DC voltage will be even higher). I have national semi LM317s and linear tech LT337 (thats what I have, if i need to use the same brand for some reason I can), and they can output up to 37VDC and have a maximum of 40V in/out difference. Can I run them off of this transformer?<p>Thanks!

[Chris Smith]

Actually the 37 volt range is only applied to the difference between the in and out. <p>You can take 100 volts and drop it by 37 volts just fine. <p>Back in the 80s I came across another version of the 317 [don’t know the number] that would control up to 100 volts or more in the same TO package. I was using a 69 volt fluro battery to run a laser at 12 volts, and this same package design, could do the trick.<p> Im sure they still make them? [National semi]<p>You can also daisy chain the 317 one step at a time to lower such high voltages. <p>You start off with say 100 volts, and run a loop in the first part of the chain to drop it 37 volts, then another loop drops it another 37 volts, and so on. <p>There is a loss at each stage, but if power is not a issue?<p>[ April 02, 2005: Message edited by: Chris Smith ]</p>

[rshayes]

The high voltage version of the LM317 is the LM317HV. The data sheet is available on the National Semiconductor web site (www.national.com). Search for LM317HV.<p>This part can withstand a 60 volt differential between input and output. This is right on the edge if you use an 84VCT transformer with a full-wave rectifier.<p>Higher voltage transistors are available. A discrete regulator may be worth considering.

[Mike]

Thanks.<p>I'd like to use what I have, rather than buying more. So I can use the 317/337 no problem with this voltage, but just may not be able to get down to 1.5VDC?<p>Isn't the 1.5vdc to 37vdc the output voltage, not the amount it can change the voltage?

[Chris Smith]

No, it’s the difference.<p>A floating voltage, cant tell the difference.<p>Same reason a HV wire man can ground him self to a 1/4 million volt line up a pole, and not be affected. Its the potential between point A and B that is important. <p>You can make a loop voltage circuit, one where it goes through the regulator and onto a small load [resistor] to ground to keep it open or working. [Caps included] <p>Then you tap off the out put before the load resistor with another loop and onto ground, [a load resistor] and so on. <p>As long as the "Difference" doesn’t exceed the specs of 37 volts, its fine. <p>Also, each one can be "adjusted" to within the specs on their own.

[Enzo]

Look at the IC, there is no ground pin on it. Whatever the input voltage, you can vary the output voltage down to 35v less than the input. If you start with 35v, then you can dial it down to 1V. If you start with 85V, then you can dial it down to 50v.

[viveguy]

You don't mention the rectifier configuration that you are using. You are going to have a voltage multiplication of 1.414 for a full wave bridge. You can get around the voltage differential problem by putting a zener on the reference (ground) pin. But, that will raise your minimum output voltage.<p>I suggest that you go to www.national.com for application information.

[Mike]

I am using a standard full wave rectifier. So, if I have 42VAC I will have 59.388VDC?? So I would only be able to get it down to 22VDC max?<p>IF I were to use two chips, one set permanently to bring the voltage down to 37VDC and another hooked up to the potentiometer to allow adjustable voltage anywhere from 0 to 35.8VDC, how much current would I end up losing? The original power supply delivered up to 1.5A, could I get that much, or would I have to get the HV regulator. Only problem is I don't see a negative version, (unless I just missed it) and the transformer is not dual secondary, it has a center tap, so I couldn't use two positive regulators.

[Dean Huster]

I may have missed, but there is one point about high input voltages to those regulators and that's the power that'll be dissipated by the regulator when you start pulling lots of current. Always subtract the output voltage of a regulator from the input voltage and multiply by the highest expected current that you'll draw to see how much power will be dissipated by the chip. Although you may be within spec for voltage, the regulator may not be able to handle the power dissipation you're slapping to it.<p>Dean

[Chris Smith]

They also make a negative voltage adjustable regulator.<p>The LM 350K [+] model holds 3 amps at 33 volts.<p>Negative Adjustable VoltageRegulators....
1.2V to 32V NTE 935 TO3 3 35W 5...50
1.2V to 33V NTE 970 TO3 3 35...3 Internally Limited
1.2V to 33V NTE 1929 TO220 5 35...3 Internally Limited
1.2V to 37V NTE 956 TO220 3 40 1.5...15
1.2V to 37V NTE 1900 TO92 3 43 0.1 0.625
2.85V to 36V NTE 1942 TO220 8 40 2 Internally Limited
4.5V to 40V NTE 1928 8-Lead Metal Can 8.5 40 0.045 0.8
5V to 37V NTE 1930 8-Lead DIP 9 40 0.025 0.450
5V to 30V NTE 953 4-Lead Power Tab 7.5 40 1 15<p>[ April 03, 2005: Message edited by: Chris Smith ]</p>

[MrAl]

Hi there,<p>Are you sure you want to use a linear regulator?
If you've got 60vdc coming in and 5vdc going
out at 1 amp, the linear will be dropping 55v
and at 1 amp that's 55 watts of power to get
rid of with a humongous heat sink.<p>Consider using a switching regulator.
National makes a line of these called
"Simple Switchers" that dont take too many
more parts then a linear like you're talking
about. Check out LM2576 and similar.<p>On another web site we developed a neat ps
using a tracking switcher ahead of a linear
to reduce wasted power with lower output
voltages if your interested. I have to say
though that it's not the simplest thing to
build.<p>What kind of specs are you looking for...or...
what do you want to be able to test with the
finished ps?<p>Take care,
Al

[sofaspud]

No matter how it's configured, I'm pretty sure the max power dissipation of the LM317T is 20W. I mention this because at the low end of the variable voltage range the part cannot provide the full 1.5A output. Says so in the datasheet. I don't know what your planned use for the ps is, but I think I would seriously consider making 2 supplies - one with the approx. 30V - 50V output and another with a 1.2V - 30V output.