Hoffman Amplifiers Tube Amplifier Forum

Other Stuff => Other Topics => Topic started by: shooter on April 22, 2018, 06:28:00 pm

Title: LM317T math
Post by: shooter on April 22, 2018, 06:28:00 pm

I'm trying to model this formula in excel, but I'm unsure of the last part of the formula

for context, here's the link;

http://www.circuitstoday.com/few-lm317-voltage-regulator-circuits

the formula;
(I adj * R2)  doe's it Mean current through Pin 1, (adj), I get the R2 part  :laugh:
If it is current, wouldn't that vary, then varying Vout  :dontknow:

Title: Re: LM317T math
Post by: 2deaf on April 22, 2018, 10:23:13 pm

One of the things that may be confusing you is that Circuits Today got the equation wrong.

I_ADJ is an unwanted and annoying current through Pin 1 that fools with the voltage drop across the variable resistor and, therefore, the output voltage.  Fortunately, they have reduced that current to 50uA typically with a very small variation across the operating range so that you can practically ignore it.

Title: Re: LM317T math
Post by: PRR on April 23, 2018, 12:56:18 am

There's a minimum load current to keep the '317 alive; it can be as much as 10mA.

It is *very convenient* to set the R1 R2 network to suck that 10mA so it gives the right output even when the load is absent.

The reference is 1.25V. 1.25V/10mA is 125 Ohms. (120 is about as good, and a Standard Value.)

Now you do NOT have to think about any silly 50uA stray current at the Adj pin. Less than 2% error when Vref is +/-4% and there's little justification for <5% resistors.

1.25V (1 + (R2/R1)) is the answer.

-OR- (Vout-1.25V) is the voltage across R2. Figure 10mA in that, done.

Title: Re: LM317T math
Post by: 2deaf on April 23, 2018, 09:53:28 am

The correct equation for the diagram shown in Circuits Today is:  1.25V(1+(R1/R2)) + I_ADJ(R1)

Title: Re: LM317T math
Post by: shooter on April 23, 2018, 10:08:51 am

Thank you Sir's, I did finally find "normally ignored" on the data sheet.

I want to use fixed R values, so just plug in #'s till I get 30Vout.

one value I didn't find Max Vin, found differential in/out, max out, I'll look again with fresh eyes

Title: Re: LM317T math
Post by: sluckey on April 23, 2018, 10:27:01 am

It seems reasonable to me that max V_in would be max V_out plus differential in/out.

Title: Re: LM317T math
Post by: PRR on April 23, 2018, 11:08:45 am

> It seems reasonable to me that max Vin would be max Vout plus differential in/out.

Yes; but note that at Turn-On, Vout is *zero* until the output cap charges. This hidden hurt has led a lot of folks into trouble.

Title: Re: LM317T math
Post by: shooter on April 23, 2018, 04:48:57 pm

Quote

This hidden hurt has led a lot of folks into trouble.

my middle name  :icon_biggrin:

the input is ~40ish bridge rec, filtered DC, the output "load", no power, is ~ 3k ohmed to ground, so I think I'm good to go once I get R1/R2 sorted.  I'm just making a temp supply til I can sort out the original.  I'll fire it up with a DL before I introduce it to the system, just to be sure.

again, thanks for the help

dave

Title: Re: LM317T math
Post by: shooter on April 23, 2018, 05:17:27 pm

here's my calculator fwiw  :laugh:

Title: Re: LM317T math
Post by: 2deaf on April 23, 2018, 06:10:09 pm

I can't open that file, but did you get 2.747K with a 120r reference resistor or 5.468K with a 240r reference resistor?

Title: Re: LM317T math
Post by: shooter on April 23, 2018, 07:24:04 pm

It's ancient XL  :icon_biggrin:

um, no, I got 2.2k / 120 = 29vdc, went with stock r's, and 29 is close enough to 30 for this.

there's my math;
V0 = Vref * (1 + R2 / R1)
I used 1.5 for vref  :think1: :cussing:
OK, ya, that's what I got  :icon_biggrin:

thanks for the debug  :laugh:

Title: Re: LM317T math
Post by: shooter on April 29, 2018, 07:06:27 pm

small progress with confusion;

29.7 is good, my spreadsheet doesn't think so   :dontknow:

I just did a quick unloaded test, the original PS was 30vdc .2A
going back n measuring pin 1, and try to lightly load with a 2-3K r and see

EDIT: updated schematic with CORRECT values, getting closer with the spreadsheet
put it under small load and everything was happy so time to hookup original loads n see