LM317 wrong output voltage

[VEGETA]

R_B = 1k5; R_T = 240R || R_T2 = 620R. V_O = 9V if switch broken; might be on the lower side of I_L(min), otherwise use 750R, 120R, 300R
R_T = 150R; R_B = 1k3 || R_B2 = 3k3. V_O = 12V if switch broken

so you suggest in the first line to switch the top resistor in parallel instead of bottom one. values of 750 ohms for bottom, 120 ohms for top when 9v and in parallel with 324 ohm when 12v seem fine and uses less resistors than my previous suggestion. very low values ensure Iadj won't do much.

plus it will have nice load current even without using external load... 1.25/87ohm = about 14.3 mA, 1.25/120 = 10.4 mA. adding another 750ohm resistor as a constant load will give 12v/750R = 16 mA or 9v/750R = 12 mA. both will be more than enough to make it stable (total will be 30.3 mA for 12v, 22.4mA for 9v).



will see today if i have such low values in my storage to test them.

and sorting out the power selector ic with tps2117 would seem to fix all things.


EDIT: I didn't have enough time tonight, so I will try later. On another side of the story, I tested the buck regulators and they performed perfectly fine. about 5mV total p-p ripple figure when fully loaded without noise or spikes or bad transients... can't really complain! 

[Zero999]

R11 needs to be 120 ohms. Period.

The LM317 will work with higher resistor values for R1, but one has to beware of the minium load current and I_ADJ also starts to become significant for higher values of R2, which are also required.

[VEGETA]

R11 needs to be 120 ohms. Period.

The LM317 will work with higher resistor values for R1, but one has to beware of the minium load current and I_ADJ also starts to become significant for higher values of R2, which are also required.

using higher value resistors is the cause of what i had, will try to use smaller value when i have time and report back.

you are correct, it can work with high value resistors but you need to compensate for Iadj... this is what i learned from this thread despite using lm317 for many years. 

[VEGETA]

I tried lifting the leg of the smt lm317 in that very tight space on the board but failed and broke the leg.

however, i got another through hole lm317t from ST which i got years ago, soldered it in prototype board as seen in attachments... and it worked exactly as intended.

top resistor = 120 ohms
bottom resistor = 748 ohms (combined 680 + 68)
parallel switchable top resistor = 300 ohms (combined 150 + 150)

of course they are not precisely that but close enough to demonstrate the idea.

just to be sure I soldered a fixed load (around 220 to 280 ohms, don't remember) just to ensure enough load.

i got 12v and 9v (with tolerances) as expected.

so now I can consider this solved.

next step is to build another prototype with this corrected circuit + the power switching circuit between 5v and 3.3v based on TPS2117.

NOTE: I attached 2 pictures of the product prototype I was talking about, you can see the smt lm317 there.

[Zero999]

I'm glad you've got it working.

If you don't need the full 1.5A, then there's the LM317L (100mA, TO-92 available) and LM317M (500mA, SMT only).

[VEGETA]

I'm glad you've got it working.

If you don't need the full 1.5A, then there's the LM317L (100mA, TO-92 available) and LM317M (500mA, SMT only).

yes thanks to those who helped. It really ticks me off that I didn't know this small info about lm317 despite working with it for years.

as for needed amps it won't be much since the only usage it will get is to deliver 9v or 12v to some TVs when using RGB SCART signal since it is a power supply for retro console. so this 9v or 12v when the target tv (crt tv) gets it then it knows you hooked an RGB SCART cable and will automatically switch to it (mostly in Europe only). I didn't measure this but certainly won't be anything much of couple of milliamps or tens of milliamps at most.

this is the part I got: https://www.lcsc.com/product-detail/Linear-Voltage-Regulators-LDO_Texas-Instruments-LM317DCYR_C51466.html

it is sot-223 and as i said won't have any significant load.

_______

on the other hand, I made a new load switch circuit in the attachments.

it uses tps2117 with a switch, TI support told me it should work without a problem but i asked about "priority mode" vs "manual mode" and he said he will get back to me later. won't affect circuit operation though.