1117 voltage regulator pins

5v to 3.3v asm voltage reg

hello all i am starting to dip into the world of pcb design and am a little stumped the voltage regulator 1117 smd chip has a "4th" pin and im not sure what its used for after googling it because the circ diagram i am using calls it "Vout" but it the schematic and online resources it reefers that the "output" pin is actually the "Vout" in the wiring shown i have a few of these laying around but am unable to test due to no smd compatible testing hardware, so in a nutshell:

what is the "4th pin/Vout" used for and am i correct in assuming that the schematic i am using has a defined the wiring slightly incorrect based on what ive written above?

thanks in advance you beautiful beautiful people

Pins 2 and 4 are exactly the same. Except any heat dissipated from the device will be conducted out on pin 4 (the tab). So you want a lot of top copper around pint 4. Maybe that is your ref box?

Your capacitor location is likely fine, however I would move the input to ground 90° and put it right at the pins 1 & 3, Then I would move the other to the right of the regulator.

Why would I do this you might ask. Good question. For high frequency reduction you want the leads as short as possible. The longer the traces to the capacitor the more the capacitor effect is at high frequencies. The digital lines of the Arduino generate high frequencies..
But as I said you are likely OK but not optimum. And with the 1117 pin 1 crossing between 2 and 4 you are limited on how large you can make the topside copper which limits how much power you can dissipate in the device.

John

1 Like

Thanks a million john,yes you are right the red trace was for reference and i will keep the capacitor reduction in mind for the future the form factor of the pcb on this one is the main focal point,rock on and thanks again

Many time designers will add a pattern of plated-through vias under the tab and will have a copper pour area on the board underside the vias will be solder pasted and will end up being solder filled so as to make a large heat sink area. All depends on how much heat the regulator will dissipate.

2 Likes

Almost like a mini vent area to help heat dispersion? I like it! Couldnt hurt to throw a few vias underneath thanks for the tip! im going to incorporate this now going forward

Ehm, is the PCB image inverted (flipped) ? Because if this is top view, then the footprint has been mirrored.
Also i would make the traces significantly bigger than they are now. These are powerleads after all, and if you need heatsink for the regulator, you should accommodate for the amount of current.

1 Like

Yes forgot to mention was inverted,what size would you recommend? Its going to be a step down from 5v 1 amp to 3v 1 amp?

For the traces i'd use 1mm or so.

Eh... even with sufficient heatsink, an LM1117 is rated at a maximum of 800mA, but that is a big heatsink.

1 Like

Quite right, it is so big that it is known as an infinite heat sink. This means one that does not change its temperature and is the same as the ambient temperature.

These of course do not exist, so that 800mA limit must be further derated, to match the real heat sink you are going to use.

1 Like

It depends a little on the voltage reduction as well.

Yes that is all part of the equations that govern the amount of power burn't off.

Curious, What devices / circuits require 1A at 3.3V? Most 3.3v applications are very low power.

1 Like

Esp07 in transceiver mesh node

I know nothing about the ESP07, However I found the below specification and will explain the process I would go through if I were designing a similar board. Remember I don't know the implications of being in "transceiver mesh node".

The spec suggests the average is ~71ma, so lets think of 100ma and see how we fair.

  1. (5V - 3.3V) * 0.1 A = 0.17Watts.

  2. RΦja is 62 °C ** for each watt, so the Average heat rise would be 62*0.17= 10 °C Rise, so not so bad.

  3. At the peak of 500ms the wattage would be: 0.85 Watts and the rise = 62*0.85= 53 °C still not so bad. And since it is possibly only a small fraction of the time.

  4. Lets look at temperature Rise. if your case is 40°C inside and the rise is 53 °C then the internal chip of the regulator would be 93°C. Well within the rating for the device.

  5. HOWEVER we have to see if the chip will current limit at a the above condition: The spec says it can output 800ma up to a junction temp of 125°C

My conclusion:
It will work with a decent amount of copper at the device both top and bottom. Personally I would like more safety margin.
I would consider a different package perhaps the TO-252. Its much bigger but will lower the junction temperature. And if you have the room you could use the TO-220 with either a board heatsink or standing up with a clip on heatsink.

All this assumes your original 1 amp is overstated. If its real then more design work needs to be done.

John

** Note the 62 °C/watt is the junction temperature rise for each watt dissipated. It is an estimate with a wide range of possibilities since whomever wrote the specification cannot know what your temperature of the board is.

ESP-07:
image

LM1117-xx
image

2 Likes

Noted thanks for the insight im going to read this a few more times to really digist all the info

Feel free to ask if something eludes you :slight_smile:

1 Like

For any ESP, i use the sot-223 package with significant heatsink, without issue to regulate down from 5v. I found that a total surface of 4 square centimeters is easily enough.

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.