Hello ; THANK YOU for posting your question here in this forum. Probably you get a better assistance if you change the post topic to "[Tips & Skill"](https://easyeda.com/forum/category/4)     :)

Hello! What are you going to do with the MOSFET? What are your requirements? Logic level only states that it can be switched with 0-5V. Need a bit more info on that one(Amps, Voltage, switching speed). I tried quickly looking into the TPS54394 datasheet. The PG pins are open drain, so similar to an open collector output, you'd need a pullup resistor. Looks like others have connected it to the VREG5 pin, which supplies 5.5V from an internal linear regulator. So I understand it as this: if the output is not "good", the internal FET will be "on" and PG pulled to PGND, and the voltage on PG pin will also be LOW. If the power is "good", the FET will be "off" and the pullup resistor will make the PG pin HIGH. I hope that is right and makes sense. Here are some tries to simulate your LED: The NPN transistor allows the LED to turn on from the 5V from VREG5 with the low current from the pullup resistor on the PG pin.  This could be how it looks in a schematic:  Have a look at how these 2 circuits were made, maybe that will help: This one is from a TI evaluation board: [https://www.ti.com/tool/TPS54394EVM-057](https://www.ti.com/tool/TPS54394EVM-057) [https://www.ti.com/lit/ug/slvu746/slvu746.pdf?ts=1626937623433&ref_url=https%3A%2F%2Fwww.ti.com%2Ftool%2FTPS54394EVM-057](https://www.ti.com/lit/ug/slvu746/slvu746.pdf?ts=1626937623433&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FTPS54394EVM-057)<br> <br> A board from a user in EEVBlog: [https://www.eevblog.com/forum/projects/switch-mode-power-supply-problems/msg577329/#msg577329](https://www.eevblog.com/forum/projects/switch-mode-power-supply-problems/msg577329/#msg577329) Although this post is about problems with his circuit. The 5V output seems to have glitches, but also his EN1 pin goes off screen unlike the 3V3 part where the EN2 does go to the VIN2 through a pullup resistor. Maybe that is his whole problem, the post is from 2014, so could be anything. Study the datasheet, maybe head over to EEVBlog to ask a question about this, there are many professional engineers and hobbyists there and will help out. Maybe someone here will also help with this. Also, don't be afraid to ask about the PCB layout when you are thinking of ordering the PCB's. There may be things you have overlooked and could use a pair of fresh eyes. Please don't use the autorouter for this PCB. And don't forget: decoupling capacitors, decoupling capacitors, decoupling capacitors, decoupling capacitors.

"Please don't use the autorouter for this PCB." Sound advice. Read and follow the layout guidelines in the TI datasheet and apps notes for the TPS54394 and the links to the evaluation module posted by deskpro256: [https://www.ti.com/tool/TPS54394EVM-057](https://www.ti.com/tool/TPS54394EVM-057) [https://www.ti.com/lit/ug/slvu746/slvu746.pdf](https://www.ti.com/lit/ug/slvu746/slvu746.pdf?ts=1626937623433&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FTPS54394EVM-057)

A little background, I'm working on a system board for a DIY CNC mill.  On this board I'd like to have a FPGA (XC3S50AN-4TQG144C) to route and interconnect the various inputs and outputs.  I'm thinking about using use two TPS54394 switchers to power the FPGA portion of the board. I'm using this design as my template: [https://github\.com/asmi84/kicad\-projects/blob/master/S7\_Min/S7\_Min\.pdf](https://github.com/asmi84/kicad-projects/blob/master/S7_Min/S7_Min.pdf)  In this design a MOSFET (DMN2020L) is used to implement a LED power good indicator.  Because I want to have my board assembled through JLCPCB Assembly Service I need to use components that are available for this service.  I was struggling to find a JLCPCB Assembly substitute part for the DMN2020L. <br>  I understand that this is quite an ambitious project and I'm prepared for numerous iterations.  I have an FPGA development board (Digilent Nexys 2 Board) and I'm using it to test what I can out of circuit. So far so good.  I already have a working system board but without the FPGA. @deskpro256, thank you for the detailed explanation of how the PG setup works.  I made some incorrect assumptions but your explanation set me straight.  This is my first time implementing a switch mode power supply for anything.  I'm basically copying the reference design for the switcher IC, including the layout.  Almost all the component selection was done with [Ti's WebBench Power Designer](https://webench.ti.com/power-designer/switching-regulator/customize/1?VinMin=5&VinMax=5&O1V=3.3&O1I=3&base_pn=TPS54394%2F2&AppType=None&Flavor=None&op_TA=30&origin=pf_panel&lang_chosen=en-US&optfactor=3&Topology=Buck&flavor=None&VoltageOption=None).  I further understand that the auto-router is a no go for this project.  Please check out my work in progress here: [https://oshwlab.com/Kajetan321/jlcpcb-assembly-spartan-3a](https://oshwlab.com/Kajetan321/jlcpcb-assembly-spartan-3a) Decoupling capacitors have not yet been implemented.<br> <br> As always, any comments or suggestions are very much appreciated. <br> <br>

Is there a reason an LED and resistor can't be connected from 5V to the PG pin? Ahh yes, the LED will light when power is not good. Wouldn't a simple solution be a PNP with a base resistor?  Mike.

@pommie, The logic sense of your suggestion is inverted. The PG pins are open drain, so similar to an open collector output, an external pullup resistor is needed to create a high level voltage on the pin. PG pin low = power not good PG pin high impedance = power is good.

Ahhh, you are of course correct. However a simple NPN will do the trick.  Or, did I miss something else? Mike.

Nice, you have plenty of work and learning ahead of you. For the MOSFET for the LED, a basic part from JLCPCB's parts library could be something like this 2N7002, JLCPCB Part #  C8545, costs $0.0247 in 1 quantity. [https://datasheet\.lcsc\.com/lcsc/1810151612\_Changjiang\-Electronics\-Tech\-CJ\-2N7002\_C8545\.pdf](https://datasheet.lcsc.com/lcsc/1810151612_Changjiang-Electronics-Tech-CJ-2N7002_C8545.pdf)<br> <br> Another thing would be to use the same MOSFET for the LED as for something else you might have on the board. Less unique parts can also lead to cheaper BOM even if the part is more expensive. You can use a more powerful transistor for this if you are using some for other power hungry things. Depends on your project. Also, the E/D pin is unconnected, is that the Enable/Disable pin, the datasheet doesn't say much about the usage, but they have used the pin 1(E/D) with a switch to GND in their test circuit, maybe it is outputting clock when E/D pin is connected to GND. I'm not sure about that 15pF capacitor, C18, they have an asterisk*, but haven't given info on that anywhere, maybe that is there for the frequency counter test lead inductance. The Kicad project you show also doesn't use it. Doing FPGA stuff can get costly quick, for 20$ a pop you better get things tested good :D Maybe even create separate cheaper PCB's for testing the circuits and the make one big boy, I don't know, hope this works well for you!

> Also, the E/D pin is unconnected, is that the Enable/Disable pin, the datasheet doesn't say much about the usage, but they have used the pin 1(E/D) with a switch to GND in their test circuit, maybe it is outputting clock when E/D pin is connected to GND. > I'm not sure about that 15pF capacitor, C18, they have an asterisk*, but haven't given info on that anywhere, maybe that is there for the frequency counter test lead inductance. The Kicad project you show also doesn't use it. You must be looking at a different datasheet than I. I don't see an E/D pin - I do see EN1 and EN2 pins whose description is in the table on page 5. I don't see a test circuit or a 15pF capacitor. Do you have a link to the datasheet you're looking at? Thanks, Mike.

@pommie, deskpro256 'a proposal using an NPN emitter followed is less power hungry for the PG pin in a low state as it allows a high resistance pullup to be used.

@Kajetan321 , deskpro256, A simple way to add a power good LED (LED on = power good) is to have a or around that value 1k pullup from PG to the VREG5 pin and a LED from the PG pin to ground. It wastes power when the PG pin is low but only a little more than when the PG pin is high. :)

Note of course that the high level voltage at the PG pin will be clamped at the forward voltage of the LED so this simple solution is not well suited for driving other logic inputs compared to using a transistor LED switching circuit.

Thanks everyone for your input.  I wanted to go with a MOSFET mainly for two reasons.  First, the working design I'm trying to adapt for my needs uses a MOSFET.  Second, I think of a BJT as a current controlled current source while a MOSFET I think of as a voltage controlled resistor. I find the voltage controlled resistor easier to conceptualize. My first reaction when looking at the data sheet of the 2N7002 was: WOW 7ohms Rds @ 5Vg.  Isn't Rds usually measured in miliohms?  Then I realized this is not a big deal since there will be a current limiting resistor in series with the MOSFET anyway to control the current through the LED and so no big deal :) I have updated my design with the MOSFET deskpro256 suggested. @deskpro256, what software did you use in the simulation?  Can you post the sim files.  I'd like to give this a try with the LED.  I understand that EasyEDA comes with simulation capability.  It would be really convenient to be able to do it from within EasyEDA.  Will be looking at more EasyEDA vids shortly. Actually now that I think of it,  perhaps simulation would be better with the TI tools, I would imagine they would have an accurate model of the switcher chip.  I wonder if I can import component model data into EasyEDA from TI. Cheers. <br> <br> <br> <br>

There is a TI PSpice model for the TPS54394: [https://www.ti.com/lit/zip/slvm613](https://www.ti.com/lit/zip/slvm613) It should run ok in LTspice as used on EasyEDA but: 1) A Spice Symbol will need to be created for it and; 2) It may found that the simulation will hit the EasyEDA run time/resources limit before the simulation runs long enough to show a useful elapsed time from start up. It may therefore be better running a simulation on a locally installed copy of LTspice or on the PSpice tools that TI supply. If however all that is required is to study a design for a LED driver from the PG pin then this simulation is adequate. [https://oshwlab.com/andyfierman/tps54394-pg-output](https://oshwlab.com/andyfierman/tps54394-pg-output)<br> <br> To learn about simulation in EasyEDA, please read the Simulation Tutorial: [https://docs.easyeda.com/en/Simulation/Headings/index.html](https://docs.easyeda.com/en/Simulation/Headings/index.html)

@pommie > You must be looking at a different datasheet than I. I don't see an E/D pin - I do see EN1 and EN2 pins whose description is in the table on page 5. > I don't see a test circuit or a 15pF capacitor. > Do you have a link to the datasheet you're looking at? I was talking about the crystal, didn't realize I hadn't mentioned that, I guess it was only in my head, sorry. :D But I had opened the project in EasyEDA, and just looked at the crystal and checked the part number, the E/D pin interested me, because usually the 4th pin is unconnected or just another GND. After clicking the "view datasheet" button it lead to a part not found page, but after checking the part number in JLCPCB, a similar crystal was found. [https://jlcpcb.com/parts/componentSearch?isSearch=true&searchTxt=C112970](https://jlcpcb.com/parts/componentSearch?isSearch=true&searchTxt=C112970)  Here is the datasheet: [https://datasheet\.lcsc\.com/lcsc/1809021111\_Yangxing\-Tech\-O705050MEEA4SC\-50MHZ\-25PPM\_C112970\.pdf](https://datasheet.lcsc.com/lcsc/1809021111_Yangxing-Tech-O705050MEEA4SC-50MHZ-25PPM_C112970.pdf)<br> <br> And the part about the test circuit in page 2:  @Kajetan321 Here is where I have a quick go at simulations: [https://www.falstad.com/circuit/circuitjs.html](https://www.falstad.com/circuit/circuitjs.html)  But Andy has done some fine stuff on his simulations with EasyEDA, check them out. > My first reaction when looking at the data sheet of the 2N7002 was: WOW 7ohms Rds @ 5Vg.  Isn't Rds usually measured in miliohms?  Then I realized this is not a big deal since there will be a current limiting resistor in series with the MOSFET anyway to control the current through the LED and so no big deal :) Thats the max i could be, typical is about 1 Ohm, maybe less, but JLCPCB has 2 sections for mosfets in the parts library: Mosfet and Mosfets, maybe there are better ones in the other category :D If you can reuse others, or just use a resistor as Andy suggested. There are many ways to skin this cat.  @andyfierman > A simple way to add a power good LED (LED on = power good) is to have a or around that value 1k pullup from PG to the VREG5 pin and a LED from the PG pin to ground. > It wastes power when the PG pin is low but only a little more than when the PG pin is high. That's real nice actually! Forgot about that one, have used similar stuff in other LED stuff for a Li-Ion charger IC. And this being a CNC thing, the little power consumption might not be a factor at all :)

Thank you @deskpro256 and everyone, I have updated my desgin to use the C112970 crystal just so that all my parts have a spec sheet.  I will leave pin 1 unconnected for now as it is not clear to me what it does. I'll be looking into simulations in the near future. Thank you again for all the assitance, it is greatly appricated. Best,