MOSFET LED Switch Not Working

ci139

Joined Jul 11, 2016
1,898
but it’s not working
trivia : http://www.everlight.com/file/ProductFile/67-31EP6-UY0500H-AM.pdf
https://www.diodes.com/assets/Datasheets/ds30120.pdf ← (Please see figure 4 -- the normalized current means 1x 2x ... Nx of the 7.5Ω nominal?)
(5V - 2.19V.LED(if higher the current will be lower...) ) / I.drain = 75Ω + 7.5Ω.MosFet(with 10kΩ or greater R.drain / with lesser - worse !) → I.drain.max = 2.81 / 82.5 Ω = 34mA and below
__
Please , specify ? LED always ON ? LED always OFF ? LED state hard to determine ...
dbl-chk
 
Last edited:

Thread Starter

Gpelon

Joined Sep 17, 2019
7
Please , specify ? LED always ON ? LED always OFF ? LED state hard to determine ...
dbl-chk
The LED is always on in a dim state, it never completely shuts off. But it will brighten when the GPIO pin get pulled high manually.
 

cmartinez

Joined Jan 17, 2007
8,218
What did you do with the other nFet included in the chip? That is, have you connected pins 1 and 2 to ground, or did you leave them floating?
 

Thread Starter

Gpelon

Joined Sep 17, 2019
7
Another thing, what MCU are you using to drive the nFet? Have you made sure its GPIO is configured as a push-pull output?
The other channel on this nFet has this identical circuit. And I’m using a stm32 which is configured for push/pull.
 

cmartinez

Joined Jan 17, 2007
8,218
They are not down to 0. Vg at LED1 is 0.04 V and Vg at LED2 is at 0.24 V
So, most likely there's something going on with the GPIO's .... I would disconnect the gates from the MCU's pins (but leaving the 10k pull-down resistors), and then test them by by manually connecting and disconnecting them to Vcc. If they work as planned, then definitely there's something going on with the MCU...

There are four ways one can normally configure a GPIO:
  1. push-pull output
  2. high-z input
  3. open collector output
  4. quasi-biderectional I/O
Maybe you've accidentally configured them as #3 or #4?
 

cmartinez

Joined Jan 17, 2007
8,218
Another thought that comes to mind, is that you might inadvertently be applying a PWM signal to the gates. Do you have an oscilloscope?
 

cmartinez

Joined Jan 17, 2007
8,218
hi Gp,
Do you a photo of the project that you could post which shows the FET and LED.?
E
Good call... if he's working with an SMT device, then surely it's mounted on a makeshift PCB, and maybe its traces are short circuited at some point... that would explain the non-zero voltages at the gates.
 

Thread Starter

Gpelon

Joined Sep 17, 2019
7
Another thought that comes to mind, is that you might inadvertently be applying a PWM signal to the gates. Do you have an oscilloscope?
I’ve actually already been trying to manually pull the signal high or low on a breadboard and I’m still getting the same issue, so it’s not the MCU. What are some ramifications of overheating the mosfet during soldering? Maybe that caused an issue?
 

Thread Starter

Gpelon

Joined Sep 17, 2019
7
Good call... if he's working with an SMT device, then surely it's mounted on a makeshift PCB, and maybe its traces are short circuited at some point... that would explain the non-zero voltages at the gates.
I actually have a prototype PCB that I’m using this on, but I cannot send pictures of it as it is proprietary
 

cmartinez

Joined Jan 17, 2007
8,218
I actually have a prototype PCB that I’m using this on, but I cannot send pictures of it as it is proprietary
No problem, we understand. My only advice would be for you to be meticulous and do a step by step check of all possible scenarios. And also, we'd love to hear about it when you finally solve it... we're all of a curious nature in this place. ;)
 

ci139

Joined Jul 11, 2016
1,898
manually pull ... same issue
-- try 100kΩ or higher to gate from +7V +10V +14V -- while so log V.drain to GND voltages (keep LED at where it is +5V)
!! but also see the thermal characteristics (410°C/W J²A with minimum landing) and SOA of the 7002 d/s .. before !!
-- there may be issue with "power to MCU" - pattern (being) not sufficient for your attempted operation =
= PWM 2 GPIOS with 10k to GND . . . ((unlikely))

- - - i had to solder BSS139 (depletion mode) from 12V battery with –DC - iron because it fried with ~AC 1
 
Top