TheBackShed.com

Forum Index : Microcontroller and PC projects : BSS138 N-Ch MOSFET question....

Author

Message

Grogster

Admin Group

Joined: 31/12/2012
Location: New Zealand
Posts: 9610

Posted: 07:39am 24 Aug 2018

Copy link to clipboard

Print this post

Hi all.

I have an application, where I am using some BSS138 N-channel MOSFET's as a low-side switch. PDF attached. BSS138_N-Ch_MOSFET_SOT-23.pdf

They DO work fine, but I am getting some 'bleeding current' when the MOSFET is supposed to be off. Not much - about 0.152mA(152uA), but that is enough(believe it or not), to light up a super-bright LED.

I would prefer to reduce this current.

The circuit I have around all these MOSFET's, is 100k from gate to ground, and 1k from gate to the MM I/O pin. The SetPin command DOES NOT make use of the optional PULLUP.

Can I assume that I need to drop the 100k down to about 10k or so? I tend to settle on 100k gate-to-ground resistor with P-channel MOSFET's(to hold them off), but perhaps N-channel require something different?

The MOSFET's are already on the PCB and are SOT-23, along with 1206 resistor arrays(four separate resistors in one 1206 size package), so I don't really want to start hacking that unless I can get some confirmation that what I think is correct.

OK all you MOSFET gurus! Your turn!

Smoke makes things work. When the smoke gets out, it stops!

Volhout
Guru

Joined: 05/03/2018
Location: Netherlands
Posts: 5089

Posted: 08:53am 24 Aug 2018

Copy link to clipboard

Print this post

Hi Grogster,

The maximum leakage current in a BSS138 is something like 5uA (at 50V Vds).
If you measure 30x more you either have

- Vgs > 0 (the gate voltage is not zero, simply measure with a multimeter). Since the BSS138 has a really low threshold voltage, 100mV (0.1V) can already cause this.

- Vds > 50V (in other words, you have more than 50V on your LED's.).

As you already guess, it could be option 1. Simple to detect, maybe more difficult to repair. Can be caused by a lot of things. Just to name some:

- The Micromite output does not go to zero all the way
a. there is an external pullup causing a voltage divider on the lower FET resistance
b/ you are using a PWM that does not completely switch OFF (always leaves a little spike, hardly detected by a voltmeter, but the BSS138 will switch shortly).
c/ the Vss of the Micromite chip, and the source of the FET doe not have the same voltage due to voltage drop across ground wires.

I hope this helps

Volhout

Edited by Volhout 2018-08-25

PicomiteVGA PETSCII ROBOTS

Frank N. Furter
Guru

Joined: 28/05/2012
Location: Germany
Posts: 949

Posted: 08:54am 24 Aug 2018

Copy link to clipboard

Print this post

Hi Grogster,

I believe it! I had the same problem with a W�rth 150080GS75000 LED. I was already irritated when the LED began to glow weakly during soldering with my supposedly potential-free soldering iron...

Have you tried pulling down the output pin in software yet?

Frank

Bill.b

Senior Member

Joined: 25/06/2011
Location: Australia
Posts: 235

Posted: 09:16am 24 Aug 2018

Copy link to clipboard

Print this post

Hi Grogs

I use mosfets on a number of different projects, I find a 10K from
gate to ground is required to ensure the mosfet turns completely of
when the output of the processor is low.

Bill

In the interests of the environment, this post has been constructed entirely from recycled electrons.

zaphod

Regular Member

Joined: 03/06/2018
Location: United Kingdom
Posts: 93

Posted: 10:59am 24 Aug 2018

Copy link to clipboard

Print this post

Unfortunatly there is another sinister posibility and that is the device has been damaged by ESD during handling or soldering, this has happened to me before! Check your soldering iron tip is actually earthed and ensure you use full anti-static precautions when transfering from packaging to pcb and that pcb is sat on an anti-static mat that you keep touching. ESD damage often makes the devices leaky as you describe. There are no gate protection diodes on this device shown on tha data sheet so this insidious damage is highly likely.

Cheers Roger
1Kwp DIY PV + Woodburner + Rainwater scavanger :)

Grogster

Admin Group

Joined: 31/12/2012
Location: New Zealand
Posts: 9610

Posted: 11:03pm 24 Aug 2018

Copy link to clipboard

Print this post

Hi everyone.

@ Volhout - Good idea. Should have thought of that myself. Gate voltage when 'Off' is 156mV, which explains quite a bit.

@ Frank N. Furter - PULLDOWN option cannot be used on outputs, only inputs.

@ Bill.b - Yeah, this is what I am thinking too. I think I will change ONE of the 1206 resistor arrays, and see if that helps. If it does, I can change the others.

@ zaphod - I always use a fully grounded workstation, iron, and anti-static strap while handling or working with any static sensitive components, so I think I am OK there. I also touch a few metal earthed things such as my microscope and PSU before I touch anything, to ensure I am at the same potential. I know how sensitive the gates on MOSFETS can be, so I went in armed.

I think the problem is the stray 156mV on the gate when the I/O pin is supposed to be zero.

EDIT: Voltage on MM pin when 'Off' is 16mV. I don't know how that translates to 156mV at the gate.

I will swap one of the 100k arrays for a 10k one and report back.

EDIT: Swapped one 100k for a 10k one, but exactly the same problem. Voltage on gate is now 172mV, so we have GAINED 20mV gate voltage by lowering the ground resistor to 10k. A bit mystified by this. Anyone got any more hints or suggestions?

EDIT: Have elected to drop all the MOSFET's and resistors in favour of a couple of ULN2003A relay driver chips. They can handle more current anyway, and I have used these before. Not sure why I did not think to use them in the first place....Edited by Grogster 2018-08-26

Smoke makes things work. When the smoke gets out, it stops!

Print this page

To reply to this topic, you need to log in.