LED dimming ???

Thread Starter

banzai

Joined Jan 3, 2014
71
HI

I made a simple circuit which works great except one thing. I have attached a hand drawn diagram of the circuit. The problem is, when the MOSFET kicks in without a load on it except the indication LED, the LED works fine (no dimming), when i connect another load to the MOSFET like a small relay, it works fine except the LED now suddenly dims (goes almost fully dark) when the MOSFET is energized and slowly brightens to about 80% of the brightens that i would get when the only load on the MOSFET is the LED by itself. What am i missing here ???



Thanks in advance guys
 

Thread Starter

banzai

Joined Jan 3, 2014
71
What is the "inductive load"? Resistance? Current required to operate?
The load is an automotive relay that runs on 12vdc and has a resistance of 82 ohms and which would be almost 150mA. I have tried to run other things and ended up with the same result except the time that the LED brightens are different from load to load, the higher the current the slower the LED brightens.
 

MikeML

Joined Oct 2, 2009
5,444
When the LM324 output is high, what is the gate voltage? What is the drain voltage? How warm or hot does the NFET get?
 

Thread Starter

banzai

Joined Jan 3, 2014
71
When the LM324 output is high, what is the gate voltage? What is the drain voltage? How warm or hot does the NFET get?
The LM324 output is at 3,7V when HIGH. The drain voltage is at 0V when the LED is run by itself, but if the relay is connected the drain voltage gets little above 3V and starts to drop until it stops as 1,3V. The MOSFET is not getting even near hot
 

MikeML

Joined Oct 2, 2009
5,444
Ok, to turn on that crummy old NFET, its Vgs needs to be approaching 10V. Either get a modern "logic level" gate NFET, or redesign the LM324 circuit.

The reason that the LM324 is only pulling to 3.7V is because it is powered with only ~4.8V. Is there any reason for this?
 

elec_mech

Joined Nov 12, 2008
1,500
I assume you're trying to use the MOSFET as a switch, is this correct?

If so, you're using a standard MOSFET which requires at least 10V at the gate to turn on. 3.7V is allowing the gate to pass a little current, but not much.

You could eliminate the voltage regulator and use 12VDC throughout your circuit - however, I'm sure you've done this for a reason and I don't know what VTC is or does, so this may not be a viable option. Alternately, you try using a logic level MOSFET like the IRL520. It requires 5VDC at the gate to fully turn on, so you may need to increase the voltage regulator output to 5VDC if this still allows everything else to work. There are MOSFET's available that turn on at a lower voltage like 1.8V, so 3.7V would be fine, but the few that I found with a similar ~9A rating were surface mount.
 

Thread Starter

banzai

Joined Jan 3, 2014
71
The reason that the LM324 is only pulling to 3.7V is because it is powered with only ~4.8V. Is there any reason for this?
The plan was to make the lm317T output 5V to be able to handle a good range of a voltage drop caused by a battery discharging
 

Thread Starter

banzai

Joined Jan 3, 2014
71
I assume you're trying to use the MOSFET as a switch, is this correct?

If so, you're using a standard MOSFET which requires at least 10V at the gate to turn on. 3.7V is allowing the gate to pass a little current, but not much.

You could eliminate the voltage regulator and use 12VDC throughout your circuit - however, I'm sure you've done this for a reason and I don't know what VTC is or does, so this may not be a viable option. Alternately, you try using a logic level MOSFET like the IRL520. It requires 5VDC at the gate to fully turn on, so you may need to increase the voltage regulator output to 5VDC if this still allows everything else to work. There are MOSFET's available that turn on at a lower voltage like 1.8V, so 3.7V would be fine, but the few that I found with a similar ~9A rating were surface mount.
if this is the case i am thinking of putting in a small transistor between the lm324 and the MOSFET and have the transistor source 12v to the MOSFET. Would that be a good idea ?
 

MikeML

Joined Oct 2, 2009
5,444
Power the LM324 with the unregulated 12V instead of the 5V. Increase the 12K hysteresis resistor to about 47K, and try it again.

The trip points will be slightly effected by the variation in the 12V supply. If this is a problem, we can fix it by adding a diode.
 
Last edited:

Thread Starter

banzai

Joined Jan 3, 2014
71
I do get the idea now thanks to you guys. One thing just to make sure of, does driving a transistor or a mosfet at lets call it (not full gate voltage) have any damage effect on the semiconductor ?
 

MikeML

Joined Oct 2, 2009
5,444
I do get the idea now thanks to you guys. One thing just to make sure of, does driving a transistor or a mosfet at lets call it (not full gate voltage) have any damage effect on the semiconductor ?
Only if it heats and melts. You were not even turning it on...
 

MikeML

Joined Oct 2, 2009
5,444
I'd follow MikeML's advice - it will result in fewer parts and fewer things to go wrong.

But if you really want to try adding a transistor, I've attached a circuit that might work. I'm no kind of transistor expert so this may or may not work.
I will not work in this case. The OP needs gain between the LM324 and the gate. This circuit is an emitter-follower, and has no voltage gain.
 

Thread Starter

banzai

Joined Jan 3, 2014
71
what is the explanation behind the LED brightening, is it because of the mosfet capacitvily charging up and as it does that the LED gets brighter (MOSFET putting out more voltage) ?
 

elec_mech

Joined Nov 12, 2008
1,500
I will not work in this case. The OP needs gain between the LM324 and the gate. This circuit is an emitter-follower, and has no voltage gain.
Doh, well thank you. Well, I was half right, I'm no transistor expert by any means. I think what is needed is a common emitter circuit, but whether I use an NPN or PNP, the signal is inverted, so two transistors would be required to do what the OP wants with transistors (unless I'm missing some vital logic), so it really isn't worthwhile.
 

Alec_t

Joined Sep 17, 2013
14,330
The FET may even be oscillating due to wiring inductance. That would give weird results.
If your 12V supply is more than a few inches from the LM317 you may need an input decoupling cap for the LM317.
 

ronv

Joined Nov 12, 2008
3,770
The gate threshold voltage for your FET is between 2 and 4 volts. This is the point where it just begins to turn on. (250ua.) So it is slightly on, but cant allow much current to flow. So it may light the led, but when it needs more current it can't. But as it warms up a little from the added current the threshold goes down a little so it may supply a little more current.
 
Top