Fix up pulsed DC signal?

Thread Starter

kenw232

Joined May 18, 2009
54
Hello, I'm an amateur. I'm curious how to increase the integrity of a pulsed DC signal from a mosfet. I have a IRFZ44N that is simply turning a LED on and off. It works fine. When I measure across the LED
with an oscilloscope I see a bit of a wobble in the signal at 0V, see the attachment called without_cap.jpg (circled in green).

So I add a cap in parallel with the LED and then I get with_cap.jpg which helps smoothen out the 0V signal (green circle) but the drop off is curved more (the purple circle) now. Being a novice I don't know what to do to try to get it to cleanly drop off and keep the signal flat when theres suppose to be no voltage. Anyone know what I'm talking about or what would be causing this?
 

Attachments

#12

Joined Nov 30, 2010
18,224
Without a schematic, I'm just guessing, so I guess that the 60 Hz wobble is coming from the power supply or a weak ground on the scope probe.
 

Thread Starter

kenw232

Joined May 18, 2009
54
I've attached a crude mockup of my circuit. Its very simple. The +5V square wave going into the gate of the mosfet is at a frequency of about 5Hz. I measure across the LED sometimes at A & B or A and -.

Again, I don't know what I'm doing. I just see a bit of a ripple as I mentioned and would like to know what is causing that or how to fix it if anyone knows by any chance.

Note the LED might be backwards in the diagram but it works fine.
 

Attachments

tshuck

Joined Oct 18, 2012
3,534
Are you sure that is direction of the LED? It's reverse-biased. Since it works you didn't hook it up as shown...

Are all off your grounds connected (oscilloscope/power supply/square wave generator)?
 

BC107C

Joined Apr 3, 2011
14
IMHO it is the high impendance in the drain causing the noise pickup at the probe.
To verify, add a 1k to 20K resistor from V+ to the MOS Drain( if scope gnd is connected to V+) or 100k from drain to ground( somehow in paralel to the LED and series resistor, if any when scope GND is connected to system V-) then recheck. The waveform should follow the Gate signal without noise pickup.
 

tubeguy

Joined Nov 3, 2012
1,157
You don't show a current limiting resistor in series with the LED.
If this is a 'normal' LED with 2.1v drop 20 ma max current, then put a 470 ohm resistor in series with the LED, unless it has a built-in resistor.
Without current limiting you will draw excessive current which can cause excessive ripple.
 

Thread Starter

kenw232

Joined May 18, 2009
54
Are you sure that is direction of the LED? It's reverse-biased. Since it works you didn't hook it up as shown...

Are all off your grounds connected (oscilloscope/power supply/square wave generator)?
The LED is probably backwards in the diagram.

Yes, most of the time all my grounds are connected together. Not sure if that is correct. Can someone confirm that is ok?
 

Thread Starter

kenw232

Joined May 18, 2009
54
I went to add a 470 ohm resistor in there but I can't reproduce my problem now, my ripple is gone. I am clueless, maybe you all should ignore me.

So what is the proper way to measure the signal across the LED? One scope probe on each side of the LED or one lead on negative (-) and the other on one side of the LED?
 

WBahn

Joined Mar 31, 2012
29,976
I have a IRFZ44N that is simply turning a LED on and off. It works fine.
If it "works fine" then what is wrong with it? If there is something wrong with it that needs fixing, then it doesn't "work fine". Instead, you need to give some kind of indication of what DOES count as "working fine".

Remember -- a circuit is intended to serve a purpose and how well it is working needs to be determined witin the context of what purpose it is intended to serve.

In this case, I gather you are just trying to understand why you don't have a clean square wave when you measure it. That's fine and worthwhile from an educational standpoint. But always try to give a decent description of what it is you are trying to achieve, otherwise you can cause a lot of miscommunication and end up "fixing" things that have nothing to do with what you are trying to achieve.

I've attached a crude mockup of my circuit. Its very simple. The +5V square wave going into the gate of the mosfet is at a frequency of about 5Hz. I measure across the LED sometimes at A & B or A and -.
I suspect that here is your problem. Most scopes are not capable of making differential measurements. The scope probe ground is connected to the chassis ground of the scope which, in turn, is generally connected to the third prong on the power cord (i.e., earth or "safety" ground). Thus you can't just connect the probe ground to any point you want. If your circuit is referenced to ground via the power supply or any other means, then you will introduce ground loops that can cause all kinds of noise (or worse).

I went to add a 470 ohm resistor in there but I can't reproduce my problem now, my ripple is gone. I am clueless, maybe you all should ignore me.

So what is the proper way to measure the signal across the LED? One scope probe on each side of the LED or one lead on negative (-) and the other on one side of the LED?
In most cases the best way is to use two probes and two channels and then use the scope's ability to subtract the two channels, assuming it has such an ability.
 
Top