interesting conceptual question

Thread Starter

durable126

Joined Feb 20, 2016
56
So here is a question

I was fooling around with trying to get an LED to blink at a specific frequency

A trigger from my frequency source is connected to the gate of an N chanel logic level get well within the VGS voltage. Basically i have a square wave oscillator on the gate

I put my scope on the drain and i can tell that my mosfet is switching like it should be and at the right hz

However if i put the scope probe on the other side of the LED i get garbage. How is this possible?

So the first scope image is from the drain.

The second is from the other side of the LED

I was trying to figure out why my LED was not blinking at the 38Khz i have it triggered. The messed up scope image is obviously to blame but I'm not exactly sure why. A BS170 N chan should be more then happy switching at 38khz

Any ideas?
Ive included a basic schematic to help get the point across

''

Please excuse the cap and resistor in this picture. Its part of the trigger.
Nothing to do with the LED circuit


Scope image from the gate and the same at the drain. Showing my switching is working correctly



Now a shot of my signal on the other side of the LED
Obviously not switching correctly

Anyone know why?
 

WBahn

Joined Mar 31, 2012
24,964
What do you mean by "other side" of the LED?

The second scope capture says that the frequency is 38.02 kHz, which is essentially the same thing the first one is saying.

My guess is that the problem is that you are auto-triggering and that your trigger threshold is being set to something that causes it to trigger at different points in the waveform. Set it to trigger on a manually set level and then adjust the trigger level to stable point in the waveform.

If you are probing the anode of the LED (the top of the part in your schematic) then you are measuring the voltage drop in whatever is driving the LED and not really the LED itself.
 

Thread Starter

durable126

Joined Feb 20, 2016
56
By other side i mean the source side of the LED.

The 5Volt ref for the second scope picture and right at the drain for the first scope image

So yes the anode side of the LED

Please explain if you do not mind why this would be measuring the drop across the source. which is just a dc power supply
 

Thread Starter

durable126

Joined Feb 20, 2016
56
You were right it was the trigger.

By manually adjusting up the threshold it stabilized.

Thanks for the suggestion. Of course this has given way to a whole new bunch of problems

Im sure you had guessed, but this is part of an IR transmitter i built

Now i have been able to confirm that the IR LED is blinking at 38khz.

Sadly, the receiver is still not seeing.

I have been able to test the receiver with a regular TV remote control so i know it works.

I am at a loss now i guess..
 

WBahn

Joined Mar 31, 2012
24,964
By other side i mean the source side of the LED.

The 5Volt ref for the second scope picture and right at the drain for the first scope image

So yes the anode side of the LED

Please explain if you do not mind why this would be measuring the drop across the source. which is just a dc power supply
If a DC voltage source is applied to the anode and you connect your probe to the anode, aren't you connecting your probe to a DC voltage source?
 

Thread Starter

durable126

Joined Feb 20, 2016
56
If a DC voltage source is applied to the anode and you connect your probe to the anode, aren't you connecting your probe to a DC voltage source?
true i suppose

i did make a mistake in the schematic, the resistor is before the LED not the way it looks now
 

WBahn

Joined Mar 31, 2012
24,964
That's reassuring because otherwise it looked like you were using a very poor voltage source -- but coin cells used in some circuits like this are very poor voltage sources.

Okay, so in that case you are measuring the change in voltage across that resistor, which effectively means you are measuring the current in the resistor. But that works for your purposes.
 

Thread Starter

durable126

Joined Feb 20, 2016
56
That's reassuring because otherwise it looked like you were using a very poor voltage source -- but coin cells used in some circuits like this are very poor voltage sources.

Okay, so in that case you are measuring the change in voltage across that resistor, which effectively means you are measuring the current in the resistor. But that works for your purposes.

Right, now i just need to figure out what my IR RX is not triggering given the fact that the light is at the proper frequency
 

wayneh

Joined Sep 9, 2010
16,124
Many receivers will not stay triggered with a constant signal received. They're looking for a pulsed waveform.
 

Thread Starter

durable126

Joined Feb 20, 2016
56
the receiver is literally just the discrete parts. there is no encoding hardware on it

https://www.sparkfun.com/datasheets/Sensors/Infrared/tsop382.pdf

Nothing more then a vss power

after reading the data sheet, i can see you guys are right

This is not good. Anyone know of a RX that will respond to continuous signal ?

I heard about a Sharp GP1U58X that works as a beam break but i don't see them for sale anywhere
 
Last edited:

WBahn

Joined Mar 31, 2012
24,964
What, exactly, is it you are trying to achieve?

Note that this question shouldn't include a description of the circuit or signal you are trying to work with, but the actual problem you are trying to solve.
 

Thread Starter

durable126

Joined Feb 20, 2016
56
What, exactly, is it you are trying to achieve?

Note that this question shouldn't include a description of the circuit or signal you are trying to work with, but the actual problem you are trying to solve.
It would take me all day to explain it. This is a small piece of a much larger project.

From what i just read in the last 5 minutes i think my design will still work, i just need to find a way to add 15 ns or so delay in between the pulses and it should pick it up.

The Sharp GP1U58X would be the best option given it detects continuos signals but can't find them anywhere.

of course this could be easily done with a micro controller and i could save myself a mass amount of headache.

Still trying to get her done with discrete parts.

Any ideas how to add a delay?
perhaps a small bit of logic?
 

WBahn

Joined Mar 31, 2012
24,964
We really need a fuller idea of what you are trying to accomplish. We don't need details of the larger project, but we need specifics related to THIS part of the project.

Saying that you want to add 15 ns of delay between the pulses doesn't give us much to work with. What pulses? Remember that all we have seen described so far is a circuit that strobes an IR LED at 38 kHz. Are those the pulses that you want to delay by 15 ns?

How do you envision this being easily done with a micro? You would have to clock the micro at over 66 MHz just to have a clock-cycle that is 15 ns. To do anything meaningful with 15 ns resolution you would probably need many times that.

Say you have a black box that your transmitter would be a part of. It has a logic signal input going to it. You have another black box that your receiver would be a part of. It has a logic signal coming out of it. What do you want the relationship to be between the input signal to the first box and the output signal of the second box?
 

Thread Starter

durable126

Joined Feb 20, 2016
56
We really need a fuller idea of what you are trying to accomplish. We don't need details of the larger project, but we need specifics related to THIS part of the project.

Saying that you want to add 15 ns of delay between the pulses doesn't give us much to work with. What pulses? Remember that all we have seen described so far is a circuit that strobes an IR LED at 38 kHz. Are those the pulses that you want to delay by 15 ns?

How do you envision this being easily done with a micro? You would have to clock the micro at over 66 MHz just to have a clock-cycle that is 15 ns. To do anything meaningful with 15 ns resolution you would probably need many times that.

Say you have a black box that your transmitter would be a part of. It has a logic signal input going to it. You have another black box that your receiver would be a part of. It has a logic signal coming out of it. What do you want the relationship to be between the input signal to the first box and the output signal of the second box?
The relationship between the TX and RX is nothing important. I just need to detect change. The data is unimportant. ones zeros whatever i don't care. I just need to know when the signal changes.

basically a beam break

Only in this case I'm using a narrow field of view IR LED (7 degrees) to track something that is moving.

in other words. when the RX gets a "high" signal from the IR LED aka the "continuous signal at 38khz" i can ascertain that the something is facing the sensor.
 

Thread Starter

durable126

Joined Feb 20, 2016
56
Why not just start with an IR photodiode or phototransistor and then filter the output as you see fit?

ah yes good idea, perhaps a photodiode would work. I was unaware they made ones for IR

however thee one part i forgot to mention is there are multiple of these units in place. I need to be able to distinguish between them. Given the fact that IR receivers are available in different freqs configurations i had intended to use a different one for each sensor to be able to figure out which one is going off

So using an IR photo diode would kill that deism requirement i guess
 
Top