0

I am working on a simple VLC link. enter image description here A square wave input with 1 MHz signal is coupled with the direct current (DC) by a bias tee (mini circuits ZFBT-4R2GW-FT+) to drive the lighting of the white LED. Next, two plano-convex lenses are placed in front of the LED. These guide the emitted light onto the PIN photodetector. After that, the received signal is processed offline in MATLAB software.

There are few questions based on this setup:

  1. Why did the oscilloscope (bandwidth 1 GHZ and sample rate of 5GS/sec) fail to display any pulse signal?
  2. Does the position of my lenses affect the received optical power?
  3. Is the bias tee really needed for faster modulation? No pulse signal appeared on the oscilloscope even without the bias tee for square signal with 500 Hz frequency.
Ms H
  • 1
  • 1
  • Could you tell LED and photodiode types, capacitance value, amplitude? To start, reduce the frequency, increase amplitude, remove lenses, put photodiode near to the led, to find any signal at all. Also the input-resistance/capacitance of the scope might reduce the photodiodes current. – datenheim Jan 17 '23 at 07:54
  • Proper circuits are needed (not block diagrams). – Andy aka Jan 17 '23 at 08:28
  • @datenheim I used Mounted LED (Peak at 406 nm; max current = 125 mA ;fwd voltage = 6.3V ), Si Biased Detector (350 to 1100 nm; Rise Time at 632 nm = 14 ns; Output voltage = Output Voltage) . I got signal (straight line) around 9.20V when tested the LED near the PD without any lenses ranging from 100 kHz to 1 MHZ. The pulse only appeared on the OSC when I set the input less than 600 HZ. – Ms H Jan 18 '23 at 06:29
  • @Andy aka, I don't have circuit. I am using the available equipment in my lab. – Ms H Jan 18 '23 at 06:36
  • That doesn't mean that a proper circuit can be avoided. EEs work with schematics and not block diagrams. The devil is in the detail and, without the detail, it's guesswork @MsH – Andy aka Jan 18 '23 at 09:11
  • 1
    What's that LED driver you're using? Are you sure it's ok with a bias? – user1850479 Jan 18 '23 at 12:53
  • Ok, the bias tee might be ok. How about the LED driver (datasheet?) and the LED? I'd recommend a single color LED (red, gree, blue, never white because of the "phospor" topic. – datenheim Jan 18 '23 at 21:18

1 Answers1

1

Your connections to the bias tee are reversed. The DC bias should be applied through the inductor, and the modulation signal should be applied through the capacitor. If your actual experiment matches the schematic you provided, then this entirely explains why you don't see any signal at the receiver. You should also not be seeing much of any light at all emitted from the LED.

Does the bias tee is really needed for faster modulation?

You'd have to say what alternatives you are considering.

A bias tee is usable up into the 10's of GHz.

But there are also IC configurations that can combine a DC source and a modulation signal in the same frequency range.

Most of us don't have the facilities to design and construct ICs for hobby projects, but a bias tee is fairly easy to buy or build on a limited budget.

The Photon
  • 126,425
  • 3
  • 159
  • 304
  • I made mistake on the figure above. In my setup, the DC port is connected to the DC power supply and a function generator with 1 MHz square wave is connected to the RF port. The output of RF& DC is then connected to the BNC port on the LED driver. The light from the LED is on. Unfortunately, it is not captured at he RX. I am not sure the problem is coming from the biased detector of the oscilloscope. The sample rate of the OSC is 5GS/sec with 1 GHZ bandwidth. About the bias tee, I bought it from minicircuits (max voltage = 30 V ; input current =500mA). And, set the DC power to 10 V. – Ms H Jan 18 '23 at 06:17
  • @MsH, Most likely the LED driver circuit is not passing the modulation signal to the actual LED. You could connect the bias tee between the driver and the actual LED and it might work. – The Photon Jan 18 '23 at 17:17
  • But it might not. You might want to filter the light reaching the detector so only the blue components are captured. This is because the white LED is really a blue LED illuminating a phosphor that produces the other color components, and the phosphor won't respond much to the modulation signal. But usually the LED component comes out of the LED pretty strongly so you should still be able to detect the modulation. – The Photon Jan 18 '23 at 17:21
  • thank you for the valuable input. – Ms H Jan 21 '23 at 05:13