Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Building a fiber optic circuit

Status
Not open for further replies.

fstspec

New Member
Hi

I am trying to build a analog optical coupler that is able to transmit data of at least 14Mbps. I have checked that there is no such optocoupler available with such high data rates, so i am thinking of a prototype that can transmit data using fiber optics.

Does anyone has any good recommendations on how could I go about doing it.
 
Do you have a miniature semiconductor fab hidden in your garage? LOL. Your chances of building a discrete opto-coupler for those speeds are between small and vanishingly small.
 
Last edited:
actually i have tried sourcing for agilent's optocoupler and i found out that couplers at higher speed are digital. Also, i have just graduated from uni, so this is actually my first project.

i have seen data being transmitted using fiber optics. since there is no available ic, i am thinking of building a prototype on my own. it might be bulky to come up with the prototype, at least it will help in my experiment.

do you know how can i go about finding schematics to build a fiber optic links, please enlighten me:D
 
I'm not sure what you are driving at when you say that the higher speed opto-couplers from Agilent(formerly HP and now Avago Technology) are digital. Can you elaborate on how you came to this conclusion?

For a discrete opto-coupler I would probably start with an IR LED and an IR detector. I've not heard of IR LEDs being used at frequencies like 14MHz. but I guess anything is possible. You will probably need a pulsed, high current driver for the IR LED. The IR detector will need to respond to the IR detector at whatever frequency you decide on. The output of the IR detector might go to a comparator, but there are precious few discrete comparators that will operate at those speeds.

Good luck and let us know how it works out.
 
I saw from the datasheets that the acpl 700 is a 3.3V digital ttl optocoupler. please enlighten me.

this project of mine is actually trying to test how much common mode current can be reduced by using optocoupler as isolation in analog signals. the input signal will be a sine wave will be oscillating from 4-30Mhz with a pk to pk voltage of 1-3volts.
 
There's a project floating around where they use a LED to transmit 10Megabit/5MHz ethernet data, and IRDA (the oddball infrared protocol who nobody uses) has a 4megabit mode, with 1/4 modulation(?) which is effectively a 8MHz pulse. One thing to look at might be using some laser drivers - Maxim makes a couple of high speed communication laser drivers, and they might be able to modulate analog at a pretty high frequency. I'm pretty sure high speed fiber modules - 100 megabit and above are laser based.

Keep in mind isolation transformers are typically the first choice when dealing with high frequency isolation, but if the point is to be testing opto isolation...
 
ok, i will keep a look out for the laser drivers. I know that high speed optocouplers at 100Mbd are digital version. Can i is it feasible to use a ADC at the front end and a DAC at the output of the digital optocoupler. Since my job is to lower EMI, does anyone has idea whether it will cause more EMI if i use this method?
 
How much resolution/accuracy do you need? Multiply by the number of bits, and you'll get an idea of how many optocouplers you'll need. 8 bits @ 60MHz sampling rate (nyquist requires 2x sampling rate at an absolute minimum)=> 500mbit/sec. There are gigabit class fiber interface modules, but as far as I know, they're a couple hundred USD each. (But then again, I don't especially have any need for one).

Going from analog to digital and back is really roundabout, plus 60MHz A/D converters aren't exactly "quiet". What kind of system is this thing going to be integrated into?
 
the system already has a 4-21Mhz coupling transformer to provide the isolation. the optocoupler will be placed after the transformer so as to provide better isolation.
 
So are you trying to reduce any sort of capacitive coupling between the stages?

Well, in any case, one thing that I might try is to essentially do a "1 bit" Analog -> Digital -> Analog conversion - where the Digital is just a PWM signal. The idea is to have a high speed triangle wave and a comparator on the incoming side which generates a PWM signal which gets sent over the opto. A filter on the output of the opto regenerates the analog signal. I think this general appoach is used when trying to isolate the feedback of a power supply circuit from the input/switching side. Never done it myself, but it sounded like a good idea at the time.

James
 
Now I understand better. You said "analog opto coupler" and "transmit data at 14 Mbits/sec". What you want is a "linear opto-coupler" whose output is proportional to the input. You have signals up to 14 MHz. which you need to get across the opto barrier with minimal distortion. You have your work cut out for you since I've never heard of this being done. My original comment about having a fab in your basement was prescient.

The real problem with opto couplers is that they are inherently current devices, not voltage devices. The current on the output will have the same basic shape but the scaling will be all over the map. There is a property of opto couplers called the current transfer ratio or CTR. On a typical opto coupler it varys from 20% to 800%. Mnaufacturers grade their products and put them in bins which have a more restricted range. Even so with that kind of spread, each channel would have to be calibrated and recalibrated since they degrade with time, temerature, and current.

I wish you good luck in your search for suitable components, but I fear it will be a long and fruitless one.
 
sorry for causing the confusion, but with your help so far, i am able to understand optocouplers more. this project of mine is more of research basis and it will last me one year. even if it might be fruitless one, but i hope that it will be a meaningful one :D
 
I look forward to hearing of your results. Please keep us informed.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top