Is this really smart or stupid?Replacing resistor with Zener

Go To Last Post
24 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Good Day,
Firstly refer to the attached pic.
In the top layout there are the usual resistors to set in the photo diodes input current. However if I have lots of these inputs, I have to have lots of resistors. Currently my design has about 64 of these resistors I will have to solder.
However! If I know the forward drop of the photo diode is 1.2v, and I know my input is 5V, can I just put a Zener Diode of say 3.9V and run all the lines to this zener, thus saving lots of components?
Any reason why it should not work? Long term robustness?
Thanks for any insights.

Attachment(s): 

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I would work through what happens at X11 1, X11 2 and X11 3 when you make X11 4 HIGH. In other words, are the other 3 input ports unaffected by happens to the 4th port?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

hi,

this will not work properly.
Photo diodes, zeners and even your 5V will have tolerances and temperature drift. Even if the current through one photo diode would match, the current with 2 or more diodes will not.
But you can use resistor arrays with a common pin. Connect the common to gnd and the others to cathodes of leds. There are arrays with 8 r`s and 9 pins.


Beware of the bear

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Just to add on what they said, this setup will not limit the current running through the pins, so you would likely burn something me thinks

Kim

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

1. Zener is a the and this means that drop out voltage is different from one to other. A very small change on this can cause a big enough current flowing.
2. Different current causes different drop out voltage.
3. In practice zeners smaller than 6V8 need more than 6mA to be polarized.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

davef wrote:
I would work through what happens at X11 1, X11 2 and X11 3 when you make X11 4 HIGH. In other words, are the other 3 input ports unaffected by happens to the 4th port?

Do you mean work through it in spice, or build it, or use some gray matter?
Well to me, If X11_4 is ON, then current flow through the 4th photo diode and through the zener, causing the zener to turn on. The 3.9V across the zener then reverse biases the other other photo inputs (1 to 3) making them more "off". I cant see anything wrong with that?

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

tossia wrote:
hi,

this will not work properly.
Photo diodes, zeners and even your 5V will have tolerances and temperature drift. Even if the current through one photo diode would match, the current with 2 or more diodes will not.
But you can use resistor arrays with a common pin. Connect the common to gnd and the others to cathodes of leds. There are arrays with 8 r`s and 9 pins.

I also like the idea of resistor arrays, move the resistor to the cathode side of the photo diode, then connect the common of all the resistor arrays to gnd. The only problem is I have to order them over the internet cause my supplies in my country stock a few rare sizes, grr!!! While I was pondering wheter to size up down the current limiting resistor, it made me unhappy, then i tried to think of another way.

I can get some super tiny smd resistor packages but sheesh their dimensions are microscopic. I wonder if our future children will even know what DIP/DIL meant.

Oh, and thanks for confirming what I suspected to be true, but I was optimistic about the idea. Ok back to pondering attainable resistor network values and which one is the lesser evil.

Thank you for you advice!

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

kherseth wrote:
Just to add on what they said, this setup will not limit the current running through the pins, so you would likely burn something me thinks

Kim

Is the current not limited by the voltage applied to it? If I put a 1.1V source across the opto, would it eventually burn out?

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Another related question if you people dont mind...
The data sheet says the photo diode can only with stand a maximum reverse V of 6V. This worries me because the input V is 24V. And if some1 corrects it the wrong way round... you cant put the smoke back in.

So I could place an anti parallel diode with the photo diode (according to literature I have researched), but that would again = lots of components. Is it ok if I just connect one diode at the common path to GND? I used a schottky diode for a low V drop.
I plan of testing this tomorrow when I have access to some tools. But would still like some confirmation if it is ok in the long run even if it passes my testing.
Thank you, you knowledgeable ones!

Attachment(s): 

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes with 1.1V the current will be as rated, but what about 1.2V, which it could be because of temperature drift and the fact that you have 5%(?) error in the zener.
What current would pass through then?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Double....

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

OK I know understand!! wow! I didnt realise it was that extreme. Thank you for your persistence! in explaining it.

With a resistor the current cant run away because if will steal photo diodes biasing voltage, which will decrease the current. But with the zener, things can go pear shaped fast. Thanks once again.

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Oh, and the reason I am asking about the reverse voltage protection, is because I dont understand if the photo diode is destroyed when applied voltage is greater than 6V because a high reverse leakage current starts flowing, in which case, a common diode to all of them (D18 in the figure above) will protect all the optos. However if it is just the presence of the high voltage that destroys the optos, then even if the diode blocks the current, they still have a voltage applied that is far too great. It seems some optos must die for the cause tomorrow....

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I gather you mean the LED, a photo diode is another thing. Yes, the LED in the optocoupler will not appreciate a reverse voltage. Usually one would put a normal diode across the LED to take care of this. A common diode will not work as you expect - the current will then run through the other LEDS.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ah thank you for correcting me Kartman, yes I dide mean LED.

For any one who is interested, I found the TLP620-4 optocoupler. It is about 3x the price of a normal 4CH opto, but has two LEDs back to back so it accepts AC inputs. So for me the internal reverse parallel LED will act as protection and save me putting in lots of diodes.

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have been thinking about it, and I cant see why the common diode will not work.

Quote:
- the current will then run through the other LEDS.

I stare and stare at all possible combinations I can think of and cannot see it. Please see the attached figure.

The top schematic is under normal operation. Input lines driven high create a V drop across the LEDS of 1.1V. Linput lines driven low create a reverse volt drop 0.5V, which is ok (much less than 6V).

The bottom schematic is under inverse polarity connection(user mistake).
There are only two possible conduction paths:
1) From the common 24V, to an input GND.
This is not possible because of the common diode D18.

2) From an Input 24V to an input GND.
If current flows from an input 24V, it must flow through 1 resitor, through a LED, and only then through the reverse LED. Now Assuming the reverse LED is at worst a short, that would create a reverse current of 20mA, I dont know if this would damage the LED, the datasheet says nothing about the max reverse current, only the reverse Voltage.

Any people who can think out the box?

Attachment(s): 

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ok now I am really confused, I did some testing, and even if put maximum reverse voltage that my variable power supply can give (28V), directly on the LED in reverse. It doesnt blow (0uA flow, diode test passes fine, still operates fine)! These things are pretty tough. I am trying to make it blow, but it just want to keep on working.

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You did not say in how many different ways it is possible to connect the polarities wrong.

Is it possible that one optocoupler input is connected okay, but another is wrong? In this case the diode would not work.

Think about it, +24V on one input, -24V on one input. Still 24V in reverse over one led if grounds for the 24V inputs are separate (floating). If 24V comes from the same source, well, it was already short circuited so there is 0V over all inputs.

It is not the reverse voltage itself that destroys the LED, it is the reverse current. It also may not destroy immediately, it just degrades over time. Reverse current is usually in the orders of microamperes, not milliamperes, and LED is not a short circuit when reverse biased - if it is, it has already given out it's magic smoke.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

24V goes through LED#1,2 so the pink line is 24V-1.2V(approx). this finds it's way back to 0V through LEDS#3,4 - reverse biased. Bzzt.

Either a diode per opto or your magic AC optos. Or make sure you don't wire the inputs backwards!

Get yourself a real PLC and reverse engineer it. The cost of the PLC will be saved by saving you time .You'll also learn what they do on industrial grade equipment. How do you think I learned? It's not like you walk out of university with a piece of paper and are an expert on designing everything. You got to find out somehow, or make a lot of mistakes learning.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

When you have lots and lots of same-value resistors to use, the easiest way is to use resistor packs, like this:

http://search.digikey.com/script...

This is an 8-pack, but they make em in up to 16 per package as well...

EDIT: Re-reading this I see now this was mentionned a few posts above... ;)

If you need to choose higher or lower than design, always go a little bit higher in resistance rather than lower. As long as you supply enough current to the LED for it to light up, it`ll work.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:
24V goes through LED#1,2 so the pink line is 24V-1.2V(approx). this finds it's way back to 0V through LEDS#3,4 - reverse biased. Bzzt.

Either a diode per opto or your magic AC optos. Or make sure you don't wire the inputs backwards!

Get yourself a real PLC and reverse engineer it. The cost of the PLC will be saved by saving you time .You'll also learn what they do on industrial grade equipment. How do you think I learned? It's not like you walk out of university with a piece of paper and are an expert on designing everything. You got to find out somehow, or make a lot of mistakes learning.

lol at "magic AC optos" :D
You right, I am basically making a PLC! I never even realised it. I should of googled "PLC input schematic" ages ago....
My circuit just somehow grew and grew as I wanted to add more and more to it.

Thats some good advice kartman! I dont know why you hang around and throw bones to noobs like me, but I appreciate it!! So I took your advice and took apart a PLC and had a looksie. And it is just SMD central. Identifying parts was a mission. After a lot of googling this is what I have deduced from a certain manufacturers input circuit(see figure below):

The two diodes D1 and D2 I think is diode package. It only says "A7-" on it. The "inductor" at the input is still a mystery to me. I only assume it must be an inductor because the series resistor 1K5 already sets the LED current to the datasheet's recommend If of the LED (15mA). Hence I assume what ever it is directly at the input, it must have a low resistance or else it will drop the applied voltage too much. Turns out to measure 2K7, so there goes my theory...

What I see and like is when a reverse voltage is applied, it is not only bypassed by a reverse anti parallel diode, but is also block by an additional diode.

The "inductor" & "cap" I can only guess are to smooth the voltage across, and the current through the LED. The "cap" might be some sort of transzorb that clamps the input if it gets too high. It measures 0nF on a fluke multimeter.

So after seeing all the effort the big dogs put into each input, even though I am not a speedy robot positioner, I think I will throw some additional diodes at my design. Maybe even add that bypass CAP from the input to GND too. I am still puzzled by the mystery component at the input...

Attachment(s): 

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Jepael wrote:
You did not say in how many different ways it is possible to connect the polarities wrong.

Is it possible that one optocoupler input is connected okay, but another is wrong? In this case the diode would not work.

Think about it, +24V on one input, -24V on one input. Still 24V in reverse over one led if grounds for the 24V inputs are separate (floating). If 24V comes from the same source, well, it was already short circuited so there is 0V over all inputs.

It is not the reverse voltage itself that destroys the LED, it is the reverse current. It also may not destroy immediately, it just degrades over time. Reverse current is usually in the orders of microamperes, not milliamperes, and LED is not a short circuit when reverse biased - if it is, it has already given out it's magic smoke.

Oh, I should of been more clear, there is only a 0V and +24V from the power supply, no -24V. But that is a good point, it may be a tiny weeny reverse current that slowly breaks it down. I did have my multimeter set to 400uA range, and it always red 0uA even when the reverse voltage was 28V. I still think I will scrape my common diode idea *me takes a moment to say goodbye to it* :cry:

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

UNiXWHoRe wrote:
When you have lots and lots of same-value resistors to use, the easiest way is to use resistor packs, like this:

http://search.digikey.com/script...

This is an 8-pack, but they make em in up to 16 per package as well...

EDIT: Re-reading this I see now this was mentionned a few posts above... ;)

If you need to choose higher or lower than design, always go a little bit higher in resistance rather than lower. As long as you supply enough current to the LED for it to light up, it`ll work.


Hmm good point, rather safe then sorry. Thanks for the input man!

Just a noob in this crazy world trying to get some electrons to obey me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Unixbeatch, you really want to have a bit more current through the LED rather than just enough. LEDs age so you want to make sure there's always a little more than enough current to make it work.

fever: I dare say the tube device is just a resistor as it sounds like a MELF package. Calc the watts in the resistor and you'll find it's a little more than you can get away with in a 0805 size SMD. The Twido PLCs I've looked at use a 1W SMD resistor at 3K3 if I recall correctly.