Go Down

Topic: optoisolator question (Read 1 time) previous topic - next topic

MarkT

There is no abs max rating for pulse output from the pin, you use 40mA as the abs max if you
want to play safe.  100ms is a long time for a single output pad on a microcontroller (pad die
area perhaps 0.001 sq mm), probably reaches thermal equilibrium in a couple of ms.
[ I won't respond to messages, use the forum please ]

TempleClause


Let's try a test:
If you hook the opto-isolator up as in the attached drawing, what does the Ohm meter read.
If you then disconnect pin 1 of the opto-isolator from +5V  what does the Ohm meter read.


It reads 137 Ohm
and then wen I disconnected pin1 it shows 0 Ohm

Grumpy_Mike


50 ohms resistor is fine here, if you want it could be even smaller.

AVR 328 tech sheet;-

Absolute Maximum Ratings:
DC Current per I/O Pin: 40.0 mA

But here is continuous DC current not pulse.

Sorry this is rubbish. The ATmega's pins are not rated for pulses so although damage due to heat can be mitigated with a short pulse the two other damage mechanisms can not.
These are :-
1) Mechanical strain on the bonding wires due to the magnetic field the pulsed signal generates interacting with other fields.
2) Depletion of charged carriers in the doped substrate.

Ratings in a data sheet are there for a reason and only an idiot would make up stuff around absolute ratings.

LarryD

#18
Jul 01, 2013, 12:20 am Last Edit: Jul 01, 2013, 12:25 am by LarryD Reason: 1
Quote
It reads 137 Ohm and then wen I disconnected pin1 it shows 0 Ohm

Now connect the flash up to pins 4 & 3. For it to work the most positive lead has to connect to pin 4 of the opto...
Lets say the tip is +, connect the tip to pin 4 and the ring to 3.
If it doesn't work try the tip to pin 3 and the ring to pin 4.

Are you using the PC123 rather than the PS2532?
The way you have it in your schematic isn't the same as how you have it wired up!

sonnyyu


Sorry this is rubbish. The ATmega's pins are not rated for pulses so although damage due to heat can be mitigated with a short pulse the two other damage mechanisms can not.
These are :-
1) Mechanical strain on the bonding wires due to the magnetic field the pulsed signal generates interacting with other fields.
2) Depletion of charged carriers in the doped substrate.

Ratings in a data sheet are there for a reason and only an idiot would make up stuff around absolute ratings.


Nonsense,  60 mA with 100 ms pulse v.s. 40 mA continuous DC current and non repeat will cause 1) +2) ?
if you read comment of http://www.glacialwanderer.com/hobbyrobotics/?p=10 , the big resistor will not work. Your imagination v.s. working case, I vote second one.

sonnyyu

#20
Jul 01, 2013, 02:30 am Last Edit: Jul 01, 2013, 02:44 am by sonnyyu Reason: 1

There is no abs max rating for pulse output from the pin, you use 40mA as the abs max if you
want to play safe.  100ms is a long time for a single output pad on a microcontroller (pad die
area perhaps 0.001 sq mm), probably reaches thermal equilibrium in a couple of ms.


I guess if single output pad is safe for 40 mA continuous DC current and should be work fine with single 60 mA with 100 ms pulse, repeat in few minutes even longer.


BTW, Where did you get AVR 328 pad die area perhaps 0.001 sq mm? Any link?

Grumpy_Mike

Sonny you just because you link a post from some other idiot does not mean you are right.
Read the data sheet 40mA is NOT A SAFEL LEVEL it is the point where damage is done.

It is clear that you think you know more that you actually do. This forum is not a haven for those sorts of people. Please learn what this forum is about before posting any more rubbish.

As I said before
Ratings in a data sheet are there for a reason and only an idiot would make up stuff around absolute ratings.

If the cap fits.

sonnyyu

#22
Jul 01, 2013, 02:06 pm Last Edit: Jul 01, 2013, 02:43 pm by sonnyyu Reason: 1

...
It is clear that you think you know more that you actually do. This forum is not a haven for those sorts of people. Please learn what this forum is about before posting any more rubbish.
...


Form time to time, I do make mistake (above average) as every one does. This is forum for the place to learn, exchange information, help new comer as well as  refresh own knowledge. I learn a lot from my own mistake as well as an other person's one. I would not call it rubbish, for me it is gold. However so far I can't find any one could beat you at the speed, frequency, repeat same mistake (rubbish in your word) again and again  posting.  You are the winner.

Ok, step back  say 30 mA is safe level,  60 mA pulse current have same roughly effect as 30 mA will be 100 ms on and 100 ms off,  but we have 100 ms on and few minutes off. It is in safety zone.  Count flash trigger some time need high current to trigger, and only high CTR of photocoupler is not enough. It is proof at field application as well.  

Grumpy_Mike

OK so slagging some one off and then putting a strike through is fine is it, I think it is only the action of someone who not quite in control of all their faculties.

The mistake that you keep making over and over is that you think that the only thing that is limiting the maximum amount of current is the thermal considerations. So by utilising the thermal time constant of the chip you can afford an overload because it has time to cool off before the next pulse. What you are forgetting is the two other mechanisms that can cause damage due to over current that I mentioned earlier. I have discussed this with a friend of mine who is a chip designer when working on a project that needed good reliability, I am not making this up.

Even with the thermal considerations, it depends on the thermal conditions in the local area and you can not rely on a constant thermal resistance around the chip, it depends on the silicon doping levels in the local vicinity. Do you know what this is, because I don't and without knowing the sane approach is to stick with what the manufacturer recommends. Even if you do not blow the chip up immediately there can be damage done, and that damage might take a time to show. It might not even show on every device but you can measure the damage statistically.

You think you know better than the chip manufactures and are willing to over drive the chip because you are ignorant have not got their perspective on the matter.

Remember this is a forum for beginners and advice to them must be clear and simple. It is clear that you have never done any real engineering where reliability is important. If you had you would not be pushing the manufacturers limits. Absolute is what it says not to be exceeded. Have you ever thought why there is not a recommended pulse current rating for this device when there clearly is for other devices?

When ratings are designed there is a balance between getting the maximum permissible and having the maximum life of the device.

You have a responsibility here not to give advice that could cause damage to another persons equipment. This is not a forum for caviler advice, if you want to do a slap dash job then take up plumbing. ( whit apologises to good plumbers )

AmbiLobe

Dear sonny yu,
You wrote, "Ok, step back  say 30 mA is safe level,  60 mA pulse current have same roughly effect as 30 mA will be 100 ms on and 100 ms off,  but we have 100 ms on and few minutes off. It is in safety zone.  Count flash trigger some time need high current to trigger, and only high CTR of photocoupler is not enough. It is proof at field application as well.  "

Wrong. Please do not lie to us. Some of us know a vast wisdom about electronics. You are wrong. You should stop lying to people here. Someone might believe your garbage.
I am going to get going.

sonnyyu

#25
Jul 01, 2013, 08:26 pm Last Edit: Jul 01, 2013, 08:35 pm by sonnyyu Reason: 1

...
Wrong. Please do not lie to us. Some of us know a vast wisdom about electronics. You are wrong. You should stop lying to people here. Someone might believe your garbage.


wrong...lie...wrong...lying ...garbage.

kindergarten speech?

Nick Gammon

Back to technical talk ...

Considering that manufacturers in general like to "talk up" their stuff, I'm sure that if the chip designers thought you could drive 60 mA intermittently (eg. 100 mS on, 1 minute off) then they would say so in the data sheets.

The data sheet generally lists test conditions of 20 mA, and then list 40 mA as an absolute maximum rating. I think the word absolute is pretty clear. It doesn't mean "you can exceed this provided it is for a short time".

The data sheet goes on (page 313):

Quote

This is a stress rating only and functional operation of the device at these or other conditions beyond those indicated in the operational sections of this specification is not implied.


The "operational specifications" effectively list 20 mA as the test current, not 40 mA.

I read the specs as saying: Design for 20 mA, it can survive 40 mA, after that all bets are off.
http://www.gammon.com.au/electronics

Go Up