i understand what you meant but i can even produced 20mA high output current using a very simple computation.. its all on the voltage output high even disregarding the -0.4mA..

For example, if the voltage output high is 3V and you have to drive a led which has a voltage drop of 2V and you want 20mA to flow through it you'll just have to add a resistor and then apply the ohms law to determine the resistane that will allow 20mA to flow (assuming from output pin to led and resistor in series to ground).. R=[(VOH - Vled)/20mA] = 50 ohms.. and that's it.. a 20mA flowing through your circuit in series..

but what im really trying to ask is can LS TTL drive a LED.. but in real life and in practice it can..

and its not a coincidence for absf... its possible..

 

That this current is the same as Ioh

Allen

Yeah, I would say it's just soincidence.

 

try to implement the circuit i have posted in this forum.. and tell me what can you say.. try to measure high level output voltage and IOH..even Whbann agree with the current output depending on the leds drop voltage and resistance..

 

WHAT IS YOUR POINT?

SO WHAT if the gate is able to deliver more current that it is spec'ed at? Of COURSE if can! No manufacturer in their right mind would do it any other way!

And of course the actual current drawn depends on the actual load! How could it be any other way? These are NOT constant current outputs.

So what is your point with all this?

 

sorry i think there is a misunderstanding... i just want to ask if LS TTL can drive one single LED by source current using one outpin pin when logic is high..?

AS i have been telling everyone here that LS TTL accdg to some references that it cannot drive LED through current source when high but in fact in an actual implementation it can..

i just want to ask, what can an expert say about this.. should i start connecting transistor this time just to drive one led or should i just take advantage of its sinking current to its outpin pin and disregard the idea that it can light up a led using source current..? or whatever? i just need clarification, information, proof.. no offense though..

 

Well you tried it and it does work. What more proof do you need?

However, since a TTL output is better at sinking current, one usually puts the LED and resistor on the high side and output a LOW LEVEL to turn on the LED.

 

Any source that you have that categorically states that it cannot drive an LED just because the spec sheet shows that it is not guaranteed to be able to is wrong.

The spec sheet makes no claim in this regard.

Now, anyone that makes a claim that it CAN drive an LED is just as wrong because the fact that you can find 100 gates that can drive an LED is no proof that the 101st gate can, too. It might be a strong indicator, but what you also have to take into account is how much of the life have you taken off that part by abusing it that way?

Any design you make that relies on any component performing beyond its rated specs is a bad design. Now, if you want to abuse components in your private little electronic torture chamber, be my guest. We all (probably all) have done it, do do it, and will continue to do it. But you should develop the habit of not abusing parts without a damn good reason -- and damn good reasons exist and whether a particular reason is "good enough" depends on the context.

 

wow amazing.. i just love your explanation.. i have been reading this many times and i have realize, i have been wrong since then.. i guess its time to correct things up..

thank you very much for your patience, i was just enlightened thanks to you man.. sorry for the stubborn arguments, its just that ive been blind.. maybe i was one of those people who are doing things blindly..

 

Now, anyone that makes a claim that it CAN drive an LED is just as wrong because the fact that you can find 100 gates that can drive an LED is no proof that the 101st gate can, too. It might be a strong indicator, but what you also have to take into account is how much of the life have you taken off that part by abusing it that way?
.

so stubborn of me believing the fact that i had found 1000 gates or more can drive led without thinking that 1001 or 9001 cannot.. thanks for opening my eyes.. your so cool!

 

wow amazing.. i just love your explanation.. i have been reading this many times and i have realize, i have been wrong since then.. i guess its time to correct things up..

thank you very much for your patience, i was just enlightened thanks to you man.. sorry for the stubborn arguments, its just that ive been blind.. maybe i was one of those people who are doing things blindly..

No need to apologize -- stubborness is generally a good trait in an engineer.

I tend to be the same way and all the way back in junior high I had a habit that really annoyed some of my teachers. If I didn't understand something and couldn't get it explained quickly in a manner that allowed me to understand it, then (without realizing it for many years) I would shift to an adversarial role in which I would defend my understanding of the topic, even though I KNEW I was wrong, and force the other person to prove I was wrong while I tried to use MY understanding to "prove" them wrong. Somewhere in this process one of us would finally hit on the key little concept that had been eluding me, at which point I would happily and cheerfully pronounce, "THANK YOU! I see the light and completely agree with you!"

But, being a socially myopic engineer-in-the-wings, it took me a long time to realize that most people don't appreciate this approach, though most other techies can tolerate it if you make it explicitly clear that this is the approach you are using. So I have gotten fairly good at detecting when this is the case (and it is a LOT easier to do in person than on a forum!) and stopping the conversation and saying something like, "Look, I know I'm wrong. I'm not seeing something and that is keeping me from grasping this, but I can't seem to see what that something is. What I would like to do is pretend that my flawed thinking is actually correct and try to defend my understanding and attack yours. If you will go along with this little courtroom drama, there is a good chance that, at some point, one of us will reveal the tidbit that is escaping me."

 

so stubborn of me believing the fact that i had found 1000 gates or more can drive led without thinking that 1001 or 9001 cannot.. thanks for opening my eyes.. your so cool!

And it probably would -- but sooner or later you will probably find one that won't and, when you do, you can't make any complaint about how that one "doesn't work right". As long as it meets spec, it is working right.

I am occasionally confronted with some variant on the following common statistics problem:

You toss a fair coin a thousand times and each time it comes up heads. What are the odds that it will come up heads on the next toss?

The answer they are always looking for is that the odds are 50/50 and the point they are almost always trying to make is that each toss is independent and, therefore, unaffected by the prior history of outcomes.

Surpisingly, the most common wrong answer is that it is more likely to come up tails next time because, on average, it has to balance out to 50/50 and so the next thousand tosses must come up mostly tails. This is an amazingly common mispercetion and you see it manifest itself in strange ways in real life all the time.

My answer is that I expect it to come up heads because the evidence on hand overwhelmingly argues that it is a "fair" two-headed coin!

That does not always go over well (but sometimes it does).

 

so stubborn of me believing the fact that i had found 1000 gates or more can drive led without thinking that 1001 or 9001 cannot.. thanks for opening my eyes.. your so cool!

My friend is much more stubborn than you. He has a design for lighting 24 LEDs with 16F877A for a marine telegraph system. He just connected the 24 LEDs directly to the PIC outputs without any current limiting resistors. I told him so many times that the resistors were needed and he kept arguing that the MOSFET inside the PIC would self limit the current to <25mA.

I was too lazy to argue with him. I am just waiting for the day that he will finally hit the brick wall.

Allen

 

thank you guys for sharing your thoughts.. im happy to see the light... thanks for the realization..

keep on sharing your expertise..

 

hello guys i came across with this notes from the webpage

http://www.allaboutcircuits.com/worksheets/ttl.html

here it goes:
=================================================================
Notes:
It should be noted that totem-pole TTL gates can actually source far more current than what is advertised without sustaining damage. The severe limitation on sourcing current is more a function of staying within the permitted output voltage margins for TTL than it is a function of chip heating. Thus, you may generally use a totem-pole TTL gate to source 20 mA to an LED without harm, though the "high" state output voltage (when the LED is lit) will be significantly below the acceptable threshold for a TTL gate input.
=================================================================
its from question 13.. is this true?

 

hello guys i came across with this notes from the webpage

http://www.allaboutcircuits.com/worksheets/ttl.html

here it goes:
=================================================================
Notes:
It should be noted that totem-pole TTL gates can actually source far more current than what is advertised without sustaining damage. The severe limitation on sourcing current is more a function of staying within the permitted output voltage margins for TTL than it is a function of chip heating. Thus, you may generally use a totem-pole TTL gate to source 20 mA to an LED without harm, though the "high" state output voltage (when the LED is lit) will be significantly below the acceptable threshold for a TTL gate input.
=================================================================
its from question 13.. is this true?

How is this any different from what has been stated by several people in several ways in this thread? What more are you looking for?

Yes, TTL (and all logic gates) can source/sink more current than the specs advertize. But this is because the specs are a guarantee of the ability to deliver that current while maintaining a specified voltage. So of course you can draw more than that from them -- but it is not guaranteed!

Most straight TTL gates can probably source 20mA if you don't care about the voltage level that results. But you may not be able to get ALL of the gates in a given package to do so at the same time. But so what? It is BAD DESIGN to do this in the first place! In the words of Ian Malcolm, you are so busy asking if you could that you are not bothering to ask if you should.

 

yes i understand.. beyond the specs the device does not guarantee to deliver..

i am not trying to argue here.. as you can see in this thread.. the case was already been solved.. i just want to share and ask comments from veterans regarding this matter specially the tutorial came from the same website and the man that noted this was very sure though.. nothing else..

and i know its a bad design...