Two IC's outputting at the same time: Damage or Not?

[PauloConstantino]

Howdy pirates,

A question has come to mind about two IC's of the same logic family and brand that would have their outputs connected together. Let's say for example the output of an OR gates is connected together with that of an AND gate of say the 74HC logic family.

Looking at the datasheets it says that the output impedances are symmetrical, so IOH == IOL, so the sourcing current is the same as the sinking current. In this case, there shouldn't be problems right? If the maximum output current given by a IC is 20mA, an the maximum sinking current is 20mA, then what problems could emerge if you keep the same logic family and brand?

Then if you change the logic family or brand, but IOH still is IOL, could problems emerge? In other words, would any IC's get toasted for breakfast ?

[PauloConstantino]

Howdy pirates,

A question has come to mind about two IC's of the same logic family and brand that would have their outputs connected together. Let's say for example the output of an OR gates is connected together with that of an AND gate of say the 74HC logic family.

Looking at the datasheets it says that the output impedances are symmetrical, so IOH == IOL, so the sourcing current is the same as the sinking current. In this case, there shouldn't be problems right? If the maximum output current given by a IC is 20mA, an the maximum sinking current is 20mA, then what problems could emerge if you keep the same logic family and brand?

Then if you change the logic family or brand, but IOH still is IOL, could problems emerge? In other words, would any IC's get toasted for breakfast ?

Paulo, please take a look at this: http://www.ti.com/lit/an/sdya009c/sdya009c.pdf

[MK14]

If I understand you correctly (and ignoring some exceptions, such as tri-state busses, open collector etc). You DON'T connect outputs to each other. Especially when the outputs can disagree (one is at 1, while the other is at 0).

Those currents are guaranteed minimums, NOT guaranteed maximums. So considerably more current could flow.

tl;dr
Please don't short circuit outputs together (unless they are designed to be connected that way, such as tri-state, open-collector and some other exceptions).

Controversially, some people connect outputs together to get a higher drive current, when the logic levels, are connected in such a way, that the outputs will always agree with each other. Such as multiple inverters in a hex inverter package. I.e. the inputs are also connected together.

[PauloConstantino]

If I understand you correctly (and ignoring some exceptions, such as tri-state busses, open collector etc). You DON'T connect outputs to each other. Especially when the outputs can disagree (one is at 1, while the other is at 0).

Those currents are guaranteed minimums, NOT guaranteed maximums. So considerably more current could flow.

tl;dr
Please don't short circuit outputs together (unless they are designed to be connected that way, such as tri-state, open-collector and some other exceptions).

Hi MK. I really want to connect two outputs together. Why can't I do this?

[MK14]

If I understand you correctly (and ignoring some exceptions, such as tri-state busses, open collector etc). You DON'T connect outputs to each other. Especially when the outputs can disagree (one is at 1, while the other is at 0).

Those currents are guaranteed minimums, NOT guaranteed maximums. So considerably more current could flow.

tl;dr
Please don't short circuit outputs together (unless they are designed to be connected that way, such as tri-state, open-collector and some other exceptions).

Hi MK. I really want to connect two outputs together. Why can't I do this?

You would be basically creating a near short circuit, between the two power rails. It may well also overload/overheat the chips. Especially if lots of outputs are connected that way.

You are free to do anything you like. In the same way the electronics is free to let out the magic smoke. At anytime.

[alm]

Two outputs connected together with opposite value is what can cause a glitch. This could be a hell to debug back when people still used simple logic gates on large numbers. Why would you intentionally create this?

If you somehow must do this, include series resistors for both outputs to limit the current to some sensible level.

[PauloConstantino]

Because the maximum output/input currents in my IC's are 20mA, so where could the problem be ?

[MK14]

Because the maximum output/input currents in my IC's are 20mA, so where could the problem be ?

When the datasheet says MAXIMUM output current.
It means the MINIMUM current that it is guaranteed to produce. It can and probably will allow MORE than that to flow.

Some chips datasheets, allow one output to be shorted to a power rail.

Analogy:
If I said you can borrow a MAXIMUM of £20 from me.
It means that you can borrow between £0 and £20.

BUT I may have a lot more money on me. E.g. £100.

So me saying a MAXIMUM of £20 to lend to you, does NOT mean if I am robbed, I only lose £20. I would actually lose £100.

[PauloConstantino]

Because the maximum output/input currents in my IC's are 20mA, so where could the problem be ?

When the datasheet says MAXIMUM output current.
It means the MINIMUM current that it is guaranteed to produce. It can and probably will allow MORE than that to flow.

Some chips datasheets, allow one output to be shorted to a power rail.

Look, how can MAXIMUM mean MINIMUM? Are you ok lad?

It's the MAX output current not the MINIMUM.


I am looking at the ABSOLUTE MAXIMUM RATINGS.

Ok mate ?

[MK14]

Because the maximum output/input currents in my IC's are 20mA, so where could the problem be ?

When the datasheet says MAXIMUM output current.
It means the MINIMUM current that it is guaranteed to produce. It can and probably will allow MORE than that to flow.

Some chips datasheets, allow one output to be shorted to a power rail.

Look, how can MAXIMUM mean MINIMUM? Are you ok lad?

It's the MAX output current not the MINIMUM.


I am looking at the ABSOLUTE MAXIMUM RATINGS.

Ok mate ?

You have got it the WRONG way round. Sorry.

They mean the MAXIMUM safe value. In practice a lot more current can flow if there is a short circuit.

[Rerouter]

That is the maximum you can feed in to / out of those pins without damaging then, it is not the short circuit current which can be much higher,

a mosfet push pull stage may appear as 1 ohm to voltage rail when switched on, now on a 5V device that will be way more than that 20mA limit. and puff goes some piece of silicon

If you wish to have the 2 connected fit a resistor inbetween the 2 so that should they output opposite signals it is still within this limit. e,g, for a 5V device you would fit a 470 ohm resistor between the 2 pins, this limits both to a maximum fault current of 10ma (Its unwise to live on the maximum ratings), and both pins should remain safe.

[MK14]

A MOTORWAY has an ABSOLUTE MAXIMUM speed limit of 70 MPH, in the UK.

BUT, you can drive at 200 MPH if you want (to lose your driving licence).

[sokoloff]

Look, how can MAXIMUM mean MINIMUM? Are you ok lad?

It's the MAX output current not the MINIMUM.


I am looking at the ABSOLUTE MAXIMUM RATINGS.

Ok mate ?

MK14 has it right. The datasheet is telling you that you can rely on sourcing (or sinking) that amount of current without damage to the chip or misbehavior of the circuit.

It is not telling you that if you tie that pin to ground via a multimeter set on Amperes (so a direct short) and drive it high that the meter will read 20mA (as if the pin were a constant current supply).

[Rerouter]

To try and clear things up, a logic IC has powerful mosfet switches to ensure a fast transistion time. but very thin bond wire because your not expected to have high currents except for at those transitions. if you constantly pull a large amount of current through them you can damage those bond wires, this is what sets the maximum

So while the mosfets may be able to switch 2A, the wire will melt if much more than 20mA flows through it

[PauloConstantino]

Look, how can MAXIMUM mean MINIMUM? Are you ok lad?

It's the MAX output current not the MINIMUM.


I am looking at the ABSOLUTE MAXIMUM RATINGS.

Ok mate ?

MK14 has it right. The datasheet is telling you that you can rely on sourcing (or sinking) that amount of current without damage to the chip or misbehavior of the circuit.

It is not telling you that if you tie that pin to ground via a multimeter set on Amperes (so a direct short) and drive it high that the meter will read 20mA (as if the pin were a constant current supply).

So what parameter defines the maximum current an IC can provide then mate?

[sokoloff]

Take two deep breaths and read rerouter's explanation.

[PauloConstantino]

Ok I was wrong. I thought the maximum ratings were the most current the IC could provide. Apologies and thanks for correcting me. 


So what defines the maximum current possible?

[MK14]

So what parameter defines the maximum current an IC can provide then mate?

If you buy a battery, and it says on it 6V at a maximum of 1 Amp.

It means it provides 6V, and you can use up to 1 Amp from it. Safely WITHOUT it overheating or something.

But if you short-circuit it. Massively more than 1 Amp may flow. E.g. 8.5 Amps.

If the Battery datasheet says Absolute Maximum Current (taken out of) battery is 1 Amp.
It does NOT mean it can NEVER give more than 1 Amp if short-circuited.

I.e. It can take up to 1 Amp, WITHOUT damaging it.

[PauloConstantino]

I agree it's the maximum allowable current before damage, but it's nothing to do with MINIMUM !!! MK has brought this minimum business up. There's no minimum here. It's maximum allowable current before malfunction.

[sokoloff]

I agree it's the maximum allowable current before damage, but it's nothing to do with MINIMUM !!! MK has brought this minimum business up. There's no minimum here. It's maximum allowable current before malfunction.

No, it's the minimum that any given chip (with the inherent manufacturing tolerances) can supply without damage. Some chips will necessarily be able to supply more than that without damage, but the manufacturer does not claim that.

It's like a tire speed rating. The tire speed rating is a safety rating. It does not guarantee that your H rated tire will come apart at 131 mph. It merely guarantees that the worst H rated tire will be safe at 130 mph. That is the sense in which it is a minimum rating.

[Rerouter]

Ok, defining how much the internal switch can actually handle. for that look at the output low and output high specs at a few mA

e.g. https://assets.nexperia.com/documents/data-sheet/74HC_HCT08.pdf

Bottom of page 4,

So for a 6V supply, the typical switch resistance is in the ballpark of 50 ohms (thats something i will remember next time i go high speed),
(6V - Out High (5.81V)) / 4mA load = 47.5 ohms, so as a hard short it should be able to sink 126mA or atleast in that ballpark, and be dissipating 0.76W in that one output, which will not last long,

[PauloConstantino]

So what parameter defines the maximum current an IC can provide then mate?

If you buy a battery, and it says on it 6V at a maximum of 1 Amp.

It means it provides 6V, and you can use up to 1 Amp from it. Safely WITHOUT it overheating or something.

But if you short-circuit it. Massively more than 1 Amp may flow. E.g. 8.5 Amps.

If the Battery datasheet says Absolute Maximum Current (taken out of) battery is 1 Amp.
It does NOT mean it can NEVER give more than 1 Amp if short-circuited.

I.e. It can take up to 1 Amp, WITHOUT damaging it.

OK agree. But why did you bring up MINIMUM ratings? It is the minimum current at which it will break down, ok. But it's also the maximum before breakdown. So it's both the minimum and the maximum before breakdown. Why bother with minimum? You bright this up. There's no need to speak of minimums. It already says maximum. I think you were just trying to contradict me again.

[sokoloff]

OK agree. But why did you bring up MINIMUM ratings? It is the minimum current at which it will break down, ok. But it's also the maximum before breakdown. So it's both the minimum and the maximum before breakdown. Why bother with minimum? You bright this up. There's no need to speak of minimums. It already says maximum. I think you were just trying to contradict me again.

He was trying to break through your confusion in this post:

Because the maximum output/input currents in my IC's are 20mA, so where could the problem be ?

You were clearly confused (at that time) about what that spec meant, and MK14 was trying his best to help you understand that that rating was a design spec guaranteed minimum, not a maximum in the way you were understanding it.

He brought up minimum in a genuine effort to try to help you. If you persist in your aggressive responses, the problem you're experiencing with people trying to help you may resolve itself.

[PauloConstantino]

I agree it's the maximum allowable current before damage, but it's nothing to do with MINIMUM !!! MK has brought this minimum business up. There's no minimum here. It's maximum allowable current before malfunction.

No, it's the minimum that any given chip (with the inherent manufacturing tolerances) can supply without damage. Some chips will necessarily be able to supply more than that without damage, but the manufacturer does not claim that.

It's like a tire speed rating. The tire speed rating is a safety rating. It does not guarantee that your H rated tire will come apart at 131 mph. It merely guarantees that the worst H rated tire will be safe at 130 mph. That is the sense in which it is a minimum rating.

"No, it's the minimum that any given chip (with the inherent manufacturing tolerances) can supply without damage. "

No, the minimum it can supply without damage is 0.

I think we are entering the domain of semantics here now. And deeply flawed logic.


The minimum current the IC can source or sink without damage is 0.

The minimum it can source or sink, (and if you go higher) damage will occur is 25mA (in this case).

The maximum it can sink or source, after which damage will occur is 25mA.



You guys are confusing the semantics and logic of minimum with that of maximum.




Anyways, I'm done with this thread now thanks.

[MK14]

So what parameter defines the maximum current an IC can provide then mate?

If you buy a battery, and it says on it 6V at a maximum of 1 Amp.

It means it provides 6V, and you can use up to 1 Amp from it. Safely WITHOUT it overheating or something.

But if you short-circuit it. Massively more than 1 Amp may flow. E.g. 8.5 Amps.

If the Battery datasheet says Absolute Maximum Current (taken out of) battery is 1 Amp.
It does NOT mean it can NEVER give more than 1 Amp if short-circuited.

I.e. It can take up to 1 Amp, WITHOUT damaging it.

OK agree. But why did you bring up MINIMUM ratings? It is the minimum current at which it will break down, ok. But it's also the maximum before breakdown. So it's both the minimum and the maximum before breakdown. Why bother with minimum? You bright this up. There's no need to speak of minimums. It already says maximum. I think you were just trying to contradict me again.

I agree with you, in that I should NOT have mentioned minimums, as it could cause confusion.

The reason I mentioned minimum, was as follows.

Let's say a power transistor, is rated for up to 1 amp current. With an absolute MAXIMUM current of 1 amp.

As regards the absolute maximum, it is a MAXIMUM of 1 Amp.

But if you accidentally short circuit the transistor, a lot more than 1 amp, may flow from it.
Hence that is why I said 1 Amp is the minimum short circuit current. What I really meant was that the 1 Amp was just the safe limit. But it could go a lot higher, if the transistor was short-circuited.

Anyway, I should have refrained from calling it minimum, as it was making it too confusing. Sorry.