safe to drive inputs on an unpowered IC

[[email protected]]

I am trying to put together a battery-powered cicruit for our lab that
does nothing more than show the state of logic lines between two pieces
of equipment using LEDs. Since one of these is the output of an 82C55,
I need to power the LEDs separately, e.g. using a hex inverter. I've
put together this kind of thing before, but never in an application
where parts of the circuit, in this case the buffer/inverter, LEDs,
etc. would only be powered occasionally (we only need to see the state
of the logic lines occasionally, and would want to conserve battery
power in between).

So, my question is, is it safe to have the inputs of the inverter (e.g.
74HC04) tied to active logic lines when the IC itself is unpowered? If
not, is there an easy workaround (e.g. another kind of IC)?

Thanks ahead of time.

 

[[email protected]]

So, my question is, is it safe to have the inputs of the inverter (e.g.
74HC04) tied to active logic lines when the IC itself is unpowered? If
not, is there an easy workaround (e.g. another kind of IC)?

Thanks ahead of time.

Depends what you mean by safe. The unpowered IC will probably power
itself through its protection diodes leading to unexpected behavior.
Use the LCX or VHC family of logic.
eg 74vhc04 or 74lcx04

 

[Al]

I am trying to put together a battery-powered cicruit for our lab that
does nothing more than show the state of logic lines between two pieces
of equipment using LEDs. Since one of these is the output of an 82C55,
I need to power the LEDs separately, e.g. using a hex inverter. I've
put together this kind of thing before, but never in an application
where parts of the circuit, in this case the buffer/inverter, LEDs,
etc. would only be powered occasionally (we only need to see the state
of the logic lines occasionally, and would want to conserve battery
power in between).

So, my question is, is it safe to have the inputs of the inverter (e.g.
74HC04) tied to active logic lines when the IC itself is unpowered? If
not, is there an easy workaround (e.g. another kind of IC)?

Thanks ahead of time.

Possibly only if the inputs are subjected to negative spikes, i.e.,
below ground. If, for some reason, power is applied to the IC while the
input is negative, it may go into latch-up and let the smoke out.

Al

 

[Luhan]

I am trying to put together a battery-powered cicruit for our lab that
does nothing more than show the state of logic lines between two pieces
of equipment using LEDs. Since one of these is the output of an 82C55,
I need to power the LEDs separately, e.g. using a hex inverter. I've
put together this kind of thing before, but never in an application
where parts of the circuit, in this case the buffer/inverter, LEDs,
etc. would only be powered occasionally (we only need to see the state
of the logic lines occasionally, and would want to conserve battery
power in between).

So, my question is, is it safe to have the inputs of the inverter (e.g.
74HC04) tied to active logic lines when the IC itself is unpowered? If
not, is there an easy workaround (e.g. another kind of IC)?

Thanks ahead of time.

I believe a ULN2003 inverter can be safely driven whether powered or
not.

Luhan

 

[Luhan]

I believe a ULN2003 inverter can be safely driven whether powered or
not.

Correction: ULN2003 does not have any power feed. Its just a bunch of
darlingtons with some diodes for the VCC connection.

Luhan

 

[Joerg]

I am trying to put together a battery-powered cicruit for our lab that
does nothing more than show the state of logic lines between two pieces
of equipment using LEDs. Since one of these is the output of an 82C55,
I need to power the LEDs separately, e.g. using a hex inverter. I've
put together this kind of thing before, but never in an application
where parts of the circuit, in this case the buffer/inverter, LEDs,
etc. would only be powered occasionally (we only need to see the state
of the logic lines occasionally, and would want to conserve battery
power in between).

So, my question is, is it safe to have the inputs of the inverter (e.g.
74HC04) tied to active logic lines when the IC itself is unpowered? If
not, is there an easy workaround (e.g. another kind of IC)?

Thanks ahead of time.

As others have written it'll try to supply the circuit through the
inputs that are high. If you have tens of kohm in series it may be ok
but you'd have to hold down VCC if you want to avoid unpredictable
behavior of the other logic that is hanging on that rail.

The real way to do it is using transistors. With BJT the input current
goes to ground and with FETs there is no input current. When using FETs
make sure there are gate protection diode against excessive Vgs in
either direction. Either version needs series resistors, of course. But
so does your 74HC04 because if you exceed the abs max and then fire up
the circuit it would latch up.

 

[John Larkin]

I am trying to put together a battery-powered cicruit for our lab that
does nothing more than show the state of logic lines between two pieces
of equipment using LEDs. Since one of these is the output of an 82C55,
I need to power the LEDs separately, e.g. using a hex inverter. I've
put together this kind of thing before, but never in an application
where parts of the circuit, in this case the buffer/inverter, LEDs,
etc. would only be powered occasionally (we only need to see the state
of the logic lines occasionally, and would want to conserve battery
power in between).

So, my question is, is it safe to have the inputs of the inverter (e.g.
74HC04) tied to active logic lines when the IC itself is unpowered? If
not, is there an easy workaround (e.g. another kind of IC)?

Thanks ahead of time.

Look at the OnSemi NL17 series. The inputs tolerate up to 7 volts,
powered up or not.

Or, A Joerg suggests, use some 2N7002's.

John

 

[[email protected]]

Look at the OnSemi NL17 series. The inputs tolerate up to 7 volts,
powered up or not.

Or, A Joerg suggests, use some 2N7002's.

John

What's wrong with dropping in a LCX or VHC part? No added parts needed!

 

[Jim Thompson]

I am trying to put together a battery-powered cicruit for our lab that
does nothing more than show the state of logic lines between two pieces
of equipment using LEDs. Since one of these is the output of an 82C55,
I need to power the LEDs separately, e.g. using a hex inverter. I've
put together this kind of thing before, but never in an application
where parts of the circuit, in this case the buffer/inverter, LEDs,
etc. would only be powered occasionally (we only need to see the state
of the logic lines occasionally, and would want to conserve battery
power in between).

So, my question is, is it safe to have the inputs of the inverter (e.g.
74HC04) tied to active logic lines when the IC itself is unpowered? If
not, is there an easy workaround (e.g. another kind of IC)?

Thanks ahead of time.

It just occurred to me that LM339 inputs can go above rail, powered or
not.

...Jim Thompson

 

[John Popelish]

It just occurred to me that LM339 inputs can go above rail, powered or
not.

Yes, by a lot, and very handy at times. Some data sheets
make this very clear, and some hide it pretty well.

 

[Jim Thompson]

Yes, by a lot, and very handy at times. Some data sheets
make this very clear, and some hide it pretty well.

Yep. Even National's data sheet is obscure about it :-(

...Jim Thompson

 

[John Larkin]

What's wrong with dropping in a LCX or VHC part? No added parts needed!

Do they have esd diodes to Vcc?

John

 

[[email protected]]

Many, many thanks for all the info/suggestions. Time to start looking
over some datasheets.

-Dave

 

[John Larkin]

Yep. Even National's data sheet is obscure about it :-(

...Jim Thompson

They're also obscure about going *below* the rail! That's a lot more
interesting.

John

 

[Jim Thompson]

Do they have esd diodes to Vcc?

John

Some of those parts (which I redesigned for ONSemi) have "magic
floating" ESD structures ;-)

...Jim Thompson

 

[Jim Thompson]

They're also obscure about going *below* the rail! That's a lot more
interesting.

John

Naaah! They say, "Don't", but there *are* ways ;-)

...Jim Thompson

 

[[email protected]]

Do they have esd diodes to Vcc?

John

The datasheet says "input protection circuit" for power down operation.
There is no diode shown on the data sheet. I'm using an LCX part in a
project where I want to avoid the classic self powering problem and so
far, it works, which leads me to think there is no diode to VCC.

 

[Joerg]

Naaah! They say, "Don't", but there *are* ways ;-)

Abs max says -0.3V. So does note 6. So does the text on page 7. Ok,
nothing in writing about performance but as we all know...

 

[neil]

I am trying to put together a battery-powered cicruit for our lab that
does nothing more than show the state of logic lines between two pieces
of equipment using LEDs. Since one of these is the output of an 82C55,
I need to power the LEDs separately, e.g. using a hex inverter. I've
put together this kind of thing before, but never in an application
where parts of the circuit, in this case the buffer/inverter, LEDs,
etc. would only be powered occasionally (we only need to see the state
of the logic lines occasionally, and would want to conserve battery
power in between).

So, my question is, is it safe to have the inputs of the inverter (e.g.
74HC04) tied to active logic lines when the IC itself is unpowered? If
not, is there an easy workaround (e.g. another kind of IC)?

Thanks ahead of time.

May be stupid response but ....
how about keeping the 74HC04 powered all the time, and removing power to the
LEDs?
Might be ok if the switching speed isn't too great, so the 04 doesn't use
too much power internally.
hth
Neil