Is it OK to have output clamp diod conducting in 74HC07?

[ark1]

Hi,

Well, a simple problem. I have a design with all logic at Vdd = 3.3V. I
need then to control a number of high side PNP transistors powered from
12V rail. I use currently NPN transistors for that purpose (bases of
PNP transistors being connected through resistors to collectors of
NPN's).

I would like to simplify this and replace all NPN's with one chip of
open collector drivers like 74xx07.

The problem is I cant use TTL part as I have VDD only 3.3V, and for
CMOS variant the problem is that they are specified for Vout no more
than Vdd (they have clamping diodes to Vdd at they output pins).

My question is whether it is OK to use these parts somewhat beyond the
recommended conditions, connecting output collectors through resistors
to say 12V (to bases of PNP transistors in my case). It seems I am
still safe since the diodes will not allow Vout to rise above Vdd +
0.5V which is still within "Absolute maximum ratings" of the part.

Is there any better solution?

TIA

-Arkady-

 

[Spehro Pefhany]

Hi,

Well, a simple problem. I have a design with all logic at Vdd = 3.3V. I
need then to control a number of high side PNP transistors powered from
12V rail. I use currently NPN transistors for that purpose (bases of
PNP transistors being connected through resistors to collectors of
NPN's).

I would like to simplify this and replace all NPN's with one chip of
open collector drivers like 74xx07.

The problem is I cant use TTL part as I have VDD only 3.3V, and for
CMOS variant the problem is that they are specified for Vout no more
than Vdd (they have clamping diodes to Vdd at they output pins).

My question is whether it is OK to use these parts somewhat beyond the
recommended conditions, connecting output collectors through resistors
to say 12V (to bases of PNP transistors in my case). It seems I am
still safe since the diodes will not allow Vout to rise above Vdd +
0.5V which is still within "Absolute maximum ratings" of the part.

Then the current will flow through the protection diodes and turn your
PNPs partially on.

Is there any better solution?

Why not a jellybean ULN2003/2803? Or if you just need a few, there
are dual NPNs with base resistors built in in tiny packages. Or use
source drivers and get rid of all your discrete transistors.


Best regards,
Spehro Pefhany

 

[Klaus Kragelund]

ark1 skrev:

Hi,

Well, a simple problem. I have a design with all logic at Vdd = 3.3V. I
need then to control a number of high side PNP transistors powered from
12V rail. I use currently NPN transistors for that purpose (bases of
PNP transistors being connected through resistors to collectors of
NPN's).

I would like to simplify this and replace all NPN's with one chip of
open collector drivers like 74xx07.

The problem is I cant use TTL part as I have VDD only 3.3V, and for
CMOS variant the problem is that they are specified for Vout no more
than Vdd (they have clamping diodes to Vdd at they output pins).

My question is whether it is OK to use these parts somewhat beyond the
recommended conditions, connecting output collectors through resistors
to say 12V (to bases of PNP transistors in my case). It seems I am
still safe since the diodes will not allow Vout to rise above Vdd +
0.5V which is still within "Absolute maximum ratings" of the part.

Is there any better solution?

With the 4000 logic series it is possible to drive current in the
protection diodes. Just a couple of days ago we did some tests on the
CD4093 and the current through the diode in the input (causing an
simelar die current) was over 20mA before the supply collapsed and the
function of the gate failed due to SCR latchup.

But, we did opt to do a work-around instead of allowing current through
the diodes. Mainly because the manufactors does not specify the allowed
current and the latest datasheets specify the input voltage to the chip
to be between the GND/VDD rails and not as in the older days when they
specified 0.5V in excess of the supply rails. (the old 4000 series
datasheets has lots more info than the newer ones)

Regards

Klaus

 

[John B]

Hi,

Well, a simple problem. I have a design with all logic at Vdd = 3.3V.
I need then to control a number of high side PNP transistors powered
from 12V rail. I use currently NPN transistors for that purpose
(bases of PNP transistors being connected through resistors to
collectors of NPN's).

I would like to simplify this and replace all NPN's with one chip of
open collector drivers like 74xx07.

The problem is I cant use TTL part as I have VDD only 3.3V, and for
CMOS variant the problem is that they are specified for Vout no more
than Vdd (they have clamping diodes to Vdd at they output pins).

My question is whether it is OK to use these parts somewhat beyond the
recommended conditions, connecting output collectors through resistors
to say 12V (to bases of PNP transistors in my case). It seems I am
still safe since the diodes will not allow Vout to rise above Vdd +
0.5V which is still within "Absolute maximum ratings" of the part.

Is there any better solution?

TIA

-Arkady-

UDN2981 or TD62783

 

[Mark]

or use the gate to turn on NPNs and the collector of the NPN pulls down
on the base of the PNP....ie you make your own open collector
outputs... you might need that to get enough current gain anyway,,,how
much current are the high side PNPs carrying?

Mark

 

[ark1]

or use the gate to turn on NPNs and the collector of the NPN pulls down
on the base of the PNP....ie you make your own open collector
outputs...

This is exactly the configuration I have now. What I want is to avoid
using discrete components.
I think I'll try one of the source drivers ICs as was suggested by
others.

-Arkady-