LV/5V level shifting using 74HC gates

ent

ve

to

A grounded emitter npn? several hundred ohms in the base and a few k ohms on the collector. Take the output from the collector... this inverts things. You may need a schottky diode from base to collector. (Baker clamp)

George H.

ent

ve

to

Vihmin for a 74hc04 at 5V is ~3.5V so ..

you can get voltage trnaslators s with separate supplys for each side e.g. 74LVX3245

or you can use a 74hct04, Vihmin is 2.0V

-Lasse

HCT devices are good for that--TTL went away, so 3.3->5V level shifting is the only reason they still exist in any quantity. Alternatively, you can put a diode in series with the HC's VDD pin to drop its supply to

Cheers

Phil Hobbs

Right. With Vcc=5 and Vin=3.3, expect shoot-through currents in the sub-mA range for an HCT04.

Bruce Varley schrieb:

Hello,

it is not a good idea to feed a 74HC04 with a high signal of 3.3 V or even less. If the HC04 has a 5 V supply, the high level on input should be close to 5 V. A 74LS04 should be compatible with the logic levels of the LV signals, a 74HCT04 should be possible too. Just compare the output levels of the LV signal with the input specification.

Bye

-OR-

Page 2 of this:

-OR-

-OR Simply-

I follow the opposite levels.

Vcc on the 74HCT04 gets 3.3 volts and the outputs go to the lower voltage side.

The first line of this data sheet states:

"The inputs include clamp diodes that enable the use of current limiting resistors to interface inputs to voltages in excess of VCC."

A simple series resistor between the +5V output from a microcontroller is all that's needed.

On the other side 3.3V output into the input of a microcontroller is more then high enough.

hamilton

So for really-o truly-o _TTL_ levels, you're already fine.

If you want 5V CMOS, of course, you're screwed.

(Jumping on the bandwagon)

Avoiding SMD isn't always possible, of course; one path that hasn't been mentioned yet, since you can tolerate relatively slow rise/fall times, is a biased transistor. These are one-input/one-output devices, and will always invert the logic. UNR4211 from DigiKey is typical.

There's still leaded parts available; buy 'em quick, though! Three terminals is typical (input, output, ground); you just add a pullup resistor on the output.

The OP used mixed metaphors... "1 LSTTL load" and "feed the LV to

I posted several possible ways to address that. ...Jim Thompson

___ 5 VOLT o-----o|___|-o----------o Pull Up | | 5 V logic | |\ o o-----o-----o-| >O- Vdd 3.3 | 1200 |/ |/ o--o|___|-o----o--o| NPN | |>

|-----o----| o | | | | o---------------o | | IO | | | uC | | | | | | | | | | | | | | | |----------- -

Something like that?

Jamie

Cascodes are an excellent way to level shift; they work for signals and power (extending the voltage range of an already sufficiently high current switch), and they tend to go fast. If the base is biased inbetween rails, the logic level also provides turn-off current. PNPs work great for below-ground work, etc.

Tim

-- Deep Friar: a very philosophical monk. Website:

"Jamie" wrote >

esent

I've

by

ge

ed

er to

=A0 |

=A0 =A0 =A0| =A0 =A0 =A05 V logic

=A0 =A0 =A0| =A0 =A0 =A0 |\

--o-----o-| >O-

=A0 =A0|/

It works, but for a simple 3.3-5 translation it would use more power and be slower than a HCT gate, NXP spec max 590uA per pin at Vin =3D Vcc -

-Lasse

CMOS 3.3V LVC can drive 1 U.L. LSTTL directly, no translator required. Is that simplified enough for you?

simplified enough for you?

Fred, You missed...

"...arrangement is to feed the LV to 74HC04 inputs, with a 5v supply."

OP has very loose understanding of LSTTL vs HC and HCT.

In...

Date: Mon, 27 Aug 2012 09:52:23 -0700 Message-ID:

I posted multiple ways to attack the problem. ...Jim Thompson

that simplified enough for you?

NXP specs HCT, max 590uA per pin @ Vin = Vcc - 2.1V

first lots of parts

last two will use more current than HCT at wc

-Lasse

present

to

that simplified enough for you?

Sounded to me like the 74HC04 was his cut at level translating to drive an ultimate load of LSTTL. There's no way a HC anything looks like LSTTL loading.

present

I've

to

that simplified enough for you?

ultimate load of LSTTL. There's no way a HC anything looks like LSTTL loading.

I'm not sure _what_ the OP really has ;-)

In my ASIC world I've been doing the device-level equivalent of...

Page 1 of this:

Dating back to my (patented) USB work with Intel... ~18 years ago. ...Jim Thompson

The HC wont work but the HCT will as long as he can get his rise/fall times in the In...