74HC164 shift register maximum current per pin?

The datasheet for the Texas Instruments part only says “±4-mA Output Drive at 5 V” in the description, and says nothing in the Electrical Characteristics section except to rate it for 5.34 minimum output voltage at 6.0V VCC when “IOH = –5.2 mA”. Anyway, I’m forced to conclude that the pin “can” source “up to” 5.2mA but I have no idea how hard I dare to push it before risking damage?

I want to multiplex 5mm RGB LEDs. I would like to use the 74HC164 outputs to switch the common cathodes via 8 TO-92 2N2222 NPN transistors. So, I will only turn on one transistor at a time. The transistor will switch on for up to 200uS with a 162mA maximum Ic. Traditional wisdom says I need to drive the transistor base with at least 8mA (1:20 B:E).

Think I can get away with sourcing 8mA per 74HC164 pin or do I limit myself to 5.2mA and hope the light duty cycle on the transistors will not cause them to cook themselves at 1:31 B:E?

An idea just occurred to me – I could use 8 1K? pullup resistors and make each 74HC164 output pin sink 5mA when LOW. Then when it goes HIGH, I have a few extra mA to turn the transistor on a little harder. Thoughts?

There max rating is 25mA per pin it can handle 8 mA I don't understand where they came up with the 4mA and the 5.2 at cause it tells you in the data sheet IO Continuous output current VO = 0 to VCC ±25 mA and I've used them with leds set to use 20mA with no problems at all. Now you can't turn on all 8 at 25mA but one or two is fine and at 8mA you could have 5 on easy

Thanks be80be. What manufacturer's datasheet are you looking at? I found the VO section you mention on the Texas Instruments SN74HC164N, but it doesn't list 25mA there. :(

tylernt: The datasheet for the Texas Instruments part only says "±4-mA Output Drive at 5 V" in the description, and says nothing in the Electrical Characteristics section except to rate it for 5.34 minimum output voltage at 6.0V VCC when "IOH = –5.2 mA". Anyway,

I'm guessing 4mA at 5V is equivalent to 5.2mA at 6V (Ohm's law, rounding 4.3mA to 4mA)

tylernt: I'm forced to conclude that the pin "can" source "up to" 5.2mA

...except you're not running at 6V.

The absolute max is 25mA for a single pin, the absolute max for the supply or ground is 50mA, so if you drive all 8 outputs your absolute max per pin is 6.25mA. 4mA seems like a sensible nominal maximum to me.

tylernt: Thanks be80be. What manufacturer's datasheet are you looking at? I found the VO section you mention on the Texas Instruments SN74HC164N, but it doesn't list 25mA there. :(

http://www.ti.com/lit/ds/symlink/sn74hc164.pdf

page 3

Maximum output ratings

IO Continuous output current VO = 0 to VCC ±25 mA

You might want to try changing your IC to CD74ACT164E.

dave-in-nj: http://www.ti.com/lit/ds/symlink/sn74hc164.pdf

page 3

Maximum output ratings

IO Continuous output current VO = 0 to VCC ±25 mA

Cool. I guess I completely skipped the Absolute Maximums sections because I was more interested in the Recommended Conditions, but in this case it made me look rather foolish. :drooling_face: I guess even 16mA should be no big deal, being only 64% of max.

Thanks all.

There's also this part, which is an open-drain version good for up to 50mA I believe. http://www.ti.com/lit/ds/symlink/tlc6c598-q1.pdf DigiKey has them for $1.09 each.

I guess even 16mA should be no big deal, being only 64% of max.

That'll only work for only up to 3 outputs. Otherwise, you'll exceed the maximum total current the chip can handle (±50mA). The CD74ACT164E can handle a total max current of ±100mA, so it could ±10mA on all 8 outputs (80% of max for the IC).

Roger that. I'm strobing the rows of a multiplexed LED grid, so only one pin at a time will ever be HIGH.

Thanks rmetzner49 and dlloyd for the pointers to better chips.

<<<I want to multiplex 5mm RGB LEDs. I would like to use the 74HC164 outputs to switch the common cathodes via 8 TO-92 2N2222 NPN transistors. So, I will only turn on one transistor at a time. The transistor will switch on for up to 200uS with a 162mA maximum Ic. Traditional wisdom says I need to drive the transistor base with at least 8mA (1:20 B:E). >>>

Why use http://www.ti.com/lit/ds/symlink/tlc6c598-q1.pdf when tylernt needs 162 mA?
TPIC6A595 can sink at least 300 mA.

Pelle

Already have the 74HC164 and NPNs in hand... and already over-budget on hobby purchases. ;)