Basic IC Question

[VJR85]

I am taking an introductory course to digital. We use a trainer board at school, but I don't have one at home.

I am trying to build the circuit at home to learn how to do it without the trainer board, but one thing seems wrong.

The IC is a 74LS08 with four two-input AND gates.

http://www.alldatasheet.com/view.jsp?Searchword=74L

I am just trying to run a LED off it, but if I hook the LED anode to any output gate (cathode grounded) with only the VCC and GND hooked to the IC, it still lights??

Shouldn't I have to apply voltage to both of the AND gate inputs for it to light?

[rshayes]

The circuit in a TTL gate is such that the input assumes a high state if nothing is connected to the input. A certain amount of current, usually a fraction of a milliamp, must flow to ground to bring the input terminal of the gate to the low level (below .8 volts). It appears that in your circuit the inputs are unconnected and are floating to the high level. Since both inputs are high, and it is an AND gate, the output is also properly high, and the LED will light.

If you connect one or both of the inputs to ground, the LED should go out.

Normally inputs are not left unconnected. They should either be connected directly to ground or to the positive power supply using a pull up resistor in the 1K to 5K range. Floating inputs can pick up stray signals and cause unpredictable effects.

[VJR85]

Thanks!

[haklesup]

That's right and to make matters worse, CMOS device inputs left unconnected are undetermined, they could be read in as a high or a low. Furthermore, floating inputs on a CMOS (74HCT08 for example) will also usually cause high supply current. The nice thing about a CMOS part though (other than low current) is that the outputs tend to be more powerful and can be used to source or sink current.

At least on a TTL, if you want a high input you can float it but that is not a good practice. Always define input only pins. In the future you may run into devices with pins which say they have an internal pullup or pulldown, these pins can be floated if you want the default logic it is pulled to.

[Bob Scott]

You might consider using a digital logic probe, meter, or 'scope to measure the output of the LS08. TTL is just on the borderline of being able to light an LED with a grounded cathode. The logic is not designed to directly interface with LEDs. However, the "logic high" current of the LS08 is limited to -.8mA, or just enough to give you a low LED brightness. In normal use of an indicator LED you'd drive it to 5mA.

Luckily, directly connecting the LED to the LS08 output is not likely to cause any part failure or generate any smoke.

Which school are you enrolled in that is still teaching this course? What type of certification, diploma or degree are you aiming for?

Regards,
Bob

[VJR85]

It's a local college in Texas. University of Houston.

Degree is just a bachelors in Electrical Power Technology.

[Dean Huster]

Myself, I'm delighted to see a school still teaching a digital logic course based on TTL (or CMOS). In this world of PICs and other "shortcut" chips and electronics programs being eliminated all over the country, you sure learn a lot about sourcing, sinking, pull-up, pull-down, logic, etc. -- things that are often basic questions asked by folks delving into PICs and stuff as first-time electronics.

TTL drives LEDs wonderfully if you use the active LOW output of the chip to drive the LED (anode of the LED to +5v through a 150- to 330-ohm current-limiting resistor rather than cathode to ground through same, although the LED lights up when the output goes LOW.

Dean

[bigkim100]

Myself, I'm delighted to see a school still teaching a digital logic course based on TTL (or CMOS). In this world of PICs and other "shortcut" chips and electronics programs being eliminated all over the country, you sure learn a lot about sourcing, sinking, pull-up, pull-down, logic, etc. -- things that are often basic questions asked by folks delving into PICs and stuff as first-time electronics.

I have to agree with Dean whole-heartedly. Deann has been aswering questions about electronics projects since the dawn of mankind (or something close to it)
Dean has been involved with some of the most useful, well designed,and most importantly FUN magazines in the electronics field.... ever published, that inspired most electronics enthusiasts that are around today.

[MrAl]

Hi there,

One thing you can do is use the high brightness LEDs.
These things will still give a good indication at currents
under 1ma. Even with a 4.7k resistor in series you should
have no problem seeing a high brightness LED light up
when the output goes high, but only if the output goes high
enough. You can also help the output go higher with a
resistor to +5 (maybe 2k or something).
If you are going to test circuits it would be best to use a transistor
driver to drive the LED. This way you can use a 10k resistor
to the base of an NPN to sense output state, and have the NPN
drive the LED though a 1k resistor. Should be nice and bright.