Hi, I have made a little project with TTL gates and flip flops. I am using a dc power supply as the power source. The dc adapter is rated as 5V, but actually outputs a little higher. I have connected it to my project, and it outputs 5.64V. I am wondering if the source voltage has to be exactly 5V or I could get away a little higher. I know TTL is sensitive to this.

This is what the data sheet for the 7474 D flip flop says. The other chips have very similar characteristics.

ABSOLUTE MAXIMUM RATINGS

VCC Supply voltage –0.5 to +7.0 V

RECOMMENDED OPERATING CONDITIONS

MAX 5.5V

From your experience, is 5.64V supply voltage safe for TTL chips, regardless of what the data sheet says?

 

Your on the edge of reliability, but if you place a diode on the source you could hit 5.00V just perfectly.

 

The Vcc range for 74xx series is 4.5v to 5.5v.

If you operate the ICs outside of that Vcc range, don't expect them to perform to their guaranteed specifications.

Many DC adapters aka "plug supplies" aka "wall warts" are not regulated; they have a specification like "5v @ 500mA" or similar; which means that when the load is 500mA, the nominal output voltage will be 5v.

You can use a simple diode in series with the positive supply to reduce the adapters' voltage by 0.6 to 0.7v. A 1N400x series diode should work OK for up to a 1A load.

Make certain to use a reasonably large capacitor after the diode to provide a reasonably constant 5v. Plug supplies typically have pretty small capacitors in them, if at all.

You will also need a 0.1uF/100nF bypass cap across Vcc/GND on each and every 74xx series IC, otherwise you will have odd problems. See the sticky thread.

 

I was going to use a diode anyway. Thanks. As far as the bypass capacitors on Vcc and gnd of each 74XX chip, I didn't do that. I'm using about 15 of these chips.

My design works perfectly so far. I tested it many times. If I don't use the caps, will I have problems later? I soldered my design on a pcb board already. Everything is running perfectly.

 

I gave up on old high power TTL about 35 years ago when I discovered low power Cmos logic ICs. Ordinary CD4xxx Cmos ICs work from an unregulated supply from 3V to 18V.

 

I was going to use a diode anyway. Thanks. As far as the bypass capacitors on Vcc and gnd of each 74XX chip, I didn't do that. I'm using about 15 of these chips.

Then you need to add 15 0.1uF/100nF bypass capacitors, one per IC.

My design works perfectly so far. I tested it many times. If I don't use the caps, will I have problems later?

Without the required bypass capacitors, the odds are not in your favor.

I soldered my design on a pcb board already. Everything is running perfectly.

If this is for an assignment, your instructor/professor should reduce your score for omitting the required bypass capacitors. Had you done such a thing for a project at a company, they might just show you the door.

Bypass capacitors are not optional.

 

About the bypass capacitors. This is my own personal project, not for a class. Also, hey I'm still learning right. We used many 74XX chips in class before and they never told us to add bypass capacitors. You guys know many things from experience that they don't teach us in school. That's why I ask you.

 

It seems that the teachers of electronics don't know anything about electronics.

 

SgtWookie you mean like this?

By the way, the clock of my logic is 1hz. I think the frequency is too small for this to make a real difference.

 

The bypass cap does not need to span the chip. It does need to be close to pin 14/16. On a PCB you make a connection to the ground plane.

 

By the way, the clock of my logic is 1hz. I think the frequency is too small for this to make a real difference.

Your clock has a low frequency but the logic switches at a high frequency.

The output of a logic IC switches very quickly with a fairly high current that must come from the bypass capacitor (isolated from the power supply by the inductance of the wiring). The switching frequency is 20MHz to 60MHz even when it switches only one time per second. Without a ceramic bypass capacitor then the power supply voltage to the IC bounces up and down at 20MHz to 60MHz each time the output switches that messes up the logic.

 

Your clock has a low frequency but the logic switches at a high frequency.

The output of a logic IC switches very quickly with a fairly high current that must come from the bypass capacitor (isolated from the power supply by the inductance of the wiring). The switching frequency is 20MHz to 60MHz even when it switches only one time per second. Without a ceramic bypass capacitor then the power supply voltage to the IC bounces up and down at 20MHz to 60MHz each time the output switches that messes up the logic.

I am not sure I understand your explanation of what switches at 20-60Mhz in the TTL. Then again all I know is that inside these chips are BJT's configured in weird ways that I never studied before. I am actually going to take a class where we study the inside of logic chips. Can't wait for it.

Are you talking about the propagation delay between the rising edge of the clock and the time when the output reacts to the rising edge? Is this what is switching fast? The propagation delay is very small? I remember it was about 5ns.

 

When a TTL output goes high or goes low it changes extremely fast which is a very high frequency transient. The output current is also fairly high which causes the supply voltage to jump up and down for a while at 20-60MHz if there is no high frequency bypass capacitor, which messes up the logic.

 

About the bypass capacitors. This is my own personal project, not for a class. Also, hey I'm still learning right. We used many 74XX chips in class before and they never told us to add bypass capacitors. You guys know many things from experience that they don't teach us in school. That's why I ask you.

Interesting video of Dave from eevblog removing every bypass cap on a single board CPU with almost no ill effects. Various factors probably at least as important like ground path. Perhaps for TTL it's more important but HC HCT probably more forgiving.

 

Welcome to AAC!

Did you realize that this thread is over 10 years old?

Interesting video of Dave from eevblog removing every bypass cap on a single board CPU with almost no ill effects.

I don't believe a lot of things people on the internet say. You can't say for certain that removing bypass caps won't have any effect. Caps aren't free and putting them on boards isn't free either. Manufacturers do that to insure reliable operation.

Perhaps for TTL it's more important but HC HCT probably more forgiving.

Actually it's just the opposite. CMOS outputs have a brief time when the P and N channel MOSFETs are both conducting. The more outputs you have switching, the more spikes you're going to be putting on the power rail.

I prototype a lot of circuits on solderless breadboards and I'm pretty cavalier about using bypass caps. Whenever I commit a circuit to a PCB, I'm more diligent about adding bypass caps.

I worked for a microprocessor company and they devoted a significant amount of area to decoupling caps. I even wrote a program that inserted MIM caps on the top metal layer wherever they'd fit after a design was completed. I see MIM caps still being mentioned in semiconductor news, so some companies must still be doing it.

 

A great deal depends on the nature of the circuit in question. A great deal has changed in 50 years, but supply droop and ground bounce are still with us despite the advances in silicon technology. Doing something for grins and giggles as a hobby project is vastly different than designing something you want to mass produce. Think about the difference between a go-kart and an automobile.