Hi all,

I'm a little new here so bare with me. I'm currently working on an industrial PLC project. I am planning to power the PLC (5VDC), Sensors(5V & 6VDC), with a power supply. I have a good 30W Power supply that is variable between 5V, 6V, 7.5V, and 9V. My plan was to use a 7805 Voltage regulator, and a 7806 Voltage regulator both paired directly to the power supply to achieve my voltages needed for all my components (with filter capacitors of course).

Would this be better then two power supply's? Or I guess my bigger question is what is the average expected life expectancy (with proper heat sinking) of the 7805 and 7806.

I've tried to do my research, but haven't seen anything on life term so far.

Thanks,

 

Welcome to AAC!

Lifetime for well designed circuits is measured in decades. How many do you need?

 

Welcome to AAC!

Lifetime for well designed circuits is measured in decades. How many do you need?

Thank you for the quick reply. I would need only a handful of each to get me started. Not really going to big just yet but as far as opinion goes, what are your thoughts? Would it be better to have two power supply's or one power supply that will provide plenty of current just regulated through two regulators.

Thinking out loud/Math
If I've done my math correctly my total draw will be: 960ma if everything fires all at once on my 5V circuit. and 1500ma (or 1.5 Amps) so i'm looking at a total of about 2.5A

My power supply can provide up to 30Watts so at 9V it will give me 3.33333 Amps of power. (This is off one power supply)

Sorry if my information is a bit unorganized.

 

I would need only a handful of each to get me started.

I meant how many decades of life expectancy you required...

Would it be better to have two power supply's or one power supply that will provide plenty of current just regulated through two regulators.

As with most things, it depends.

Using one power supply reduces expense and is probably what I'd do.

If I've done my math correctly my total draw will be: 960ma if everything fires all at once on my 5V circuit. and 1500ma (or 1.5 Amps) so i'm looking at a total of about 2.5A

How are you calculating your peak current requirement? Do you have any switching logic?

LM78xx regulators will provide up to 1A of current. Your 960mA figure is close enough to the limit that you'll probably want external pass transistors on both regulators. The datasheet will usually have an example.

 

Start with 9 volts and the 7805 dissipates nearly 4 watts. That is a heat sink situation.
The 7806 dissipates 4.5 watts. Heat sink again.
Turn it down to 7.5 volts and the 7806 is in danger of not having enough voltage to work with unless you can find a low dropout 7806.
A quick search of www.mouser.com shows nothing that fits the 6 volt position in a low dropout regulator.
You could get away without the heat sinks by using a switching regulator, but I have seen so many high speed devices go bad while my analog designs from the 1970's are still running.

What? Designs from the 1970's are still running?
Yes. That's what I said.
Build it right and the math confirms real world experience.
The most common failure today is aluminum electrolytic capacitors. Your 30 watt supply most likely has some. That's where you put your money if you want reliability. Choose capacitors rated for almost double the actual voltage they will experience and use caps rated for high temperature. I think the good number is 105 C. Install resistors rated for at least 2x the watts they will experience.

Here are some 5V regulators rated at 0ver 2 amps:
http://www.mouser.com/Power/Power-M...0w8p8Z1z0vlw7Z1z0w25cZ1yzryzwZ1z0wd5eZ1z0wbdj

On a personal note, you sound kind of half way there right now. Don't be surprised if you are on your 4th time through the design when everything fits correctly.

 

Start with 9 volts and the 7805 dissipates nearly 4 watts. That is a heat sink situation.
The 7806 dissipates 4.5 watts. Heat sink again.
Turn it down to 7.5 volts and the 7806 is in danger of not having enough voltage to work with unless you can find a low dropout 7806.
A quick search of www.mouser.com shows nothing that fits the 6 volt position in a low dropout regulator.
You could get away without the heat sinks by using a switching regulator, but I have seen so many high speed devices go bad while my analog designs from the 1970's are still running.

What? Designs from the 1970's are still running?
Yes. That's what I said.
Build it right and the math confirms real world experience.
The most common failure today is aluminum electrolytic capacitors. Your 30 watt supply most likely has some. That's where you put your money if you want reliability. Choose capacitors rated for almost double the actual voltage they will experience and use caps rated for high temperature. I think the good number is 105 C. Install resistors rated for at least 2x the watts they will experience.

Here are some 5V regulators rated at 0ver 2 amps:
http://www.mouser.com/Power/Power-M...0w8p8Z1z0vlw7Z1z0w25cZ1yzryzwZ1z0wd5eZ1z0wbdj

On a personal note, you sound kind of half way there right now. Don't be surprised if you are on your 4th time through the design when everything fits correctly.

A nice reason to live longer: verifying a rugged design lasting maybe 60 years! Do not forget to post th we outcome.

 

A nice reason to live longer: verifying a rugged design lasting maybe 60 years! Do not forget to post the outcome.

Should I revive this Thread when my 1980 thermostat fails or just check in every ten years?
Seriously, I think I'm going to die before that thermostat design goes bad.

 

I built a 5VDC supply using 7805 about 50 years ago. Still using it with original components.

 

Off Topic: Wow! such a great source of experience and knowledge from everyone here is utterly amazing. Not trying to sound sarcastic, actually being very sincere, I've been posting on a different forum to try to get some answers to other issues, always receiving "bad attitude" from everyone. Here is very welcoming and I appreciate that from all of you!

On Topic:
I'll try to address everyone's response individually but know that I have read each one and I appreciate you sharing your knowledge with such an amateur like myself

I meant how many decades of life expectancy you required...
As with most things, it depends.

Using one power supply reduces expense and is probably what I'd do.
How are you calculating your peak current requirement? Do you have any switching logic?

LM78xx regulators will provide up to 1A of current. Your 960mA figure is close enough to the limit that you'll probably want external pass transistors on both regulators. The datasheet will usually have an example.

I'm calculating my peak from all the datasheets i've looked over for each part in my design. I have a relay board that says 160ma when all relays are "ON". So the odds of this happening are very slim due to how I have the PLC/microcontroller programmed but I've included the 160 as a just in case.

Switching logic I'm not exactly sure what you mean. I have relays that will be switching but they are solid state using phototriac isolation.

As far as the regulator max current I actually didn't even think about that, It is very close to my limit.

Start with 9 volts and the 7805 dissipates nearly 4 watts. That is a heat sink situation.
The 7806 dissipates 4.5 watts. Heat sink again.
Turn it down to 7.5 volts and the 7806 is in danger of not having enough voltage to work with unless you can find a low dropout 7806.
A quick search of www.mouser.com shows nothing that fits the 6 volt position in a low dropout regulator.
You could get away without the heat sinks by using a switching regulator, but I have seen so many high speed devices go bad while my analog designs from the 1970's are still running.

What? Designs from the 1970's are still running?
Yes. That's what I said.
Build it right and the math confirms real world experience.
The most common failure today is aluminum electrolytic capacitors. Your 30 watt supply most likely has some. That's where you put your money if you want reliability. Choose capacitors rated for almost double the actual voltage they will experience and use caps rated for high temperature. I think the good number is 105 C. Install resistors rated for at least 2x the watts they will experience.

Here are some 5V regulators rated at 0ver 2 amps:
http://www.mouser.com/Power/Power-M...0w8p8Z1z0vlw7Z1z0w25cZ1yzryzwZ1z0wd5eZ1z0wbdj

On a personal note, you sound kind of half way there right now. Don't be surprised if you are on your 4th time through the design when everything fits correctly.

LOL, this is actually very ironic. I had my project setup on my desk for about a week before we transferred it to the machine it runs. After rewiring EVERYTHING onto the machine. Hit the start button and nothing happened. It was jaw dropping. I actually had to go through rewire again to find my error. Luckily it didn't take to long.

I did intend to use heat sinks just in case even if they didn't logically need them. Our facility gets a bit warm in the summer time so I'm also considering installing a small fan just to keep air moving inside the control box. As far as the 1970's go, we all started from somewhere and I agree, some of the best tools, components, etc are the "older" ones. At the same time, to me at least, it is quite funny how technology has advanced so much in the past couple of decades. Going from a computer the size of a bedroom to a computer that unplugs and goes anywhere (laptop). I'm getting off topic with my endless rambling...

I'm going to do my best at building it right, theres always room for error, part defects and such but I'm hoping for my design to last as long as possible. My boss would like to keep things simple so if/when things burn up or go bad, its an easy troubleshoot. Where my theory is to build it so that it should never go bad. So a little arguing is defiantly in store.

A nice reason to live longer: verifying a rugged design lasting maybe 60 years! Do not forget to post th we outcome.

That would be great if it will last that long. I will defiantly be posting a diagram and specs after I revamp some of my design to include some of the responses from here. Plus a double check from experts is always welcomed.

 

I'm going to do my best at building it right, there's always room for error, part defects and such

There is a thing called, "infant mortality" in IC's. My experience was 3% when running production lines installing commercial grade chips. All failures showed up within 30 days. If it's running rock solid after thirty days, and you did a good design, you can assume you're safe.

I did intend to use heat sinks just in case even if they didn't logically need them. Our facility gets a bit warm in the summer time so I'm also considering installing a small fan

Those regulators definitely need heat sinks!
A little bit of air flow does amazing things, but a fan is far more likely to fail than a transistor. Just use some ridiculous size heat sinks and fugget abowdit.