Correct use of "sup" direction pins?

Am 13.11.2014 um 04:00 schrieb Ryan Pettigrew:

Is it correct use of a "sup" direction pin for the following, and if

not, why not?

 

• Regulated Voltage Outputs (Battery Charge Controllers, Regulator

outputs)

• Reference Volage Outputs

• Drive Voltages Outputs (outputs for loads)

 

--

To view any images and attachments in this post, visit:

http://www.element14.com/community/message/131292

 

Direction SUP are only für voltage symbols as VCC GND 0V +5V +12V ....

A "output" of a regulator is a OUT.

***

Um das Original zu lesen, und auch zuverlässig angehängte Dateien

bereitstellen zu können, benutzen Sie news.cadsoft.de und einen

funktionierenden News-Reader wie Thunderbird!

***

Mit freundlichen Grüßen / Best regards

Alfred Zaffran

--

______________________________________________________________

Alfred Zaffran              Support

CadSoft Computer GmbH       Hotline:   08635-698930

Pleidolfweg 15              FAX:       08635-698940

84568 Pleiskirchen          eMail: <alf@cadsoft.de>

                             Web:   <www.cadsoft.de>

Registergericht: Amtsgericht Traunstein HRB 5573

Geschaeftsfuehrer: Mark Whiteling

______________________________________________________________

Am 13.11.2014 um 04:00 schrieb Ryan Pettigrew:

Is it correct use of a "sup" direction pin for the following, and if

not, why not?

 

• Regulated Voltage Outputs (Battery Charge Controllers, Regulator

outputs)

• Reference Volage Outputs

• Drive Voltages Outputs (outputs for loads)

 

--

To view any images and attachments in this post, visit:

http://www.element14.com/community/message/131292

 

Direction SUP are only für voltage symbols as VCC GND 0V +5V +12V ....

A "output" of a regulator is a OUT.

***

Um das Original zu lesen, und auch zuverlässig angehängte Dateien

bereitstellen zu können, benutzen Sie news.cadsoft.de und einen

funktionierenden News-Reader wie Thunderbird!

***

Mit freundlichen Grüßen / Best regards

Alfred Zaffran

--

______________________________________________________________

Alfred Zaffran              Support

CadSoft Computer GmbH       Hotline:   08635-698930

Pleidolfweg 15              FAX:       08635-698940

84568 Pleiskirchen          eMail: <alf@cadsoft.de>

                             Web:   <www.cadsoft.de>

Registergericht: Amtsgericht Traunstein HRB 5573

Geschaeftsfuehrer: Mark Whiteling

______________________________________________________________

Isn't the point of a regulator to provide such a supply voltage as 5V, 12V, and so on? Shouldn't its output have the direction set to sup?

On 14/11/14 00:33, Ryan Pettigrew wrote:

Isn't the point of a regulator to provide such a supply voltage as 5V,

12V, and so on? Shouldn't its output have the direction set to sup?

 

No, it shouldn't. As you've already been told by one of CadSoft's

engineers. Why don't you believe him?

Perhaps what you actually meant to ask was:

Oh, okay, but I'm not clear why. Could you explain the logic, please?

If so, it's something like this:

A pin of 'direction' "SUP" has a special meaning. It creates a supply

net of the same name as the pin (which is a special net in some other ways).

A pin of 'direction' "PWR" is also special, in that it automatically

connects to such a net if there is one with its name on the sheet as

another part of the device, even if the symbol it's part of has not been

invoked.

The output of a voltage regulator is going to be called "out" or "Vout"

or some such. The regulator device in the Eagle library will be a 78LXX

with no committed voltage until you place it. Even if it's a dedicated

5V regulator, you may choose to have three separate 5V rails (say, VCC,

VDD, V+, for core, digital I/O and analogue) all provided by otherwise

identical regulators. That's three instances of a 78L05 device, each

with different nets connected to the output pin.

Now, given the special meaning of "SUP" and "PWR" pins, you can

hopefully see that Alfred's answer is the only possible correct one.

"A. Zaffran"  skrev i nyhetsmeldingen: m42atc$1mp$2@cheetah.cadsoft.de ...

Direction SUP are only für voltage symbols as VCC GND 0V +5V +12V ....

A "output" of a regulator is a OUT.

I get ERC multiple pages of error messages "OUTPUT and SUPPLY pins mixed on

net ..." when doing this. Using SUP for regulator output pins cause less

headace when you use power symbols for page connectors.

Ive suggested changes to the use of SUP in suggest forum on several

occations.

CadSoft Guest wrote:

 

On 14/11/14 00:33, Ryan Pettigrew wrote:

Isn't the point of a regulator to provide such a supply voltage as 5V,

12V, and so on? Shouldn't its output have the direction set to sup?

 

No, it shouldn't. As you've already been told by one of CadSoft's

engineers. Why don't you believe him?

Why don't you believe that regulators output specific voltages, as per their datasheets? This has nothing to do with believing him or not. In modern circuits, without some sort of means of regulating an unregulated supply, there aren't any sources such as 5V, 12V, and so on. If the parts of your circuit need to be connected to a "sup" direction pin, why shouldn't your actual voltage supply be that pin?

CadSoft Guest wrote:

 

Perhaps what you actually meant to ask was:

Oh, okay, but I'm not clear why. Could you explain the logic, please?

 

If so, it's something like this:

 

A pin of 'direction' "SUP" has a special meaning. It creates a supply

net of the same name as the pin (which is a special net in some other ways).

 

A pin of 'direction' "PWR" is also special, in that it automatically

connects to such a net if there is one with its name on the sheet as

another part of the device, even if the symbol it's part of has not been

invoked.

"Something like" is irrelevant; it's "exactly like" that matters. Vague nonsense doesn't matter.

CadSoft Guest wrote:

 

The output of a voltage regulator is going to be called "out" or "Vout"

or some such. The regulator device in the Eagle library will be a 78LXX

with no committed voltage until you place it. Even if it's a dedicated

5V regulator, you may choose to have three separate 5V rails (say, VCC,

VDD, V+, for core, digital I/O and analogue) all provided by otherwise

identical regulators. That's three instances of a 78L05 device, each

with different nets connected to the output pin.

No one is talking about that particular regulator. And while a linear regulator may not have a fixed voltage per se, it is still nominally something; and 5V and 12V aren't meant to be precisely those values, they are only nominally those values; so I see no difference in that case, either. Even if what you are saying about multiple regulators with the same output voltage is true, that doesn't mean that's how the software should, or necessarily does, work. Further, if I wanted 3 separate ideal voltage sources with the same value in my schematic, to represent redundant supplies, I should be able to get it without having difficulty. Are you saying that Eagle is going to throw a fit if I try to do that, too?

Hi,

Since this post is in the 'support' forum  I guess that  the right answer is in the answer of 'Alfred Zaffran'.

But, It's true the way Eagle manage supply could be improved, but that should be post in the 'suggestion' forum. There are already some requests about this (for example, look at the post http://www.element14.com/community/message/124500).

By the way, even if it's not in the suggestion forum, I would give my opinion (and it's only my point of view) about this "automatic connection feature of supply" as described :

 

A pin of 'direction' "SUP" has a special meaning. It creates a supply

net of the same name as the pin (which is a special net in some other ways).

 

A pin of 'direction' "PWR" is also special, in that it automatically

connects to such a net if there is one with its name on the sheet as

another part of the device, even if the symbol it's part of has not been

invoked.

This  feature is obsolete and unpractical for modern board: I never used it, I will never do and I HATE this. Why ?

Because:

- I want to see on schematic ALL connection, and hidden connection is the best way to made a big mistake !

- in modern board, you have so many supply, that and automatic connection is irrelevant : two identical devices (same library, so same pin name) can be used in different part of you design, with DIFFERENT supplies. Unless if in your library, you duplicate this device and rename the supply pin, each time you had to use it with a different supply: you will end up with N (with N>>1) versions of the same amplifier, one for the [0..5V] supply, one for the [-5V..+5V] supply, one for the [0..5V_A] supply,

one for the [0..VDD] supply, etc..... - > NONSENSE !

- What the point of hiding power supply, since you have to had decoupling capacitors on supply. I think it good practice, in schematic,  to place decoupling capacitor right away the power supply pin. So when you place and route, just by looking at you schematic, you know which capacitor must be place near the power pad of your IC.

- Since we can connect multiple PAD/SMD to one single pin of the symbol, there is no need to hide supply pin, anymore ! the schematic is now quite clear !

But as I said it's only my point of view....

On 2014-11-16 07:30:35 +0000, Ryan Pettigrew said:

Why don't you believe that regulators output specific voltages, as per

their datasheets? This has nothing to do with believing him or not.

Without some sort of means of regulating an unregulated supply, there

aren't any sources such as 5V, 12V, and so on. If, practically

speaking, you can't get a precision voltage like that without a

regulator, why shouldn't it be the regulator's output pin whose

direction be set to "sup"?

The output of a regulator may not always be going to a supply rail,

that is why.

Many people use things like the LM317 regulator for ... creative...

appliations. Constant current chargers are a common example. I

personally use regulators with digital pots or filtered PWM outputs to

vary the regulator output under microcontroller control and the

resulting output is used to power a device under test.

If you used the pin "direction" of "sup" for this, then the net

connected to the pin would be forced to be +VADJ or whatever you named

the pin. This absolutely is not what you want in practically every case.

Another example: I might have two 5V regulators; one for a digital

domain and one for a quiet analog domain. If you used the "sup" pin

direction for the output, they would be tied together automatically.

This isn't what you want. Instead, you would create a +5VA schematic

symbol and connect the output of the one regulator to it, leaving the

output of the other connected to your VCC or VDD or 5 or V net.

Just my $0.02 as to why the Eagle explanation of why you don't set pin

direction to "sup" for anything but dedicated power symbols, and also

why you would use the pin direction of "out" or even "pas" and not

"pwr" for regualtor outputs.

No one is talking about that particular regulator. And while a linear

regulator may not have a fixed voltage per se, it is still nominally

something; and 5V and 12V aren't meant to be precisely those values,

they are only nominally those values; so I see no difference in that

case, either. Even if what you are saying about multiple regulators with

the same output voltage is true, that doesn't mean that's how the

software should, or necessarily does, do it. Further, if I wanted 3

separate ideal voltage sources with the same value in my schematic, to

represent redundant supplies, I should be able to get it without having

difficulty. Are you saying that Eagle is going to throw a fit if I try

to do that, too?

You can absolutely get that without pain; don't use the pin direction

"sup" for the output of regulator pins.

-A.

On 16/11/14 07:30, Ryan Pettigrew wrote:

Why don't you believe that regulators output specific voltages,

Why don't you stop being a pillock and read what I wrote.

Believe it or not, I WAS being helpful

CadSoft Guest wrote:

 

On 16/11/14 07:30, Ryan Pettigrew wrote:

Why don't you believe that regulators output specific voltages,

 

Why don't you stop being a pillock and read what I wrote.

 

Believe it or not, I WAS being helpful

 

No, you were merely trying to be helpful; you were also failing, and miserably.

Your explanation essentially amounted to, "Do what Mr. Cadsoft guy says, because of magic reasons without logical explanation!"; no one should ever take such an argument as a valid reason for anything!