开云体育

I expect you are right, and I'm into learning still, but it takes time to grasp the nettle.

Creating libs etc is very easy and moving things from one lib to another
is also easy. It's a task worth getting to know.

The EDR is not there to tell you that your design won't work, its to
ensure that your connection are all there.

I've never used the LTspice or Spice function of Kicad, but I do use LT
spice stand alone.

When I started, I did only one board with tape and dots.. Fortunately it
was a training exercise, and very shortly after we got an early PCB
layout system Racal Kadet I think. The layout techs would not let us
engineers near the thing, I think they thought (probably correctly) that
we would have it in bits to see what made it tick :-]

Andy
 

 


 On Fri, 21 Jun 2024
22:10:37 +0100 "John Woodgate" <jmw@...> wrote:

I can accept all you say, but for my limited work I think that creating 
new libraries is not justified. The point about the opamp is that the 
EDR won't tell you that you have provided strong positive feedback 
because you have not noticed that the + input is at the top. An .AC 
might also not tell you, but a .TRAN would. My projects are now
'kitchen sink' - special test equipment, usually one-off, but when I
was in industry, 4000 would be a very modest run. In those days, it was
black tape on drafting film.

On 2024-06-21 21:46, Andy wrote:

Yes the pin positions are more common with -ive uppermost, but not
always.

The EDR cannot detect this because it in NOT an error it's just how
the symbol was drawn. It's down to the designer to connect the
components to the correct pin.

If you don't like it, then change the symbol...
I can't stand the American electrolytic symbol, so I change that to
the one that I've used for decades, the filled box for -ive and
unfilled for +ive

This highlights a VERY important point for ANY CAD system be it Kicad,
Eagle,  Autocad, or any other design tool.

You NEVER use the provided symbols for live work. You ONLY use symbols
that you have checked and verified as good. These you save in your own
libraries and that's what you use when building your design on the
screen. It protects you from an unknown change being distributed to
the system that you don't spot, so when you use the part again you get
something different.

Not a problem on a single board "kitchen sink" project you just curse
at it, and bodge it. But if that happens on your 4000 board
production run you might be looking for a new job :-)

When you think about it, you only use a very small handful of
components in amateur projects so it does not take long to build up
your own libraries, and most of the provided ones will be perfectly
OK so it's just a matter of copying therm across to your own libs.


Andy



On Fri, 21 Jun 2024 16:04:18 +0100
"John Woodgate"<jmw@...>  wrote:
 

As I understand it, I have to make a copy of the schematic and add
voltage sources (power and signal) and a load. Then add to that the
simulation command. But how do I introduce the models and subcircuits
for the active components?

What would be much easier for me is to be able to copy the schematic
into LTspice (as that is what I use) and do the simulation in that
environment. But is is possible to do that, i.e. can KiCad export a
schematic (or even a netlist) that SPICE understands?

This is important because it can be a costly mistake not to realise
that the KiCad opamp symbol has the + input uppermost, whereas it is
most common, I think, to see the - input uppermost. The EDR can't
detect that error, of course.

-- 
Signature OOO - Own Opinions Only
Best wishes
John Woodgate, Rayleigh, Essex UK
Keep trying

-- 
This email has been checked for viruses by AVG antivirus software.





 

-- 
Signature OOO - Own Opinions Only
Best wishes
John Woodgate, Rayleigh, Essex UK
Keep trying

Creating libs etc is very easy and moving things from one lib to another
is also easy. It's a task worth getting to know.

The EDR is not there to tell you that your design won't work, its to
ensure that your connection are all there.

I've never used the LTspice or Spice function of Kicad, but I do use LT
spice stand alone.

When I started, I did only one board with tape and dots.. Fortunately it
was a training exercise, and very shortly after we got an early PCB
layout system Racal Kadet I think. The layout techs would not let us
engineers near the thing, I think they thought (probably correctly) that
we would have it in bits to see what made it tick :-]

Andy





On Fri, 21 Jun 2024
22:10:37 +0100 "John Woodgate" <jmw@...> wrote:

I can accept all you say, but for my limited work I think that creating
new libraries is not justified. The point about the opamp is that the
EDR won't tell you that you have provided strong positive feedback
because you have not noticed that the + input is at the top. An .AC
might also not tell you, but a .TRAN would. My projects are now
'kitchen sink' - special test equipment, usually one-off, but when I
was in industry, 4000 would be a very modest run. In those days, it was
black tape on drafting film.

On 2024-06-21 21:46, Andy wrote:

Yes the pin positions are more common with -ive uppermost, but not
always.

The EDR cannot detect this because it in NOT an error it's just how
the symbol was drawn. It's down to the designer to connect the
components to the correct pin.

If you don't like it, then change the symbol...
I can't stand the American electrolytic symbol, so I change that to
the one that I've used for decades, the filled box for -ive and
unfilled for +ive

This highlights a VERY important point for ANY CAD system be it Kicad,
Eagle, Autocad, or any other design tool.

You NEVER use the provided symbols for live work. You ONLY use symbols
that you have checked and verified as good. These you save in your own
libraries and that's what you use when building your design on the
screen. It protects you from an unknown change being distributed to
the system that you don't spot, so when you use the part again you get
something different.

Not a problem on a single board "kitchen sink" project you just curse
at it, and bodge it. But if that happens on your 4000 board
production run you might be looking for a new job :-)

When you think about it, you only use a very small handful of
components in amateur projects so it does not take long to build up
your own libraries, and most of the provided ones will be perfectly
OK so it's just a matter of copying therm across to your own libs.


Andy



On Fri, 21 Jun 2024 16:04:18 +0100
"John Woodgate"<jmw@...> wrote:

As I understand it, I have to make a copy of the schematic and add
voltage sources (power and signal) and a load. Then add to that the
simulation command. But how do I introduce the models and subcircuits
for the active components?

What would be much easier for me is to be able to copy the schematic
into LTspice (as that is what I use) and do the simulation in that
environment. But is is possible to do that, i.e. can KiCad export a
schematic (or even a netlist) that SPICE understands?

This is important because it can be a costly mistake not to realise
that the KiCad opamp symbol has the + input uppermost, whereas it is
most common, I think, to see the - input uppermost. The EDR can't
detect that error, of course.

--
Signature OOO - Own Opinions Only
Best wishes
John Woodgate, Rayleigh, Essex UK
Keep trying

--
This email has been checked for viruses by AVG antivirus software.
www.avg.com

--
Signature OOO - Own Opinions Only
Best wishes
John Woodgate, Rayleigh, Essex UK
Keep trying

开云体育

Thanks; that's how I like help to be presented. I just have to find the Mounting hole library.

As others have probably said, it is easy (in the PCB editor) to add 
predefined mounting holes to the board as follows:

* Select 'Add Footprint' from the menu on the right-hand side
* Scroll to the 'MountingHole' library
* Select the desired footprint (i.e. 'MountingHole_2.2mm_M2') and click 'OK'
* Place the component as desired on the board, left-click to confirm 
placement.

I also 'Lock' my mounting holes once placed to prevent them being moved 
accidentally.

On 6/22/2024 1:55 AM, John Woodgate wrote:

I didn't know that. I've never seen it, but now I will look for it.

On 2024-06-21 17:52, Jon Escombe via groups.io wrote:

Perhaps worth noting that there is also MountingHole footprint 
library. You'll find a number of standard metric hole sizes already in 
there..

Regards,
Jon

On 21/06/2024 17:38, John Woodgate via groups.io wrote:

Thank you. I can understand that.

On 2024-06-21 17:07, Robert via groups.io wrote:

The second paragraph is "Footprint pads".?? You just need to create a
single through-hole pad of the correct dimensions.

-- 
Signature OOO - Own Opinions Only
Best wishes
John Woodgate, Rayleigh, Essex UK
Keep trying

	Virus-free.www.avg.com 

<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

开云体育

As others have probably said, it is easy (in the PCB editor) to add
predefined mounting holes to the board as follows:

* Select 'Add Footprint' from the menu on the right-hand side
* Scroll to the 'MountingHole' library
* Select the desired footprint (i.e. 'MountingHole_2.2mm_M2') and click 'OK'
* Place the component as desired on the board, left-click to confirm
placement.

I also 'Lock' my mounting holes once placed to prevent them being moved
accidentally.

开云体育

I can accept all you say, but for my limited work I think that creating new libraries is not justified. The point about the opamp is that the EDR won't tell you that you have provided strong positive feedback because you have not noticed that the + input is at the top. An .AC might also not tell you, but a .TRAN would. My projects are now 'kitchen sink' - special test equipment, usually one-off, but when I was in industry, 4000 would be a very modest run. In those days, it was black tape on drafting film.

Yes the pin positions are more common with -ive uppermost, but not
always. 

The EDR cannot detect this because it in NOT an error it's just how the
symbol was drawn. It's down to the designer to connect the components to
the correct pin.

If you don't like it, then change the symbol...
I can't stand the American electrolytic symbol, so I change that to the
one that I've used for decades, the filled box for -ive and unfilled for
+ive

This highlights a VERY important point for ANY CAD system be it Kicad,
Eagle,  Autocad, or any other design tool.

You NEVER use the provided symbols for live work. You ONLY use symbols
that you have checked and verified as good. These you save in your own
libraries and that's what you use when building your design on the
screen. It protects you from an unknown change being distributed to the
system that you don't spot, so when you use the part again you get
something different.

Not a problem on a single board "kitchen sink" project you just curse at
it, and bodge it. But if that happens on your 4000 board production run
you might be looking for a new job :-)

When you think about it, you only use a very small handful of components
in amateur projects so it does not take long to build up your own
libraries, and most of the provided ones will be perfectly OK so it's
just a matter of copying therm across to your own libs.


Andy



On Fri, 21 Jun 2024 16:04:18 +0100
"John Woodgate" <jmw@...> wrote:

As I understand it, I have to make a copy of the schematic and add 
voltage sources (power and signal) and a load. Then add to that the 
simulation command. But how do I introduce the models and subcircuits 
for the active components?

What would be much easier for me is to be able to copy the schematic 
into LTspice (as that is what I use) and do the simulation in that 
environment. But is is possible to do that, i.e. can KiCad export a 
schematic (or even a netlist) that SPICE understands?

This is important because it can be a costly mistake not to realise that 
the KiCad opamp symbol has the + input uppermost, whereas it is most 
common, I think, to see the - input uppermost. The EDR can't detect that 
error, of course.

-- 
Signature OOO - Own Opinions Only
Best wishes
John Woodgate, Rayleigh, Essex UK
Keep trying

-- 
This email has been checked for viruses by AVG antivirus software.

Yes the pin positions are more common with -ive uppermost, but not
always.

The EDR cannot detect this because it in NOT an error it's just how the
symbol was drawn. It's down to the designer to connect the components to
the correct pin.

If you don't like it, then change the symbol...
I can't stand the American electrolytic symbol, so I change that to the
one that I've used for decades, the filled box for -ive and unfilled for
+ive

This highlights a VERY important point for ANY CAD system be it Kicad,
Eagle, Autocad, or any other design tool.

You NEVER use the provided symbols for live work. You ONLY use symbols
that you have checked and verified as good. These you save in your own
libraries and that's what you use when building your design on the
screen. It protects you from an unknown change being distributed to the
system that you don't spot, so when you use the part again you get
something different.

Not a problem on a single board "kitchen sink" project you just curse at
it, and bodge it. But if that happens on your 4000 board production run
you might be looking for a new job :-)

When you think about it, you only use a very small handful of components
in amateur projects so it does not take long to build up your own
libraries, and most of the provided ones will be perfectly OK so it's
just a matter of copying therm across to your own libs.


Andy



On Fri, 21 Jun 2024 16:04:18 +0100
"John Woodgate" <jmw@...> wrote:

As I understand it, I have to make a copy of the schematic and add
voltage sources (power and signal) and a load. Then add to that the
simulation command. But how do I introduce the models and subcircuits
for the active components?

What would be much easier for me is to be able to copy the schematic
into LTspice (as that is what I use) and do the simulation in that
environment. But is is possible to do that, i.e. can KiCad export a
schematic (or even a netlist) that SPICE understands?

This is important because it can be a costly mistake not to realise that
the KiCad opamp symbol has the + input uppermost, whereas it is most
common, I think, to see the - input uppermost. The EDR can't detect that
error, of course.

--
Signature OOO - Own Opinions Only
Best wishes
John Woodgate, Rayleigh, Essex UK
Keep trying

--
This email has been checked for viruses by AVG antivirus software.
www.avg.com

When I add Test Points I add a part number for the type I want and I use GRN for Ground, BLK for Digital Ground, RED for Voltages, YEL, and BLU for various other signals...??
When I add 2-Pin Jumbers I add a separate schematic symbol and part number for the Shunt...? I have my own color code for those too...
You could add a part number for a hole so you would remember to buy screws or have a separate symbol for screws, washers...
Same for stand-offs or other hardware items...
I typically have all of those hardware items on the last page of my schematic...? This makes it easy to open a previous schematic and cut-n-paste them into the new design...? ?

73 Dallas N4DDM
CQ Field Day, CQ Field Day

开云体育

I didn't know that. I've never seen it, but now I will look for it.

Perhaps worth noting that there is also MountingHole footprint library. You'll find a number of standard metric hole sizes already in there..

Regards,
Jon

On 21/06/2024 17:38, John Woodgate via groups.io wrote:

Thank you. I can understand that.

On 2024-06-21 17:07, Robert via groups.io wrote:

The second paragraph is "Footprint pads".?? You just need to create a
single through-hole pad of the correct dimensions.

开云体育

I will study the guide again.

There's a guide in the getting started guide:



If you just want a simple bolt hole, nothing fancy, use a single pin
connector as the schematic symbol, connect it to a net or not as
required, and skip down to:



The second paragraph is "Footprint pads".?? You just need to create a
single through-hole pad of the correct dimensions.

Once you have your footprint saved, you can go back to the schematic
symbol and select it in the symbol properties.

Regards,

Robert.

* Plain text email - safe, readable, inclusive. *

Perhaps worth noting that there is also MountingHole footprint library. You'll find a number of standard metric hole sizes already in there..

Regards,
Jon

开云体育

I made a note, but I didn't buy. Now I have bought (not without difficulty, but that's life).

You should seriously get the ebooks I mentioned in one of my previous posts, it will take you right through all these aspects, and be a reference for the various questions.

On Fri, 21 Jun 2024 at 16:52, John Woodgate via <jmw=[email protected]> wrote:

Would that I had your depth of knowledge!

On 2024-06-21 16:47, Tony Casey wrote:

Exactly so. I include mechanical things like fixing holes, RF screens, and the PCB itself in the schematic as parts.

--
Regards,
Tony

On 21/06/2024 17:24, Robert via wrote:

A hole is just a single pad of the correct dimensions.?? It can have
copper layers, or none (in which case it becomes a non-plated through
hole).?? If it has copper layers, they can optionally be connected to
any net (usually the ground plane, and typically with no thermal
spokes).?? In the past I just defined the hole on the schematic with a
single pin connector, but these days I have my own symbol.?? Often the
associated footprint that I create will define a board outline and all
the holes (as pads), and typically much more (such as silk screen
elements).?? I routinely import that from mechanical CAD as a DXF.

So there are no hard and fast rules - you just do whatever cranks your
tractor on the day.

--
OOO - Own Opinions Only
Best wishes
John Woodgate, Rayleigh, Essex UK
Keep trying

Virus-free.

开云体育

Thank you. I can understand that.

There's a guide in the getting started guide:



If you just want a simple bolt hole, nothing fancy, use a single pin
connector as the schematic symbol, connect it to a net or not as
required, and skip down to:



The second paragraph is "Footprint pads".?? You just need to create a
single through-hole pad of the correct dimensions.

Once you have your footprint saved, you can go back to the schematic
symbol and select it in the symbol properties.

Regards,

Robert.

* Plain text email - safe, readable, inclusive. *

There's a guide in the getting started guide:



If you just want a simple bolt hole, nothing fancy, use a single pin
connector as the schematic symbol, connect it to a net or not as
required, and skip down to:



The second paragraph is "Footprint pads". You just need to create a
single through-hole pad of the correct dimensions.

Once you have your footprint saved, you can go back to the schematic
symbol and select it in the symbol properties.

Regards,

Robert.

* Plain text email - safe, readable, inclusive. *

开云体育

Thank you. I will investigate whether I can work on it in the SPICE environment.

I believe the most recent version (which I believe is what you are using) of kicad can export to LTspice. I haven't tried it but presume there is a specific export section in the file menu, least that is where I would expect to find it.

On Fri, 21 Jun 2024 at 16:05, John Woodgate via <jmw=[email protected]> wrote:

As I understand it, I have to make a copy of the schematic and add voltage sources (power and signal) and a load. Then add to that the simulation command. But how do I introduce the models and subcircuits for the active components?

What would be much easier for me is to be able to copy the schematic into LTspice (as that is what I use) and do the simulation in that environment. But is is possible to do that, i.e. can KiCad export a schematic (or even a netlist) that SPICE understands?

This is important because it can be a costly mistake not to realise that the KiCad opamp symbol has the + input uppermost, whereas it is most common, I think, to see the - input uppermost. The EDR can't detect that error, of course.

--
OOO - Own Opinions Only
Best wishes
John Woodgate, Rayleigh, Essex UK
Keep trying

Virus-free.

开云体育

Would that I had your depth of knowledge!

Exactly so. I include mechanical things like fixing holes, RF screens, and the PCB itself in the schematic as parts.

--
Regards,
Tony

On 21/06/2024 17:24, Robert via groups.io wrote:

A hole is just a single pad of the correct dimensions.?? It can have
copper layers, or none (in which case it becomes a non-plated through
hole).?? If it has copper layers, they can optionally be connected to
any net (usually the ground plane, and typically with no thermal
spokes).?? In the past I just defined the hole on the schematic with a
single pin connector, but these days I have my own symbol.?? Often the
associated footprint that I create will define a board outline and all
the holes (as pads), and typically much more (such as silk screen
elements).?? I routinely import that from mechanical CAD as a DXF.

So there are no hard and fast rules - you just do whatever cranks your
tractor on the day.

开云体育

Thank you, but I need a more elementary guide, step by step.

A hole is just a single pad of the correct dimensions.?? It can have
copper layers, or none (in which case it becomes a non-plated through
hole).?? If it has copper layers, they can optionally be connected to
any net (usually the ground plane, and typically with no thermal
spokes).?? In the past I just defined the hole on the schematic with a
single pin connector, but these days I have my own symbol.?? Often the
associated footprint that I create will define a board outline and all
the holes (as pads), and typically much more (such as silk screen
elements).?? I routinely import that from mechanical CAD as a DXF.

So there are no hard and fast rules - you just do whatever cranks your
tractor on the day.

Regards,

Robert.

* Plain text email - safe, readable, inclusive. *