We've hired a new PCB designer but the engineers are still doing their
own PCB layout until he starts.

I'm revising one board.

https://www.dropbox.com/s/xv5ue30bln5wwao/T502B3.jpg?raw=1

The changes involve maybe 15% of the design, but making changes is so
confusing that I just ripped most of the parts off-board and started
the layout almost from scratch. The power pours make using the old
placement even trickier. I saved some of the fast stuff upper-right.

This is picosecond stuff, so layout is super tedious.

I have two SMA connectors on 0.5" centers, which won't make me popular
in production. Maybe I'll be generous and give them another tenth of
an inch.

--

I yam what I yam - Popeye

John - what software are you using for the board layout?

Hul

jlarkin@highlandsniptechnology.com wrote:

> We've hired a new PCB designer but the engineers are still doing their
> own PCB layout until he starts.

> I'm revising one board.

> https://www.dropbox.com/s/xv5ue30bln5wwao/T502B3.jpg?raw=1

> The changes involve maybe 15% of the design, but making changes is so
> confusing that I just ripped most of the parts off-board and started
> the layout almost from scratch. The power pours make using the old
> placement even trickier. I saved some of the fast stuff upper-right.

> This is picosecond stuff, so layout is super tedious.

> I have two SMA connectors on 0.5" centers, which won't make me popular
> in production. Maybe I'll be generous and give them another tenth of
> an inch.

> --

> I yam what I yam - Popeye

On Wed, 23 Mar 2022 20:20:07 -0000 (UTC), Hul Tytus <ht@panix.com>
wrote:

>John - what software are you using for the board layout?
>
>Hul

PADS, a somewhat older version, which we prefer to the latest one.
We've been using PADS since the DOS days.

I've got everything placed!

https://www.dropbox.com/s/v8ugresrqugbfx9/T502_B4.jpg?raw=1

It still needs a ton of work.

--

If a man will begin with certainties, he shall end with doubts,
but if he will be content to begin with doubts he shall end in certainties.
Francis Bacon

John Larkin <jlarkin@highland_atwork_technology.com> wrote:

> On Wed, 23 Mar 2022 20:20:07 -0000 (UTC), Hul Tytus <ht@panix.com>
> wrote:
>
>>John - what software are you using for the board layout?
>>
>>Hul
>
> PADS, a somewhat older version, which we prefer to the latest one.
> We've been using PADS since the DOS days.
>
> I've got everything placed!
>
> https://www.dropbox.com/s/v8ugresrqugbfx9/T502_B4.jpg?raw=1
>
> It still needs a ton of work.

I used PCAD since the DOS days. Eventually it ran out of gas, and I
switched to several other expensive systems, then to DipTrace. Schematic
entry, manual and autoplace and autoroute, built-in component list, search,
modify components, trace timing, differential trace management, BOM, etc...

There is no comparison. They do everything possible to help you do your
job. Productivity is incredible, stress is zero. Built in checking makes it
impossible to make a mistake. Imports Altium, Eagle, KiCad, Orcad, Pcad,
Pads, exports to many programs. Spice netlist, etc...

And it's free. No ads.

--
MRM

On Thu, 24 Mar 2022 04:41:25 -0000 (UTC), Mike Monett <spamme@not.com>
wrote:
>
>I used PCAD since the DOS days. Eventually it ran out of gas, and I
>switched to several other expensive systems, then to DipTrace. Schematic
>entry, manual and autoplace and autoroute, built-in component list, search,
>modify components, trace timing, differential trace management, BOM, etc...
>

I`m stilll running PCAD2006. It´s still running fine for that what I
do. But now it`s really time to move on..

How was switching from PCAD (windows) to Diptrace? Moving Libs,
Designs and so on to Diptrace?

regards

Michael
- Michael Wieser
--

Michael Wieser <Mi_Wieser@web.de> wrote:

> On Thu, 24 Mar 2022 04:41:25 -0000 (UTC), Mike Monett <spamme@not.com>
> wrote:
>
>>
>>I used PCAD since the DOS days. Eventually it ran out of gas, and I
>>switched to several other expensive systems, then to DipTrace. Schematic
>>entry, manual and autoplace and autoroute, built-in component list,
>>search, modify components, trace timing, differential trace management,
>>BOM, etc...
>>
>
> I`m stilll running PCAD2006. It�s still running fine for that what I
> do. But now it`s really time to move on..
>
> How was switching from PCAD (windows) to Diptrace? Moving Libs,
> Designs and so on to Diptrace?
>
> regards
>
> Michael
> -
> Michael Wieser

I don't know. I was still on the DOS version but had been out of the
electronic design business for over two decades due to health issues. I
also lost all my files due to a catastrophic hard disk crash. When I
started to get my act together again, I suffered a series of strokes that
left me with the loss of eyesight in one eye and the loss of my legs.
However, I was still able to work in LTspice and found transferring ASC
files to Diptrace was very easy. The learning curve was the easiest I have
ever seen on any system, from programming in assembly language to any PC
CAD system that I used.

My experience with DipTrace tells me you will find transferring from PCAD
will be simple and trouble free. It will help to get familiar with the
wealth of features at your fingertips. You will be astonished at how easy
it is to learn and to use.

--
MRM

On Thu, 24 Mar 2022 04:41:25 -0000 (UTC), Mike Monett <spamme@not.com>
wrote:

>John Larkin <jlarkin@highland_atwork_technology.com> wrote:
>
>> On Wed, 23 Mar 2022 20:20:07 -0000 (UTC), Hul Tytus <ht@panix.com>
>> wrote:
>>
>>>John - what software are you using for the board layout?
>>>
>>>Hul
>>
>> PADS, a somewhat older version, which we prefer to the latest one.
>> We've been using PADS since the DOS days.
>>
>> I've got everything placed!
>>
>> https://www.dropbox.com/s/v8ugresrqugbfx9/T502_B4.jpg?raw=1
>>
>> It still needs a ton of work.
>
>I used PCAD since the DOS days. Eventually it ran out of gas, and I
>switched to several other expensive systems, then to DipTrace. Schematic
>entry, manual and autoplace and autoroute, built-in component list, search,
>modify components, trace timing, differential trace management, BOM, etc...
>
>There is no comparison. They do everything possible to help you do your
>job. Productivity is incredible, stress is zero. Built in checking makes it
>impossible to make a mistake. Imports Altium, Eagle, KiCad, Orcad, Pcad,
>Pads, exports to many programs. Spice netlist, etc...
>
>And it's free. No ads.

PADS works fine. I wouldn't want auto-place and usually wouldn't want
auto-route.

PADS can do anything in ascii files... schematic, board, library. So
we can ascii export, sic an editor or a python or basic program on it,
and re-import. Can other pcb programs do that?

Essentially all the parts that we use in the library are ones that we
created... a thousand at least. Converting the library for another
program might be messy.

We have hundreds of designs to maintain too. Best to stick with PADS.

progress:

https://www.dropbox.com/s/p0u3i1v24djsaam/T502B4.23.jpg?raw=1

I fumble-fingered something and deleted a couple of the fancy ground
pours and didn't notice until it was too late to undo. So I copied and
pasted the pours from a previous edit.

We have a disciplined process for incremental edits and for
controlling file ownership, so we drop off lots of steps. Edit B3 has
the pours I needed.

Next I have to do my least-favorite part, the power pours. This board
has two routing layers, two grounds, and two power planes. And a lot
of supply voltages.

On fast boards like this, all the action is on layer 1, all microstrip
with no fast stuff going through vias. Which means no crossovers on
diff pairs.

It needs a zillion stitching vias too, to nail the L1 pours to the L2
ground plane.

--

I yam what I yam - Popeye

jlarkin@highlandsniptechnology.com wrote:

[...]

I have no desire to get you to switch to DipTrace. I was describing my own
experience. Yours is completely different, and you have very good reasons
for wanting to stay on Pads. It makes no difference to me what software you
use. However, in the interest of academic completeness, I will answer your
questions.

> PADS works fine. I wouldn't want auto-place and usually wouldn't want
> auto-route.

You can place critical components and traces manually, and allow the
autoplace and autoroute to complete the boring stuff.

> PADS can do anything in ascii files... schematic, board, library. So
> we can ascii export, sic an editor or a python or basic program on it,
> and re-import. Can other pcb programs do that?

DipTrace can export PADS SCH, and import and export PADS PCB.

> Essentially all the parts that we use in the library are ones that we
> created... a thousand at least. Converting the library for another
> program might be messy.

True.

> We have hundreds of designs to maintain too. Best to stick with PADS.

True.
> progress:
>
> https://www.dropbox.com/s/p0u3i1v24djsaam/T502B4.23.jpg?raw=1
>
> I fumble-fingered something and deleted a couple of the fancy ground
> pours and didn't notice until it was too late to undo. So I copied and
> pasted the pours from a previous edit.
>
> We have a disciplined process for incremental edits and for
> controlling file ownership, so we drop off lots of steps. Edit B3 has
> the pours I needed.
>
> Next I have to do my least-favorite part, the power pours. This board
> has two routing layers, two grounds, and two power planes. And a lot
> of supply voltages.
>
> On fast boards like this, all the action is on layer 1, all microstrip
> with no fast stuff going through vias. Which means no crossovers on
> diff pairs.
>
> It needs a zillion stitching vias too, to nail the L1 pours to the L2
> ground plane.

Picosecond stuff is always demanding. I'm surprised you don't do more on
Rogers, but you keep the straces short. I recall the fun you had with
saucer pcbs. I'm trying to find out if Kapton flex pcb would be a suitable
replacement for Rogers.

--
MRM

On Wed, 23 Mar 2022 08:33:07 -0700, jlarkin@highlandsniptechnology.com
wrote:

>
>We've hired a new PCB designer but the engineers are still doing their
>own PCB layout until he starts.
>
>I'm revising one board.
>
>https://www.dropbox.com/s/xv5ue30bln5wwao/T502B3.jpg?raw=1
>
>The changes involve maybe 15% of the design, but making changes is so
>confusing that I just ripped most of the parts off-board and started
>the layout almost from scratch. The power pours make using the old
>placement even trickier. I saved some of the fast stuff upper-right.
>
>This is picosecond stuff, so layout is super tedious.
>
>I have two SMA connectors on 0.5" centers, which won't make me popular
>in production. Maybe I'll be generous and give them another tenth of
>an inch.

It's resequenced and done! Short of the design review and
panelization. I might play with ground vias some more.

https://www.dropbox.com/s/jpr0oolz7ewg1u7/T502B6-316.jpg?raw=1

The big chip upper-right is a $300 Hittite distributed amplifier. I
invented a new, bizarre, bias servo thing optimized for pulse swing.
After all, ABS MAX is not everything.

The reference designator situation is unfortunate. The tiny ones are
obviously useless on the real PCB, but can be seen on a PC, either in
PADS or, for manufacturing and QC, in a searchable PDF.

We'll move the sillier ref desigs to another layer so the fab house
doesn't call us and complain that they are too small to print. One of
the keys to pcb layout is to anticipate and avoid calls from the fab
house, which slows things down.

--

I yam what I yam - Popeye

On Saturday, March 26, 2022 at 3:26:07 PM UTC, jla...@highlandsniptechnology.com wrote:
> On Wed, 23 Mar 2022 08:33:07 -0700, jla...@highlandsniptechnology.com
> wrote:
> >

> The reference designator situation is unfortunate. The tiny ones are
> obviously useless on the real PCB, but can be seen on a PC, either in
> PADS or, for manufacturing and QC, in a searchable PDF.

Ref Des characters, would using less of them help?
Is there a reason we must use base-10 numbers?
I assume one could use base-36, so relabel "R10" as "RA", etc.
regards, RS

On Sunday, March 27, 2022 at 2:15:14 PM UTC-7, Rich S wrote:

> Ref Des characters, would using less of them help?
> Is there a reason we must use base-10 numbers?
> I assume one could use base-36, so relabel "R10" as "RA", etc.

As long as we're being cryptic, why not use a hash table; even leaving out
hard-to-identify I, O, Z as being too near numeric symbols, the alphabet
offers, in two-character chunks, 529 distinct labels. How many printed wiring
boards have that many components?

And, you can keep the hash table, cross-reference, and algorithm in the cloud, with the right
reference address for each wiring board to be printed on the board, in a suitable 2D barcode.
Every robot in the factory will know more about the board than any human.

Who wants to write and sell the cellphone camera apps? One for Android, one for IOS?

On Sun, 27 Mar 2022 14:15:10 -0700 (PDT), Rich S
<richsulinengineer@gmail.com> wrote:

>On Saturday, March 26, 2022 at 3:26:07 PM UTC, jla...@highlandsniptechnology.com wrote:
>> On Wed, 23 Mar 2022 08:33:07 -0700, jla...@highlandsniptechnology.com
>> wrote:
>> >
>
>> The reference designator situation is unfortunate. The tiny ones are
>> obviously useless on the real PCB, but can be seen on a PC, either in
>> PADS or, for manufacturing and QC, in a searchable PDF.
>
>Ref Des characters, would using less of them help?
>Is there a reason we must use base-10 numbers?
>I assume one could use base-36, so relabel "R10" as "RA", etc.
>regards, RS

GE did some control systems where the ref desigs on boards were just
numbers. No "R" or "C" or anything.

--

I yam what I yam - Popeye

On 25/03/2022 02:34, jlarkin@highlandsniptechnology.com wrote:
> PADS can do anything in ascii files... schematic, board, library. So
> we can ascii export, sic an editor or a python or basic program on it,
> and re-import. Can other pcb programs do that?

yes, KiCad