X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f X-Recipient: geda-user AT delorie DOT com X-TCPREMOTEIP: 63.119.35.194 X-Authenticated-UID: jpd AT noqsi DOT com Content-Type: multipart/signed; boundary="Apple-Mail=_31857315-E68F-4DE0-B306-33C89DC5F977"; protocol="application/pgp-signature"; micalg=pgp-sha512 Mime-Version: 1.0 (Mac OS X Mail 7.3 \(1878.6\)) Subject: Re: [geda-user] Project leadership X-Pgp-Agent: GPGMail 2.5.2 From: John Doty In-Reply-To: Date: Mon, 4 Jan 2016 17:18:45 -0500 Message-Id: <2B07BB75-1CF0-4A30-985C-893E8DCAEE54@noqsi.com> References: <5AC3D5C5-144F-41AE-9562-47BC34D9047F AT noqsi DOT com>

<20151222193859 DOT 26898 DOT qmail AT stuge DOT se> <20151223202851 DOT 637d5b1f AT jive DOT levalinux DOT org> <20151223195846 DOT 8392 DOT qmail AT stuge DOT se> <20151229155647 DOT GA3752 AT localhost DOT localdomain> <20151229175222 DOT GD3752 AT localhost DOT localdomain> <96A12FC1-E09C-4D63-8346-5A62FDAB4228 AT sbcglobal DOT net> <2F0AFAE8-ADAB-4DE0-A79C-DD87077E66B5 AT sbcglobal DOT net>

To: geda-user AT delorie DOT com X-Mailer: Apple Mail (2.1878.6) Reply-To: geda-user AT delorie DOT com Errors-To: nobody AT delorie DOT com X-Mailing-List: geda-user AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk --Apple-Mail=_31857315-E68F-4DE0-B306-33C89DC5F977 Content-Type: multipart/alternative; boundary="Apple-Mail=_FE4E2689-36EB-4F3D-8419-164B07B049C6" --Apple-Mail=_FE4E2689-36EB-4F3D-8419-164B07B049C6 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=windows-1252 Peter, have you looked at https://github.com/noqsi/gnet-spice-noqsi? On Jan 4, 2016, at 7:12 AM, Peter Clifton = (petercjclifton AT googlemail DOT com) [via geda-user AT delorie DOT com] = wrote: >=20 > On 4 Jan 2016 14:00, "Roland Lutz" wrote: >=20 > > Thinking about it, I'm not even sure if this behavior (treating the = slots of a package as individual SPICE objects) is even necessary. The = other SPICE backends don't seem to be using it. >=20 The history of this is complicated. There was a time years ago when a = broken libgeda wound up in distribution because of a failed attempt to = make slotting work =93right=94 for both layout and SPICE. But what=92s = right for SPICE depends on context. Can you ignore crosstalk within a = package (maybe it=92s the most important thing you=92re modeling)? If = you can, and the SPICE model you have is for one slot (a very common = scenario), you may want to accommodate that. Otherwise, you probably = want to have one object per package. This issue was my most important motivation for writing = gnet-spice-noqsi, which allows the user to control the connections of = model instances to nets through the spice-prototype attribute. This = works at the symbol level, or even the instance level, with defaults = that imitate gnet-spice-sdb. Is the current behavior of gnet-spice-sdb for slotted devices documented = anywhere? > An alternative option would be to have a (schematic or netlist based) = sub-model which maps the connections of the actual part, to 'n' internal = instances of the cell model. >=20 >=20 Gnet-spice-noqsi makes this a user choice. > It really depends on whether you are viewing the schematic and spice = simulation as that of the design with actual parts, or at a more = abstracted level where actual embodiment of the opamp cells means = picking what chip they go in. >=20 >=20 Yes. And in the process of turning an idea into a finished design, you = may go through intermediate states that mix abstract and concrete. Even = when you have a finished design, the SPICE =93test fixture=94 is usually = a rather abstract piece of circuitry. That=92s why the back end needs to = allow the user to control the expression of a device in the SPICE = netlist. It also needs good defaults, of course, so you don=92t need to = hold its hand about every resistor. John Doty Noqsi Aerospace, Ltd. http://www.noqsi.com/ jpd AT noqsi DOT com --Apple-Mail=_FE4E2689-36EB-4F3D-8419-164B07B049C6 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=windows-1252 Peter, = have you looked at https://github.com/noqs= i/gnet-spice-noqsi?

On Jan 4, 2016, at 7:12 AM, = Peter Clifton (petercjclifton AT googlemail DOT co= m) [via geda-user AT delorie DOT com] <geda-user AT delorie DOT com> = wrote:

On 4 Jan 2016 14:00, "Roland Lutz" <rlutz AT hedmen DOT org> wrote:
> Thinking about it, I'm not even sure if this behavior = (treating the slots of a package as individual SPICE objects) is even = necessary. The other SPICE backends don't seem to be using = it.

The history of this is complicated. There = was a time years ago when a broken libgeda wound up in distribution = because of a failed attempt to make slotting work =93right=94 for both = layout and SPICE. But what=92s right for SPICE depends on context. Can = you ignore crosstalk within a package (maybe it=92s the most important = thing you=92re modeling)? If you can, and the SPICE model you have is = for one slot (a very common scenario), you may want to accommodate that. = Otherwise, you probably want to have one object per package.

This = issue was my most important motivation for writing gnet-spice-noqsi, = which allows the user to control the connections of model instances to = nets through the spice-prototype attribute. This works at the symbol = level, or even the instance level, with defaults that imitate = gnet-spice-sdb.

Is the current behavior of gnet-spice-sdb for = slotted devices documented anywhere?

An alternative option would be to have a (schematic or = netlist based) sub-model which maps the connections of the actual part, = to 'n' internal instances of the cell = model.

Gnet-spice-noqsi = makes this a user choice.

It = really depends on whether you are viewing the schematic and spice = simulation as that of the design with actual parts, or at a more = abstracted level where actual embodiment of the opamp cells means = picking what chip they go = in.

Yes. And in the = process of turning an idea into a finished design, you may go through = intermediate states that mix abstract and concrete. Even when you have a = finished design, the SPICE =93test fixture=94 is usually a rather = abstract piece of circuitry. That=92s why the back end needs to allow = the user to control the expression of a device in the SPICE netlist. It = also needs good defaults, of course, so you don=92t need to hold its = hand about every resistor.

John = Doty = Noqsi = Aerospace, Ltd.

http://www.noqsi.com/

jpd AT noqsi DOT com

= --Apple-Mail=_FE4E2689-36EB-4F3D-8419-164B07B049C6-- --Apple-Mail=_31857315-E68F-4DE0-B306-33C89DC5F977 Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename=signature.asc Content-Type: application/pgp-signature; name=signature.asc Content-Description: Message signed with OpenPGP using GPGMail -----BEGIN PGP SIGNATURE----- Comment: GPGTools - https://gpgtools.org iQIcBAEBCgAGBQJWiu/GAAoJEF1Aj/0UKykRxacQAJgrl9tvwUsvhybf2j3MEinu 7XSjTlLYdRZJByX8MsPFNp6yc5nnlrQ6HB5hMOwQ/SZ/bs2riboRbLwwZ4dHzsGi cjyOGPFTQ1zC3VhuDtI6vW/Do2sfy58kBRYZODqwOOm0tOJC4cC7eJ2c7ZE2p9oK T/S4pAoPNfDsyfQ5eJw+Aufp5dMQzi87s3sZgqxS1GMzq9t6U/79bcny2ys0CIXH OpM3AUZ3yyf9B0T0jMmUMvRN0qHcGd+RAFLVPyG1O0nPp+Q5f6QESSXIOk6fy9Y4 PBXFhcXsZYpCV1yJfJCsXojSluukH2dJmL/MF4qcb2HriHedMfEwakSoGq1sts+e e1roMlDGcbrNLX+/RNGybyyf4U11mI50nuWy3BKihZqWD8SmOLkDpV+rovSgHldA Wh7va+m/8omevJcXyjhiTrL0w30h26KelsdxNVAmYzM2ao4AqKeZbDMO3BNVjWev GQvxD4sniHYMl02aUF6vVEq5Ha1LIbZVqmPcAh2pVUbr1pY6pQTDuMh4nKq38d7F Rt93qoRK8C0xz855apveq4ZTu3lXDDWnbdBm2AqqKeeH8D/4E4HYa+JpIHq4WDQ+ 6miW35qcNh2DZwCLHfXA2hQ6x7+rEnQ4vOTwFZGpSV/GnYDSqhgQZZ8JdKQIPiFH 012L4mXILrLyFQSyLgmc =mxZ5 -----END PGP SIGNATURE----- --Apple-Mail=_31857315-E68F-4DE0-B306-33C89DC5F977--