package parasitic models for Spice

[Paul Burridge]

Hi all,

For multi-gigahertz analysis, a regular Spice model of the chip/die
properties alone is of little use. You need to incorporate the device
package's parasitic properties, too. But these don't seem to very easy
to come by. Does anyone know of any sites where these additional model
parameters for UHF+ analysis can be found?

Thanks,

p.
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[Robert]

"Paul Burridge" <pb@notthisbit.osiris1.co.uk> wrote in message
news:ktt1i09jr4nfoqgqcntokvg7djp4posgpf@4ax.com...

Hi all,

For multi-gigahertz analysis, a regular Spice model of the chip/die
properties alone is of little use. You need to incorporate the device
package's parasitic properties, too. But these don't seem to very easy
to come by. Does anyone know of any sites where these additional model
parameters for UHF+ analysis can be found?

Thanks,

Depends on what package you choose. Most of the package suppliers have
simple spice models of the ones they offer.

Dean Monthei's book, has quite a few spice models for common package types
as well.

Robert

 

[Paul Burridge]

On Tue, 17 Aug 2004 17:51:25 GMT, "Robert" <robert@yahoo.com> wrote:

Depends on what package you choose. Most of the package suppliers have
simple spice models of the ones they offer.

Eh? Are you saying that dies and their packages are made by different
suppliers? I'd always assumed the transistor manufacturer also made
the packages for its dies, albeit to an industry-wide specification.
Who manufactures the packages, then?
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[legg]

On Mon, 16 Aug 2004 19:10:49 +0100, Paul Burridge
<pb@notthisbit.osiris1.co.uk> wrote:

Hi all,

For multi-gigahertz analysis, a regular Spice model of the chip/die
properties alone is of little use. You need to incorporate the device
package's parasitic properties, too. But these don't seem to very easy
to come by. Does anyone know of any sites where these additional model
parameters for UHF+ analysis can be found?

You might find what you need in some of these. If you care to identify
the particular package that you are interested in, it might help.

"Package lead inductance considerations in high-speed applications"
Stephen C. Hinkle, Jeffrey A. West
Philips AN212 '87
http://www.philipslogic.com/support/appnotes/logic/pdf/an212.pdf

Electrical Performance of Packages
Nat Semi AN1205 '01
http://www.national.com/an/AN/AN-1205.pdf

DE-SERIES FAST POWER MOSFET
George J Krausse,
Directed Energy, Inc. Doc #9300-0002 Rev 2 '00
http://www.ixysrf.com/pdf/switch_mode/appnotes/2de_series_intro.pdf

DE-Series MOSFET, DEIC420 And SOP-28 IC Device Installation & Mounting
Instructions
IXYS Doc #9300-0005 Rev 3 '03
http://www.ixysrf.com/pdf/diodes/app_notes/7de-series_mosfet_installation_instructions.pdf


"Lumped Skin Effect Model for Package Leads"
A.E.Engin
IBIS'03
http://www.eda.org/pub/ibis/summits/feb04b/engin.pdf

RL

 

[Robert]

"Paul Burridge" <pb@notthisbit.osiris1.co.uk> wrote >

Depends on what package you choose. Most of the package suppliers have
simple spice models of the ones they offer.

Eh? Are you saying that dies and their packages are made by different
suppliers? I'd always assumed the transistor manufacturer also made
the packages for its dies, albeit to an industry-wide specification.
Who manufactures the packages, then?
--

Yep. A lot of the packages moved offshore quite a while ago to the Assembly
houses in the Asian area. Thailand, Singapore, and others take the die from
the IC manufacturers and wirebond it into the various packages. Or
wedgebonds or ballbonds or ...

Do a google on Assembly IC packages and you should see a few of their sites.

Robert

 

[Paul Burridge]

On Wed, 18 Aug 2004 20:41:33 GMT, legg <legg@nospam.magma.ca> wrote:

You might find what you need in some of these. If you care to identify
the particular package that you are interested in, it might help.

[useful links snipped]

Thanks for the links.
Okay. I'm only talking about RF transistors here; nothing more
complicated than that. No specific package, but rather a variety of
the common types that are mostly used for encapsulating RF trannies.
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[legg]

On Thu, 19 Aug 2004 12:47:58 +0100, Paul Burridge
<pb@notthisbit.osiris1.co.uk> wrote:

On Wed, 18 Aug 2004 20:41:33 GMT, legg <legg@nospam.magma.ca> wrote:

You might find what you need in some of these. If you care to identify
the particular package that you are interested in, it might help.

[useful links snipped]

Thanks for the links.
Okay. I'm only talking about RF transistors here; nothing more
complicated than that. No specific package, but rather a variety of
the common types that are mostly used for encapsulating RF trannies.

I would expect that devices specifically designed to be applied in RF
applications will have these strays, as present in the recommended
board layout, in their manufacturer-generated models.

The only problem is with 'misapplied' devices in conventional
packaging.

If you have an application, then you will have a list of candidates
with specific package types that need to be addressed.

RL

RL