[ibis-macro] Re: August 30 Minutes
- From: "Mike LaBonte (milabont)" <milabont@xxxxxxxxx>
- To: <ibis-macro@xxxxxxxxxxxxx>
- Date: Wed, 31 Aug 2005 12:25:10 -0400
It's hard to explain why I wrote that we would implement a 1-input
adder, except that I was thinking in terms of how many ports there would
be IN ADDITION to the first one. OK, so I guess we decided to implement
only the basic 2-input arithmetic building blocks, and cascade them if
we need more input ports. And I can't remember why I though only the 2nd
adder input would be scaled.
I have attached a corrected file. It's a good thing someone is checking
up on me ;)
Mike
-----Original Message-----
From: ibis-macro-bounce@xxxxxxxxxxxxx
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Muranyi, Arpad
Sent: Wednesday, August 31, 2005 11:34 AM
To: ibis-macro@xxxxxxxxxxxxx
Subject: [ibis-macro] Re: August 30 Minutes
Mike,
Just a little correction. We decided that the adder, multiplier and
divider will have two input pairs, not 1-3 as you wrote it in the
minutes. Also, I don't remember hearing that the adder should scale the
second port. I implemented it as:
Scale*(A+B)
Do we really want:
A + Scale * B
as stated in the minutes?
Thanks,
Arpad
============================================
-----Original Message-----
From: ibis-macro-bounce@xxxxxxxxxxxxx
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Mike LaBonte
(milabont)
Sent: Wednesday, August 31, 2005 6:18 AM
To: ibis-macro@xxxxxxxxxxxxx
Subject: [ibis-macro] August 30 Minutes
meeting date: 30 aug 2005
attending: Arpad Muranyi, Todd Westerhoff, Ian Dodd, Bob Ross, Barry
Katz,
Ken Willis, Sam Chitwood, Mike LaBonte
-------------
AR: Ian to continue researching receiver architectures and come up with
a generic template
- Verbally sketched out, details TBD
AR: Arpad to post a note to the reflector to continue this discussion
- Done, but not much discussion.
AR: Todd/Mike to propose paper for IBIS Summit and develop presentation
-------------
Passing parameters:
- Ian: Arpad not using AMS the way it was intended.
- Something about steam engines...
- Parameters are for initialization only, ports are for runtime.
- Arpad: Voltage on a port can be converted to resistance value ...
- Todd: do we want ports to declare types like 'resistance'?
- No, all ports will be 'electrical'
Latest library from Arpad:
- C is charge conserving; changing C changes V
- L is flux conserving; changing L changes I
- If we want PWL VCVS, etc. we need to clone a template for each
instance.
- Table data can be pasted directly into the clone instead of using an
external file.
- Should we have add and multiply modules? yes
- How many inputs?
- Use port arrays?
- No, implement 1-3 inputs, fixed
- Use Scale param for both, can't be changed on the fly.
- Adder only scales 2nd port
- multiply multiplies and scales both
- Divider? yes
- Polynomial module? may be more complicated than we need
- Could have coefficient and exponent params to make it replace adder,
multiplier, and divider. Nah.
- Philosophy is to have simple module, even if we end up with more of
them.
- Ideal T-line? yes
- delay param or length and Tpd? delay
- s-param?
- make it a future enhancement
- AMS languages don't have it
- IBIS buffers
- series models?
- AMS multi-dimensional table could implement series mosfet
- AMS interpolates, can other simulators
Any other modules needed?
- triggerable input?
- PWL has it
- May not be supported by all simulators
- We'll put it off for now
- We should add new ones only as they are needed for new templates
- Can Cadence contribute more templates?
- Need a website for people to share modules
- Need to version number library files
Arpad needs help annotating the modules
AR: Mike will annotate one module, get comments, do the rest
AR: Todd and Mike start presentation
AR: Ken find out if cadence can contribute another template
-------------
Next meeting: Tuesday September 6, 2005.
---------------------------------------------------------------------
IBIS Macro website: http://www.sisoft.com/ibis-macro IBIS Macro
archives: //www.freelists.org/archives/ibis-macro
To unsubscribe send an email:
To: ibis-macro-request@xxxxxxxxxxxxx
Subject: unsubscribe
meeting date: 30 aug 2005
attending: Arpad Muranyi, Todd Westerhoff, Ian Dodd, Bob Ross, Barry Katz,
Ken Willis, Sam Chitwood, Mike LaBonte
-------------
AR: Ian to continue researching receiver architectures and come up with a
generic template
- Verbally sketched out, details TBD
AR: Arpad to post a note to the reflector to continue this discussion
- Done, but not much discussion.
AR: Todd/Mike to propose paper for IBIS Summit and develop presentation
-------------
Passing parameters:
- Ian: Arpad not using AMS the way it was intended.
- Something about steam engines...
- Parameters are for initialization only, ports are for runtime.
- Arpad: Voltage on a port can be converted to resistance value ...
- Todd: do we want ports to declare types like 'resistance'?
- No, all ports will be 'electrical'
Latest library from Arpad:
- C is charge conserving; changing C changes V
- L is flux conserving; changing L changes I
- If we want PWL VCVS, etc. we need to clone a template for each instance.
- Table data can be pasted directly into the clone instead of using an
external file.
- Should we have add and multiply modules? yes
- How many inputs?
- Use port arrays?
- No, implement 2 inputs, fixed
- Can cascade modules to get more inputs
- Use Scale param for both, can't be changed on the fly.
- Divider? yes
- Polynomial module? may be more complicated than we need
- Could have coefficient and exponent params to make it replace adder,
multiplier, and divider. Nah.
- Philosophy is to have simple module, even if we end up with more of them.
- Ideal T-line? yes
- delay param or length and Tpd? delay
- s-param?
- make it a future enhancement
- AMS languages don't have it
- IBIS buffers
- series models?
- AMS multi-dimensional table could implement series mosfet
- AMS interpolates, can other simulators
Any other modules needed?
- triggerable input?
- PWL has it
- May not be supported by all simulators
- We'll put it off for now
- We should add new ones only as they are needed for new templates
- Can Cadence contribute more templates?
- Need a website for people to share modules
- Need to version number library files
Arpad needs help annotating the modules
AR: Mike will annotate one module, get comments, do the rest
AR: Todd and Mike start presentation
AR: Ken find out if cadence can contribute another template
-------------
Next meeting: Tuesday September 6, 2005.
(minutes corrected to reflect feedback from Arpad)
Other related posts:
