IEEE 1076.1 Working Group Meeting Minutes 10 October 2001 ========================================================== Meeting attendees: Jim Barby jabarby@uwaterloo.ca Peter Frey pfrey@cadence.com Alan Mantooth (chair) mantooth@engr.uark.edu Monte Mar mmar@orora.com Darrell Teegarden darrell_teegarden@mentor.com Meeting agenda -------------- 1. Introductions 2. Brief status of working group formation and schedule 3. Review of each proposed package a. Motivation and Introduction b. Electrical c. Mechanical d. Fluidic e. Thermal f. Radiant g. Energy 4. Review action items 5. Next meeting Alan opened the meeting at 2:15pm. 1. Introductions ---------------- Each attendee spent just a few minutes to introduce themselves to the group including their current job function and interest in VHDL-AMS package standardization. 2. Brief status of working group formation and schedule ------------------------------------------------------- Alan described the formation of the working group, described the progress towards getting the PAR form approved, and outlined the intended schedule for proceeding to ballot with a standard set of packages. The intent is to take roughly one year for discussion and attempt to ballot shortly thereafter. This will depend on participation and regularity of meetings, of course. 3. Review of each proposed package ---------------------------------- Alan proceeded to lead the discussion of each of the six standard packages as they exist in draft form. These include: . energy_systems . electrical_systems (inclusive of magnetic) . mechanical_systems . fluidic_systems . thermal_systems . radiant_systems Ernst Christen had made some suggested improvements to the standard packages to Alan for the meeting. These were discussed. The suggestions fell into the following categories: . formatting . subtype declarations for composite (i.e., vector) types . some additional SI prefixes (engineering scale factors) to energy_systems . converted symbol strings to all lower case (e.g., "Coulomb" to "coulomb") according to SI spec . removal of subtype declarations for resistance/capacitance. The group decided that the original text formatting was more acceptable, so the decision was made to stick to that for now (issue of VHDL keywords being capitalized versus lower-case, etc.). The suggestion to remove resistance and capacitance was rejected as well. The rationale being simply that it doesn't hurt to have them declared and available. All other suggestions were accepted and it was decided to propagate these suggestions throughout all of the packages were applicable. energy_systems -------------- Primary highlights of the discussion here centered on the addition of a few more scale factors taking the range from 1e-24 to 1e+24 and, more importantly, the choice of standard constants. Ernst Christen had previously pointed out via e-mail that the values of certain material constants vary from source to source. His latest suggestions include comments about each constant as follows in this code snippet: -- Boltzmann's constant -- NIST value: 1.380 6503e-23 joules/kelvin -- uncertainty: 0.000 0024e-23 constant k : REAL := 1.380_6503e-23; This form has been adopted, but it leaves open the issues of which constants to include when different values exist. For example, a simple web search reveals several values for the relative permittivity of Silicon and SiO2. These values have been disputed, as there are models that use different values such as 11.8. A search on the web gave as many values for EPS_SI as I had hits, with values ranging from 11 to 14. The suspicion is that this may have to do with how the value was measured, maybe also with purity, and maybe the material is not isotropic. Ernst has recommended that these material constants/properties be removed from the standard packages, since they are subjective. Further discussion is required on this issue. electrical_systems ------------------ In the declaration of vector subtypes, a question was raised by Darrell Teegarden. Which is correct? subtype VOLTAGE_VECTOR is ELECTRICAL_VECTOR'across; OR subtype VOLTAGE_VECTOR is ARRAY (Natural RANGE <>) of VOLTAGE; The feeling was that the first was the better declaration (as indicated by Ernst Christen in his proposed updates) because it was self-correcting with respect to any changes that might occur in the electrical nature definition (however unlikely). Peter Wilson had suggested that while "ampere-turns" is commonly used for mmf, that since turns was really dimensionless it might be a better suggestion to simply use "ampere" for the symbol of mmf. No decision was made on this and it is left as is for further discussion. As noted before, resistance, capacitance and inductance were kept in the package. A significant discussion of the magnetic nature occured prior to the meeting on the reflector. Siep Onneweer brought up the idea of using flux rate as the through variable for magnetic systems rather than flux. Peter Wilson took the opposing view. Ernst Christen has proposed a way of evaluating these two views based on simulation results (posted to the reflector). No conclusions were reached at the meeting. mechanical_systems ------------------ Add the vector subtypes consistent with electrical package and for tranlational_v and rotational_v natures change the vector natures to TRANSLATIONAL_V_VECTOR from TRANSLATIONALV_VECTOR and likewise for ROTATIONAL_V_VECTOR. fluidic_systems --------------- Update with vector subtypes and attributes. Determine experts in this area to review package. thermal_systems --------------- Update with vector subtypes and attributes. Another issue raised by Darrell Teegarden was to consider having multiple natures: one with heat_flow as the through variable and the other with entropy as the through variable. Open for discussion at this point. radiant_systems --------------- It was suggested to inquire with Allen Hefner at NIST for someone more associated with this field to help define the package. 4. Review action items and open issues -------------------------------------- 1. Conclude on using flux vs. flux rate in magnetic nature 2. Add vector subtypes for all packages analogous to voltage and current in electrical_system. 3. Make symbol attributes SI-compatible and decide on name vs. symbol or both (e.g., "meter" OR "m" or both) 4. Material properties inclusion in energy_systems (included at present) 5. Names for physical constants in energy_systems (e.g., k for Boltzmann's constant vs. k_boltzmann or phys_boltzmann) 5. Next Meeting --------------- The date of the next working group meeting is not yet fixed. DATE & HDLcon in March 2002 is a possible time frame. Alan adjourned the meeting at 5:00pm.