Hicum versus design simulators

1
Hicum versus design simulators

• Hicum Workshop
• June 2002

2
Basic DC tests

• 2 different sets of parameters, one with self
  heating.
• 2 test benches for each set of parameters at 27
  and 150ºC
• Forward Gummel simulation sweep at various Vbc
  for
• the model with self-heating.
• Ic vs. Vce characteristic at Ibcst.
• The comparisons are made with Device, Pr.
  Schroters simulator.
• The tested simulators are Eldo, Spectre,
  Aplac, Hspice and ADS.
• gt ST Target less than 0.1 difference.

3
Tools tested
Simulator Tested version Availability
Eldo 5.8_1.1 (Beta Version) Forecasted this summer
Spectre IC446 MSR7 Available
Aplac 7.62e Available
Hspice Beta 2002_2 Forecasted June 02
ADS 2002 Available
4
Device/Eldo Gummel Plot without self-heating
temp 150ºC
temp 27ºC
Max of error 0.1
Max of error 1
5
Device/Eldo Gummel Plot with self heating
Device Eldo
temp 27ºC
temp 150ºC
Max of error lt0.1
Max of error 0.02
6
Device/Eldo Ic vs Vce Ibcst Plot without
self-heating
temp 150ºC
temp 27ºC
Max of error 0.03
Max of error 0
7
Device/Eldo Ic vs Vce Ibcst Plot with self
heating
Device Eldo
temp 150ºC
Device Eldo
temp 27ºC
Max of error 0.01
Max of error 0.01
8
Device/Spectre Gummel Plot without self-heating
temp 150ºC
temp 27ºC
Max of error 0.1
Max of error 0.03
9
Device/Spectre Gummel Plot with self heating
temp 150ºC
temp 27ºC
Max of error gt 10 From 0.4V lt 0.1
Max of error lt 0.1
10
Device/Spectre Ic vs Vce Ibcst Plot without
self-heating
temp 27ºC
temp 150ºC
Max of error 0.03
Max of error gt 20 From 0.4V max of error lt
0.1
11
Device/Spectre Ic vs Vce Ibcst Plot with
self-heating
temp 150ºC
temp 27ºC
Max of error gt 10 From 0.5V lt 0.01
Max of error 0.1 From 0.4V lt0.01
12
Device/Aplac Gummel Plot without self-heating
temp 150ºC
temp 27ºC
Max of error gt 10
Max of error 10 lt1.2V lt 0.1
13
Device/Aplac Gummel Plot with self-heating
temp 27ºC
temp 150ºc
Max of error gt 10
Max of error 10
14
Device/Aplac Ic vs Vce Ibcst Plot without
self-heating
temp 27ºC
temp 150ºC
Max of error 5
Max of error 1 From 0.2V 0.03
15
Device/Aplac Ic vs Vce Ibcst Plot with
self-heating
temp 150ºC
temp 27ºC
Max of error 10
16
Device/Hspice Gummel Plot without self-heating
temp 150ºC
temp 27ºC
Max of error 10 lt 0.9V lt 0.01
Max of error 10 lt 0.9V lt 0.01
17
Device/Hspice Ic vs Vce Ibcst Plot without
self-heating
temp 150ºC
temp 27ºC
Max of error 0.8 From 0.3V lt0.02
Max of error 2
18
Device/ADS Gummel Plot without self-heating
temp 150ºC
temp 27ºC
Max of error gt 10 lt 0.8V lt 0.1
Max of error 10 lt 0.9V 0.2
19
Device/ADS Gummel Plot with self heating
temp 150ºC
temp 27ºC
Max of error 0.8VltVbelt0.9V lt 1
Max of error gt 10 lt 0.8V lt0.03
20
Device/ADS Ic vs Vce Ibcst Plot without
self-heating
temp 150ºC
temp 27ºC
Max of error gt 10 From 0.3V 2
Max of error 1 From 0.3V 0.1
21
Device/ADS Ic vs Vce Ibcst Plot with
self-heating
temp 150ºC
temp 27ºC
Max of error gt 10 From 0.3V 10
22
Comparison Sum-up
Simulator Tested version Comments
Eldo 5.8_1.1 Delivery expected in summer Good results in DC.
Spectre IC446 MSR7 Available Good results at 27ºC, discrepancies at 150 ºC
Aplac 7.62e Available Still too big discrepancies especially at 150 ºC
Hspice Beta 2002_2 2002_2 available june 02 Improvements with no self-heating, but still some discrepancies. Self-heating to be tested soon by the end of the summer
ADS 2002 Available Still discrepancies big problems at 150 ºC