SynBERC:MIT/Calendar/2006-9-27: Difference between revisions
No edit summary |
|||
(2 intermediate revisions by 2 users not shown) | |||
Line 30: | Line 30: | ||
#[[User:JeremyZucker|JeremyZucker]] 19:16, 26 September 2006 (EDT) | #[[User:JeremyZucker|JeremyZucker]] 19:16, 26 September 2006 (EDT) | ||
#Ken Oye | #Ken Oye | ||
# Jake Beal | |||
==Notes== | |||
(by Reshma Shetty, feel free to modify) | |||
Logic devices | |||
#Performs logical operation | |||
#Power gain and signal restoration | |||
These two functions are often combined but don't necessarily need to be. | |||
===History of logic device families=== | |||
====RTL (resistor transistor logic) family==== | |||
*First integrated transistor family built on a single piece of silicon. | |||
*Resistor = logical operation | |||
*Transistor = signal restoration | |||
*didn't have perfect signal restoration | |||
*very slow | |||
*Tom drew a parts level diagram of a logic device. Depending on which resistor parts you chose, the device could either implement a NOR or NAND logic operation | |||
*very cheap which is why it took off | |||
====DTL (diode transistor logic) family==== | |||
*uses diodes instead | |||
*more robust control over voltage | |||
*also very slow | |||
*Motorola | |||
====TTL (transistor transistor logic) family==== | |||
*faster | |||
*push-pull | |||
*has standard power supply voltages and interfaces. Therefore, you could transition from one series to another. The series differed in speed and power dissipation. | |||
*Texas Instruments | |||
====CMOS transistors==== | |||
*made from P-MOS and N-MOS transistors | |||
*P-MOS transistors are naturally inverting | |||
*logic swings go from ground to positive power supply unlike previous logic families ... bigger swing. | |||
*RCA | |||
==Device families== | |||
Transitions from one family to another were driven by performance requirements and the technology. Transitions between device families were accompanied by transitions in related technologies like the power supply. | |||
How do you cope with poor signal restoration? | |||
*vary power supply voltage to change noise margins | |||
*it was hard to do good designs in RTL devices | |||
Fanout (~10 max) | |||
*Determined by the physical limitations of the device. It is not driven by the applications. | |||
Fanin (~8 max) | |||
*Determined by how many inputs the device is designed to take in. | |||
Wire-OR: a type of logic that may be important in biology | |||
*can do two stages of logic out of a single gate | |||
*have open collector gates | |||
*similar to letting a protein be expressed in either of two different ways | |||
Also see relevant papers on noise margins. | |||
How do you design reliable logic circuits? Gotten a lot of attention over the years. A lot of this complexity is hidden from the users. |
Latest revision as of 11:42, 27 September 2006
Anyone in the synthetic biology community is welcome to attend.
Wednesday at 1pm 32-D463, MIT ***NOTE ROOM CHANGE***
Topic of discussion
Tom Knight will be giving an introduction to different electronics device families and their specifications. This introduction should help to frame future lunch discussions regarding specification of biological device families.
Attending
Please list your name below if you are attending so that we order enough food. (To add your name to the list, type #~~~~
.)
- Reshma 14:29, 20 September 2006 (EDT)
- bruno 18:32, 22 September 2006 (EDT).
- Ilya 18:46, 22 September 2006 (EDT)
- David 23:56, 22 September 2006 (EDT)
- Austin Che 13:38, 23 September 2006 (EDT)
- Caroline10:26, 25 September 2006 (EDT)
- tk 13:09, 25 September 2006 (EDT)
- Endy 13:26, 25 September 2006 (EDT)
- hkeller 13:42, 25 September 2006 (EDT)
- Jbuck 13:50, 25 September 2006 (EDT)
- cmc 14:01, 25 September 2006 (EDT)
- BC 15:29, 25 September 2006 (EDT)
- Timp0 17:03, 25 September 2006 (EDT)
- Samantha 19:47, 25 September 2006 (EDT)
- Jason Kelly
- Julie Norville
- PatrickBoyle 11:09, 26 September 2006 (EDT)
- JeremyZucker 19:16, 26 September 2006 (EDT)
- Ken Oye
- Jake Beal
Notes
(by Reshma Shetty, feel free to modify)
Logic devices
- Performs logical operation
- Power gain and signal restoration
These two functions are often combined but don't necessarily need to be.
History of logic device families
RTL (resistor transistor logic) family
- First integrated transistor family built on a single piece of silicon.
- Resistor = logical operation
- Transistor = signal restoration
- didn't have perfect signal restoration
- very slow
- Tom drew a parts level diagram of a logic device. Depending on which resistor parts you chose, the device could either implement a NOR or NAND logic operation
- very cheap which is why it took off
DTL (diode transistor logic) family
- uses diodes instead
- more robust control over voltage
- also very slow
- Motorola
TTL (transistor transistor logic) family
- faster
- push-pull
- has standard power supply voltages and interfaces. Therefore, you could transition from one series to another. The series differed in speed and power dissipation.
- Texas Instruments
CMOS transistors
- made from P-MOS and N-MOS transistors
- P-MOS transistors are naturally inverting
- logic swings go from ground to positive power supply unlike previous logic families ... bigger swing.
- RCA
Device families
Transitions from one family to another were driven by performance requirements and the technology. Transitions between device families were accompanied by transitions in related technologies like the power supply.
How do you cope with poor signal restoration?
- vary power supply voltage to change noise margins
- it was hard to do good designs in RTL devices
Fanout (~10 max)
- Determined by the physical limitations of the device. It is not driven by the applications.
Fanin (~8 max)
- Determined by how many inputs the device is designed to take in.
Wire-OR: a type of logic that may be important in biology
- can do two stages of logic out of a single gate
- have open collector gates
- similar to letting a protein be expressed in either of two different ways
Also see relevant papers on noise margins. How do you design reliable logic circuits? Gotten a lot of attention over the years. A lot of this complexity is hidden from the users.