| Author |
Message |
pbs681
Joined: 19 Aug 2004
Posts: 154
Helped: 11
|
 13 Aug 2008 12:38 Substrate current effect |
|
|
|
Hi all,
What is the effect of the substrate current in analog design domain?
Thanks
|
|
| Back to top |
|
 |
jecyhale
Joined: 19 Feb 2008
Posts: 447
Helped: 46
Location: China
|
| 19 Aug 2008 10:52 Re: Substrate current effect |
|
|
|
| pbs681 wrote: |
Hi all,
What is the effect of the substrate current in analog design domain?
Thanks |
Substrate current? what is it?
Substrate should be connected to good gnd.
Do you mean the substrate noise current?
If yes, add more isolate guarding ring.
|
|
| Back to top |
|
|
Colbhaidh
Joined: 10 Aug 2004
Posts: 151
Helped: 16
|
| 19 Aug 2008 15:44 Re: Substrate current effect |
|
|
|
Substrate current is a measure of hot electrons injected into the Wells during worst case conditions between gate and drain voltages, for hot electron generation at the drain. If the substrate current is high, then it is generally considered that hot electrons are also being injected into the gate and degrading the quality of the oxide causing threshold shifts and Gm degradation over time. This is called hot carrier injection or HCI. Every Fab process will have a quality specificqtion that the Gm will not degrade worse than 10% over 10 years of operation and worst case conditions. Most Fabs far exceed this and for pmos, may well surpass 100 years before the degradation is seen.
For analogue applications, it has severe impact on matching if the transistors become degraded.
Another hidden effect is when the Wells are not adequately contacted to Vss or Vdd.
Substrate current (possibly from an unrelated source) can change the actual Well bias close to critical transistors but far from the Well tap connections. This causes their Vts to change during the substrate current event. This can be very hard to trace as it is a "soft failure" i.e. it does not happen all the time.
|
|
| Back to top |
|
|
protonixs
Joined: 06 Jul 2007
Posts: 122
Helped: 14
Location: SG
|
| 20 Aug 2008 7:35 Re: Substrate current effect |
|
|
|
| there must be no substrate current. otherwise latch up will be triggered.
|
|
| Back to top |
|
|
bicave
Joined: 19 Jun 2006
Posts: 60
|
| 27 Aug 2008 10:10 Re: Substrate current effect |
|
|
|
| High electrical field between Drain-Source cause collssion and generate electron, the phenomenon is hot carrier. Some electrons go to substrate.
|
|
| Back to top |
|
|
pbs681
Joined: 19 Aug 2004
Posts: 154
Helped: 11
|
| 27 Aug 2008 13:57 Re: Substrate current effect |
|
|
|
| protonixs wrote: |
| there must be no substrate current. otherwise latch up will be triggered. |
there is always a substrate current in MOSFET. The explanation by Colbhaidh is correct. Substrate current will cause Reliability issue of the transistor by changing Vt.
Added after 2 minutes:
| bicave wrote: |
| High electrical field between Drain-Source cause collssion and generate electron, the phenomenon is hot carrier. Some electrons go to substrate. |
High electric field will generate electron-hole pair at the drain. The electron will be injected into oxide while the hole will be swept into substrate. So, substrate current is caused by hole (NMOS) not by electron.
|
|
| Back to top |
|
|
selvaraja
Joined: 01 Jun 2004
Posts: 220
Helped: 3
|
| 19 Sep 2008 4:32 Re: Substrate current effect |
|
|
|
| substarte current is form of noise ...they callled substrate noise.
|
|
| Back to top |
|
|
