| Author |
Message |
analog_fever
Joined: 26 Dec 2008
Posts: 21
Location: USA
|
 03 Jan 2009 3:15 two stage op amp design |
|
|
|
|
I am a novice analog designer. I have some questions regarding the classic 2 stage CMOS opamp (or for that matter any analog ckt):
When one starts a design from scratch, how does one determine the device sizes to start with? Hand calculations? I tried following the procedure in Allen & Holberg for a 2-stage CMOS opamp, but found that with the calculated device sizes simulation is no where near the calculated performance.
I know most of the times designers pick an older circuit and work on making modifications to it. But how does one approach if he has to start from scratch?
|
|
| Back to top |
|
 |
Old Nick
Joined: 14 Sep 2007
Posts: 445
Helped: 49
|
03 Jan 2009 3:59 2 stage op amp design |
|
|
|
|
Choose you're minimum length to be a couple of times the minimum available in the process.
Choose the widths of, say the tail transistor, according to your current requirements.
Choose sensible overdrive voltages for your PMOS and NMOS devices, and you're pretty much there, bar compensation. A few sims later and some revisions to improve performance and you're ready to layout.
|
|
| Back to top |
|
|
henrywent
Joined: 29 Dec 2008
Posts: 80
Location: china
|
03 Jan 2009 7:44 two stage opamp design |
|
|
|
|
Old Nick,
what do you mean by " choose sensible overdrive voltages ", how much do you choose in your analog design for the sensible overdrive voltages. does 0.2v in 1.8v supply voltage sensible. Any comment is welcome.
|
|
| Back to top |
|
|
Old Nick
Joined: 14 Sep 2007
Posts: 445
Helped: 49
|
03 Jan 2009 16:57 2 stage opamp design |
|
|
|
|
| henrywent wrote: |
Old Nick,
what do you mean by " choose sensible overdrive voltages ", how much do you choose in your analog design for the sensible overdrive voltages. does 0.2v in 1.8v supply voltage sensible. Any comment is welcome. |
I've never used a 1.8V process, but you'll need a different overdrive voltage for P and N mos devices. Off the top of my head, I think I tend to use 250 mV and 350 mV ( for N and P respectively) for a 3.3 V process. So 200 mV sounds a bit high, although as I say , I've never used that process. It does deppend how aggressive you want to be with your values though.
As an exercise, choose a low Overdrive voltage and sim an amp, then choose a higher one and sim again and see the effect it has on the performance.
|
|
| Back to top |
|
|
analog_fever
Joined: 26 Dec 2008
Posts: 21
Location: USA
|
05 Jan 2009 19:29 two stage operational amplifier |
|
|
|
|
| Oldnick thanks for the response
|
|
| Back to top |
|
|
Google
AdSense
|
05 Jan 2009 19:29 Ads |
|
|
|
|
|
|
| Back to top |
|
|
egibson873
Joined: 20 Nov 2007
Posts: 2
|
06 Jan 2009 23:40 mos operational amplifier design |
|
|
|
|
analog_fever,
The first thing you need to decide when designing an opamp is what characteristics you want. If you want speed, you need smaller devices. If you want DC accuracy, you need bigger devices. There are other things you also have to worry about, like noise, power consumption, etc. that will also determine how big your devices need to be.
I guess the short answer to is to watch your overdrive voltages of your transistors, I usually shoot for 5-10% of the supply voltage, depending on how much speed I need to achieve and how much head room I am willing to give up.
|
|
| Back to top |
|
|
tlihu
Joined: 02 Jan 2002
Posts: 669
Helped: 5
|
07 Jan 2009 8:07 two stage op amp |
|
|
|
|
| I used to run .dc of the basic sizes of P-/N-MOS and list some of the parameters such as Gm, etc to skip hand calculation of device parameters. From there you can follow the textbook.
|
|
| Back to top |
|
|
