LTSpice and OpAmp Oscillation

Started by M. Hamed May 31, 2009
I've built a discrete transistor OpAmp, and when the input was off, I
saw it oscillating. So, I got interested in OpAmp oscillation and
compensation. The main criterion is that a Phase shift of 180 degrees
occur when the loop gain is higher than 1.

I tried to demonstrate that in LTSpice but couldn't. I'm not sure if
it's an LTSpice limitation, or something I'm doing.

I opened up the included exampled circuit for the LM741, and removed
the compensation capacitor. I ran a frequency response and it showed
that around 18 MHz, the phase shift is -180 and the open loop gain is
greater than unity, so connecting it as a follower, the loop gain
should be still greater than unity.

I turned off the AC source and ran simulation expecting it will
oscillate (after all LTSpice can do other types of oscillators). The
circuit failed to oscillate however, no matter what I did. I tried
different things, but in vain, I couldn't get it to oscillate. In fact
at 18 MHz, even the basic follower functionality wasn't working.

I wonder if somebody can comment on this cheesy experiment, and let me
know what's going on? is it me? or LTSpice? or things are just more
complicated than explained in AoE?

Regards,
"M. Hamed" <mhelshou@hotmail.com> wrote in message 
news:fd5d659d-4d4f-4cd8-ae2c-754db9be9e64@z9g2000yqi.googlegroups.com...
> I've built a discrete transistor OpAmp, and when the input was off, I > saw it oscillating. So, I got interested in OpAmp oscillation and > compensation. The main criterion is that a Phase shift of 180 degrees > occur when the loop gain is higher than 1. > > I tried to demonstrate that in LTSpice but couldn't. I'm not sure if > it's an LTSpice limitation, or something I'm doing. > > I opened up the included exampled circuit for the LM741, and removed > the compensation capacitor. I ran a frequency response and it showed > that around 18 MHz, the phase shift is -180 and the open loop gain is > greater than unity, so connecting it as a follower, the loop gain > should be still greater than unity. > > I turned off the AC source and ran simulation expecting it will > oscillate (after all LTSpice can do other types of oscillators). The > circuit failed to oscillate however, no matter what I did. I tried > different things, but in vain, I couldn't get it to oscillate. In fact > at 18 MHz, even the basic follower functionality wasn't working. > > I wonder if somebody can comment on this cheesy experiment, and let me > know what's going on? is it me? or LTSpice? or things are just more > complicated than explained in AoE? > > Regards,
Oscillators sometimes need help to get started in spice. Leave the AC source; but try setting Ncycles to a small number.
"Andrew Holme" <ah@nospam.co.uk> wrote in message 
news:pHCUl.320422$0V4.140750@newsfe25.ams2...
> > "M. Hamed" <mhelshou@hotmail.com> wrote in message > news:fd5d659d-4d4f-4cd8-ae2c-754db9be9e64@z9g2000yqi.googlegroups.com... >> I've built a discrete transistor OpAmp, and when the input was off, I >> saw it oscillating. So, I got interested in OpAmp oscillation and >> compensation. The main criterion is that a Phase shift of 180 degrees >> occur when the loop gain is higher than 1. >> >> I tried to demonstrate that in LTSpice but couldn't. I'm not sure if >> it's an LTSpice limitation, or something I'm doing. >> >> I opened up the included exampled circuit for the LM741, and removed >> the compensation capacitor. I ran a frequency response and it showed >> that around 18 MHz, the phase shift is -180 and the open loop gain is >> greater than unity, so connecting it as a follower, the loop gain >> should be still greater than unity. >> >> I turned off the AC source and ran simulation expecting it will >> oscillate (after all LTSpice can do other types of oscillators). The >> circuit failed to oscillate however, no matter what I did. I tried >> different things, but in vain, I couldn't get it to oscillate. In fact >> at 18 MHz, even the basic follower functionality wasn't working. >> >> I wonder if somebody can comment on this cheesy experiment, and let me >> know what's going on? is it me? or LTSpice? or things are just more >> complicated than explained in AoE? >> >> Regards, > > Oscillators sometimes need help to get started in spice. Leave the AC > source; but try setting Ncycles to a small number. >
Oscillators need noise to get them going. Some basic SPICE packages do not include the noise sources, so your SPICE circuit will just sit there and sulk. -- Bill Naylor www.electronworks.co.uk Electronic Kits for Education and Fun
M. Hamed wrote:
> I've built a discrete transistor OpAmp, and when the input was off, I > saw it oscillating. So, I got interested in OpAmp oscillation and > compensation. The main criterion is that a Phase shift of 180 degrees > occur when the loop gain is higher than 1.
Not strictly correct. This is a necessary condition, but not sufficient. In most cases it is correct.
> > I tried to demonstrate that in LTSpice but couldn't. I'm not sure if > it's an LTSpice limitation, or something I'm doing. > > I opened up the included exampled circuit for the LM741, and removed > the compensation capacitor. I ran a frequency response and it showed > that around 18 MHz, the phase shift is -180 and the open loop gain is > greater than unity, so connecting it as a follower, the loop gain > should be still greater than unity. > > I turned off the AC source and ran simulation expecting it will > oscillate (after all LTSpice can do other types of oscillators). The > circuit failed to oscillate however, no matter what I did. I tried > different things, but in vain, I couldn't get it to oscillate. In fact > at 18 MHz, even the basic follower functionality wasn't working. > > I wonder if somebody can comment on this cheesy experiment, and let me > know what's going on? is it me? or LTSpice? or things are just more > complicated than explained in AoE? >
Switch on the power supply with a pulse of 0 to Vsupply to get any oscilllations started. Kevin Aylward SuperSpice www.anasoft.co.uk
M. Hamed wrote:
> I've built a discrete transistor OpAmp, and when the input was off, I > saw it oscillating. So, I got interested in OpAmp oscillation and > compensation. The main criterion is that a Phase shift of 180 degrees > occur when the loop gain is higher than 1.
Not strictly correct. This is a necessary condition, but not sufficient. In most cases it is correct.
> > I tried to demonstrate that in LTSpice but couldn't. I'm not sure if > it's an LTSpice limitation, or something I'm doing. > > I opened up the included exampled circuit for the LM741, and removed > the compensation capacitor. I ran a frequency response and it showed > that around 18 MHz, the phase shift is -180 and the open loop gain is > greater than unity, so connecting it as a follower, the loop gain > should be still greater than unity. > > I turned off the AC source and ran simulation expecting it will > oscillate (after all LTSpice can do other types of oscillators). The > circuit failed to oscillate however, no matter what I did. I tried > different things, but in vain, I couldn't get it to oscillate. In fact > at 18 MHz, even the basic follower functionality wasn't working. > > I wonder if somebody can comment on this cheesy experiment, and let me > know what's going on? is it me? or LTSpice? or things are just more > complicated than explained in AoE? >
Switch on the power supply with a pulse of 0 to Vsupply to get any oscilllations started. Kevin Aylward SuperSpice www.anasoft.co.uk
"Andrew Holme" <ah@nospam.co.uk> wrote in message 
news:pHCUl.320422$0V4.140750@newsfe25.ams2...
> > "M. Hamed" <mhelshou@hotmail.com> wrote in message > news:fd5d659d-4d4f-4cd8-ae2c-754db9be9e64@z9g2000yqi.googlegroups.com... >> I've built a discrete transistor OpAmp, and when the input was off, I >> saw it oscillating. So, I got interested in OpAmp oscillation and >> compensation. The main criterion is that a Phase shift of 180 degrees >> occur when the loop gain is higher than 1. >> >> I tried to demonstrate that in LTSpice but couldn't. I'm not sure if >> it's an LTSpice limitation, or something I'm doing. >> >> I opened up the included exampled circuit for the LM741, and removed >> the compensation capacitor. I ran a frequency response and it showed >> that around 18 MHz, the phase shift is -180 and the open loop gain is >> greater than unity, so connecting it as a follower, the loop gain >> should be still greater than unity. >> >> I turned off the AC source and ran simulation expecting it will >> oscillate (after all LTSpice can do other types of oscillators). The >> circuit failed to oscillate however, no matter what I did. I tried >> different things, but in vain, I couldn't get it to oscillate. In fact >> at 18 MHz, even the basic follower functionality wasn't working. >> >> I wonder if somebody can comment on this cheesy experiment, and let me >> know what's going on? is it me? or LTSpice? or things are just more >> complicated than explained in AoE? >> >> Regards, > > Oscillators sometimes need help to get started in spice. Leave the AC > source; but try setting Ncycles to a small number. >
Oscillators need noise to get them going. Some basic SPICE packages do not include the noise sources, so your SPICE circuit will just sit there and sulk. -- Bill Naylor www.electronworks.co.uk Electronic Kits for Education and Fun
"M. Hamed" <mhelshou@hotmail.com> wrote in message 
news:fd5d659d-4d4f-4cd8-ae2c-754db9be9e64@z9g2000yqi.googlegroups.com...
> I've built a discrete transistor OpAmp, and when the input was off, I > saw it oscillating. So, I got interested in OpAmp oscillation and > compensation. The main criterion is that a Phase shift of 180 degrees > occur when the loop gain is higher than 1. > > I tried to demonstrate that in LTSpice but couldn't. I'm not sure if > it's an LTSpice limitation, or something I'm doing. > > I opened up the included exampled circuit for the LM741, and removed > the compensation capacitor. I ran a frequency response and it showed > that around 18 MHz, the phase shift is -180 and the open loop gain is > greater than unity, so connecting it as a follower, the loop gain > should be still greater than unity. > > I turned off the AC source and ran simulation expecting it will > oscillate (after all LTSpice can do other types of oscillators). The > circuit failed to oscillate however, no matter what I did. I tried > different things, but in vain, I couldn't get it to oscillate. In fact > at 18 MHz, even the basic follower functionality wasn't working. > > I wonder if somebody can comment on this cheesy experiment, and let me > know what's going on? is it me? or LTSpice? or things are just more > complicated than explained in AoE? > > Regards,
Oscillators sometimes need help to get started in spice. Leave the AC source; but try setting Ncycles to a small number.
I've built a discrete transistor OpAmp, and when the input was off, I
saw it oscillating. So, I got interested in OpAmp oscillation and
compensation. The main criterion is that a Phase shift of 180 degrees
occur when the loop gain is higher than 1.

I tried to demonstrate that in LTSpice but couldn't. I'm not sure if
it's an LTSpice limitation, or something I'm doing.

I opened up the included exampled circuit for the LM741, and removed
the compensation capacitor. I ran a frequency response and it showed
that around 18 MHz, the phase shift is -180 and the open loop gain is
greater than unity, so connecting it as a follower, the loop gain
should be still greater than unity.

I turned off the AC source and ran simulation expecting it will
oscillate (after all LTSpice can do other types of oscillators). The
circuit failed to oscillate however, no matter what I did. I tried
different things, but in vain, I couldn't get it to oscillate. In fact
at 18 MHz, even the basic follower functionality wasn't working.

I wonder if somebody can comment on this cheesy experiment, and let me
know what's going on? is it me? or LTSpice? or things are just more
complicated than explained in AoE?

Regards,