Author Topic: How do i Model This Part in LTSpice IV?  (Read 7850 times)

0 Members and 1 Guest are viewing this topic.

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9411
  • Country: my
  • reassessing directives...
How do i Model This Part in LTSpice IV?
« on: May 06, 2011, 07:03:00 am »
i want to design circuit using these Texas Instruments chips:
THS3202, 2GHz, 9KV/us, ±7.5Vss
THS3061, 300MHz, 7KV/us, ±15Vss

but first i have to simulate it in LTSpice IV (or any better suggestion?). since LTSpice only provide Linear Technology chips, so TI chips are not available in the library. i dont have the TI chip yet, so i dont know how they will behave. worst the above chips are "Current Feedback Operational Amplifier", which is quite new to me at least, and not behave like a normal "Voltage Feedback Amplifier".

1) any links to the TI chip library that can be used in LTSpice will be appreciated.
2) How do i create a new component/library in LTSpice if i have no knowledge of the real component behaviour? and the simulation output as close as the real component/circuit?

thanx.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline allanw

  • Frequent Contributor
  • **
  • Posts: 343
    • Electronoblog
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9411
  • Country: my
  • reassessing directives...
Re: How do i Model This Part in LTSpice IV?
« Reply #2 on: May 06, 2011, 07:28:07 am »
ahh that! thanx alot! i've been browsing TI site in pain. still cant get everything right... me. what a slow server! :(
edit: downloading the tina...... 1%...
« Last Edit: May 06, 2011, 07:31:50 am by Mechatrommer »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9411
  • Country: my
  • reassessing directives...
Re: How do i Model This Part in LTSpice IV?
« Reply #3 on: May 06, 2011, 12:32:34 pm »
this is cool yo! i got a new 'assy'lloscope!
it seems i'm pushing the 3062 to its GBW limit.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9411
  • Country: my
  • reassessing directives...
Re: How do i Model This Part in LTSpice IV?
« Reply #4 on: May 06, 2011, 01:10:21 pm »
but wait. 3202 datasheet said, Vpp max = ±7.5V. i just supply it with ±12V but the sim didnt smoke it. this seems (picture) more realistic. but i wonder, 3062 seems immune to distortion even with the 10X gain (exceeded its GBW far enough). o well, we'll see in the real life.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline allanw

  • Frequent Contributor
  • **
  • Posts: 343
    • Electronoblog
Re: How do i Model This Part in LTSpice IV?
« Reply #5 on: May 06, 2011, 01:34:46 pm »
Did you see this? Nice warning right after the introduction section:

 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9411
  • Country: my
  • reassessing directives...
Re: How do i Model This Part in LTSpice IV?
« Reply #6 on: May 06, 2011, 03:08:30 pm »
yes i just read that few minutes ago. i just got excited with tina earlier. i guess 3095 is the way (while stock last, 3091 is out of stock). i'm in the middle simulating 3095 right now.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline tecman

  • Frequent Contributor
  • **
  • Posts: 434
  • Country: us
Re: How do i Model This Part in LTSpice IV?
« Reply #7 on: May 06, 2011, 05:51:31 pm »
What are you doing that requires such hot (high frequency) chips.  If it is audio stuff, there are a number of other ultra low distortion, standard voltage mode chips on the market.

Although they are called current mode op-amps, I do not believe that they are what was called Norton amplifiers where you match, or mirror the input currents.  They appear to sum, similar to a standard op-amp.  Be careful, however, since the input differential voltage range for the inputs is quite small, compared to a normal op-amp.

paul
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9411
  • Country: my
  • reassessing directives...
Re: How do i Model This Part in LTSpice IV?
« Reply #8 on: May 06, 2011, 07:57:18 pm »
What are you doing that requires such hot (high frequency) chips.  If it is audio stuff, there are a number of other ultra low distortion, standard voltage mode chips on the market.
now i still get this question. ??? i need an amplifier to increase the Vpp of the output of this...

i will externally supply it with ±12V. allanw and others at Op Amp (uA741CN) Help have been around guiding me.

Be careful, however, since the input differential voltage range for the inputs is quite small, compared to a normal op-amp.
the 3095 datasheet says... Differential input voltage Vid = ±4V should be ok i think. The Hantek's output is max at ±3.5V
« Last Edit: May 06, 2011, 08:00:24 pm by Mechatrommer »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9411
  • Country: my
  • reassessing directives...
Re: How do i Model This Part in LTSpice IV?
« Reply #9 on: May 06, 2011, 08:21:07 pm »
there is in THS3095 datasheet (picture) saying the SPICE only model within the chip non-saturation region, and it will perform poorly in saturated/distorted region. maybe that explained why the THS3062 didnt distort in the above sim. there's also mentioning about small signal AC, so not sure what small is small.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 13449
  • Country: gb
  • 0999
Re: How do i Model This Part in LTSpice IV?
« Reply #10 on: May 06, 2011, 08:36:50 pm »
I've noticed that all of the op-amps specified so far are compensated to be stable at unity gain but is this really important?

If the minimum gain requirement is >1 then why not go for an uncompensated amplifier and add the required capacitance to make it stable at the desired gain?
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9411
  • Country: my
  • reassessing directives...
Re: How do i Model This Part in LTSpice IV?
« Reply #11 on: May 06, 2011, 09:02:11 pm »
I've noticed that all of the op-amps specified so far are compensated to be stable at unity gain but is this really important?
If the minimum gain requirement is >1 then why not go for an uncompensated amplifier and add the required capacitance to make it stable at the desired gain?
someone should tell us why. i guess its easier for a layman like me to estimate the BW of an opamp from the gain required. all i have right now is reading the wiki:
Frequency compensation
Current-feedback operational amplifier
Gain–bandwidth product
it is told that CFB opamp is much faster than VFB opamp.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline allanw

  • Frequent Contributor
  • **
  • Posts: 343
    • Electronoblog
Re: How do i Model This Part in LTSpice IV?
« Reply #12 on: May 06, 2011, 10:54:02 pm »
True, a de-compensated amplifier would be more useful. They are definitely less available than the unity-gain stable ones though. I don't think I've seen any at high bandwidth/slew rate.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 13449
  • Country: gb
  • 0999
Re: How do i Model This Part in LTSpice IV?
« Reply #13 on: May 07, 2011, 05:05:14 pm »
Here are some uncompensated amplifier ICs I've found using Google. I don't know what your exact requirements are (too many threads) but hopefully one of them will do.

THS4021
HA-2625
OPA685
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9411
  • Country: my
  • reassessing directives...
Re: How do i Model This Part in LTSpice IV?
« Reply #14 on: May 07, 2011, 06:24:52 pm »
thanx hero for the pointers. i will study them. this thread particularly for simulating those opamp. those opamp (topmost) are among the top "high slew rate" AND "high GBW" spec i can find (and available in ebay too, which is my "easy" shopping place). the purpose and why of that is discussed in another thread (evolved from uA741 :P)
« Last Edit: May 07, 2011, 06:28:14 pm by Mechatrommer »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf