TI has new spice models for LM3886 LM3875 et all - Page 3

KSTR · 18th May 2012, 06:25 PM

Find it attached, rename from .TXT to .ASC.

It's not a self contained complete model, only a stub, but you can just copy it into your circuits and use it for basic AC stability analysis. Don't excect that input parasitics are reflected, you must add them manually if desired.

linuxguru · 19th May 2012, 05:46 AM

Quote:

Originally Posted by KSTR

I contacted TI, now that we know that TINA also shows the problem.
Faulty SPICE model of LM3886? - Audio Amplifiers Forum - Audio Amplifiers - TI E2E Community

There's a positive response from TI on that forum, so maybe there's still hope.

KSTR · 19th May 2012, 01:43 PM

Yes, let's wait and hope the best.

Their comment on the OLG modelled unloaded is confusing me though, I see no big change with any load from 0.1R to 10R. The gain magnitude curve just barely moves, phase curve remains unchanged.

linuxguru · 19th May 2012, 03:03 PM

Quote:

Originally Posted by KSTR

Their comment on the OLG modelled unloaded is confusing me though, I see no big change with any load from 0.1R to 10R. The gain magnitude curve just barely moves, phase curve remains unchanged.

Yeah, loading shouldn't make much difference to OLG - it's a low-impedance, high-current output stage driving the load in any event. Loading definitely cannot alter the dominant pole, which is typically a miller cap between some internal VAS nodes.

jcx · 04:55 PM

actually open loop gain can change noticably with loading vs totally unloaded - factors of 2 aren't impossible

macromodeling has always been hit or miss - often the "new guy" may be assigned to familiarize him with the products - even if he's never bult a model before, hasn't even read the app notes on modeling

http://www.analog.com/static/importe...tes/AN-138.pdf

http://www.ti.com/lit/an/sbfa009/sbfa009.pdf

http://www.ti.com/lit/an/snoa265b/snoa265b.pdf

http://www.ti.com/lit/an/snoa247a/snoa247a.pdf

just checking the pin assignment on the op amp package shows any model with spice node "0" in it is suspect - particularly with any sub component of the model directly connecting form the op amp pins to Spice gnd

even the notes above often get this wrong with CM Z, leakage parts going to gnd

what is frustrating is the bad use of node "0"/gnd in modeling has been clear from before monolithic op amps and Spice - as early Philbrick/Analog Devices papers show input parasitics/compensation referenced to the power supply pins of the op amp modules

Ouroboros · 21st May 2012, 07:03 AM

I've just checked the LM3875 using 'TINA' in the inverting mode. On Friday I checked for non-inverting operation and the THD was as expected.
In the inverting mode the distortion is vanishingly small (unrealistic). I expected it to be lower as that is how the LM3875 measures in real life, but not by quite so much!

gootee · 10th September 2012, 02:11 PM

A TI newsletter mentioned that an LM3886 spice model was released in August. Has anyone looked at it yet?

I just saw the TI newsletter last night. I tried to look at the spice files but the versions I got linked to were encrypted or something. It was too late for me to pursue, last night, and now I'm at work and can't. Has anyone seen the newest LM3886 model?

Ouroboros · 10th September 2012, 02:58 PM

The files are now dated 6th July, so they have been changed from when I looked first. Taking their TINA ref schematic (inverting, unity gain) and changing R(in) to 1k to give a gain of 10, still shows unreasonably low distortion of 5E-6% at 1kHz and 7.4E-6% at 20kHz, into a 4R load at 30V pk-pk output, so something is still not right.

Discanto · 17th March 2013, 10:38 AM

any news?

Discanto · 14th April 2013, 07:46 AM

You're all dead?
Anyway...Im not expert of simulators but i try to use lm3886 TINA and PSpice model and I can't run any.
In TINA the models, both downloaded from TI or the original in TINA,the macro model does not seem to absorb any current, Pspice instead return me nodes erros, finally importing compilated pspice model in LTspice return error on the first compiled line.
The only model i can use is the LM4780 in TINA, very similar to the lm3886 but i fear is not exactly the same.
Any suggestions?

19th May 2012, 04:38 PM	#25
jcx diyAudio Member Join Date: Feb 2003 Location: ..	actually open loop gain can change noticably with loading vs totally unloaded - factors of 2 aren't impossible macromodeling has always been hit or miss - often the "new guy" may be assigned to familiarize him with the products - even if he's never bult a model before, hasn't even read the app notes on modeling http://www.analog.com/static/importe...tes/AN-138.pdf http://www.ti.com/lit/an/sbfa009/sbfa009.pdf http://www.ti.com/lit/an/snoa265b/snoa265b.pdf http://www.ti.com/lit/an/snoa247a/snoa247a.pdf just checking the pin assignment on the op amp package shows any model with spice node "0" in it is suspect - particularly with any sub component of the model directly connecting form the op amp pins to Spice gnd even the notes above often get this wrong with CM Z, leakage parts going to gnd what is frustrating is the bad use of node "0"/gnd in modeling has been clear from before monolithic op amps and Spice - as early Philbrick/Analog Devices papers show input parasitics/compensation referenced to the power supply pins of the op amp modules Last edited by jcx; 19th May 2012 at 04:55 PM.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Any spice models for the LM3886 ?	gaetan8888	Chip Amps	5	29th June 2009 09:03 PM
Simulation Models for OPA827 and TINA-TI	stef1777	Parts	8	13th March 2008 04:56 AM
Spice models LM3875 and TDA2050	Jarno	Solid State	5	3rd February 2003 11:06 PM
Spice models	JoeBob	Solid State	18	25th April 2002 02:34 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

19th May 2012, 01:43 PM	#23
KSTR diyAudio Member Join Date: Jul 2007 Location: Central Berlin, Germany	Yes, let's wait and hope the best. Their comment on the OLG modelled unloaded is confusing me though, I see no big change with any load from 0.1R to 10R. The gain magnitude curve just barely moves, phase curve remains unchanged.

21st May 2012, 07:03 AM	#26
Ouroboros diyAudio Member Join Date: Dec 2003 Location: Nottingham UK	I've just checked the LM3875 using 'TINA' in the inverting mode. On Friday I checked for non-inverting operation and the THD was as expected. In the inverting mode the distortion is vanishingly small (unrealistic). I expected it to be lower as that is how the LM3875 measures in real life, but not by quite so much!

10th September 2012, 02:11 PM	#27
gootee diyAudio Member Join Date: Nov 2006 Location: Indiana Blog Entries: 1	A TI newsletter mentioned that an LM3886 spice model was released in August. Has anyone looked at it yet? I just saw the TI newsletter last night. I tried to look at the spice files but the versions I got linked to were encrypted or something. It was too late for me to pursue, last night, and now I'm at work and can't. Has anyone seen the newest LM3886 model?

10th September 2012, 02:58 PM	#28
Ouroboros diyAudio Member Join Date: Dec 2003 Location: Nottingham UK	The files are now dated 6th July, so they have been changed from when I looked first. Taking their TINA ref schematic (inverting, unity gain) and changing R(in) to 1k to give a gain of 10, still shows unreasonably low distortion of 5E-6% at 1kHz and 7.4E-6% at 20kHz, into a 4R load at 30V pk-pk output, so something is still not right.

17th March 2013, 10:38 AM	#29
Discanto diyAudio Member Join Date: Feb 2013	any news?

14th April 2013, 07:46 AM	#30
Discanto diyAudio Member Join Date: Feb 2013	You're all dead? Anyway...Im not expert of simulators but i try to use lm3886 TINA and PSpice model and I can't run any. In TINA the models, both downloaded from TI or the original in TINA,the macro model does not seem to absorb any current, Pspice instead return me nodes erros, finally importing compilated pspice model in LTspice return error on the first compiled line. The only model i can use is the LM4780 in TINA, very similar to the lm3886 but i fear is not exactly the same. Any suggestions?