1. LTspice
  2. Теми

Working model of LM4040

 

As TI never seem to fix their model I'm wondering if anyone else have a working model of this?
 

Hi,:

(TI) has lots model of spice in pspice or TINA-TI, though not for LTspice.

But I believe they had spent lots resource on that to help their customers. It's still useful in some perspective.
I think one as a (TI)'s customer would  appreciate that help from them.

Best regards.
 

Not sure what you are trying to say here.
 

Hi.
Example simple models:
.subckt ZXRE125 1 2 Vref=1.22 rd=0.6 inom=0.1m imin=30u
d1 2 1 MyD
d2 1 2 dp
c 1 2 20p
.model MyD D(Ron={rd} Roff={Vref/imin*1.05} Vfwd={Vref-rd*inom-90u*(25-temp)})
.model dp D(Ron=1 Roff=1G Vfwd=0.6)
.ends
 
Bordodynov.


13.10.2016, 09:37, "rickard.steffensen@... [LTspice]" :

 

Not sure what you are trying to say here.
 

Thanks, i guess i could use another model for now. I would love to use the LM4040 as that's what i have planned for my circuit. However, if i can find a similar thing for 10v reference that would work too.
 

You could make your own model? Just looked at datasheet, it bascally behaves as a low Voltage Zener. You can do that with the simple diode model. HTH
 

I wish i could! Thats a bit over my head at the moment. :X

I found out that the LM4050 was almost the same as LM4040, but the spice model from TI dont work in LTspice either. Darn..
 

Hello,

There are many LM4040 unencrypted PSPICE models of the LM4040 on TI's web site.
These models work out of the box with LTspice too. I have uploaded examples.

LM4040-10-test.zip

Best regards,
Helmut

 

 
 

Hi again Helmut!
Once again you help me out, big thanks!

So, you are right. The LM4040 do indeed work on its own, like in your test-file. I was wrong, sorry for misleading anyone who reads this.

My problem must be something else then.
I am powering a small circuit from the LM4040 thru a reference buffer, and it simulates fast and properly when not using the LM4040 (using a 10V ref point from the embedded power instead), but once adding the LM4040 it simulates forever and my computer-fans go amok. Already destroyed one of my RAM-memories by letting it behave like this for a long time, so im afraid to let it run more than a few minutes now.

I guess i'll take my chances trusting the embedded 10V for preference is close enough. Its the op amp buffer that's more important anyway.

Thanks for all the help guys, appreciate it!
 

rickard.steffensen wrote:

   "The LM4040 do indeed work on its own"

The problem is that T.I. has SO many SPICE models for the part.  76 models!  Only the models listed in that list specifically as "unencrypted" are generic enough to be usable in other SPICE programs such as LTspice.

Even so, mistakes are easy to make.

   "but once adding the LM4040 it simulates forever and my computer-fans go amok."

Many SPICE models are poorly made and can cause a simulation to get very slow and run a simulator at "full tilt".  Occasionally it's not just the added model that's at fault, but rather a combination of models.  But more likely it is one badly-made model.

   "Already destroyed one of my RAM-memories by letting it behave like this for a long time"

Now that's a big concern, even though it isn't an LTspice problem.  Why was something in your computer so close to failing?  Are you over-clocking to such an extent that you can overheat a component, and your computer's thermal design is incapable of keeping up?  If so, I'd suggest not over-clocking.  Overclocking helps most on long, compute-intensive tasks, the kind that would raise the temperature to max, where the difference between 6 hours and 5 hours matters.  Overclocking doesn't make sense (to me) if it only benefits tasks that run for a second or two.  (Does it really matter if it finishes 0.2 seconds faster?)

If it was flash memory that failed, then you may not have your computer configured correctly.  Flash RAM shouldn't be used where there are very frequent write operations.  Even so, I think you are not likely to "wear it out" unless you ran the simulation for several months straight.  (But I didn't run the numbers.)

   "I guess i'll take my chances trusting the embedded 10V for preference is close enough."

Yes, it's key to know what parts of a circuit really need to be simulated.  There's no need to blindly simulate every component when some of them don't need to be simulated.

Andy
John Woodgate
 

I agree about the overheating. This should definitely not happen unless due to overclocking. Maybe your computer is full of dust?

With best wishes DESIGN IT IN! OOO – Own Opinions Only
<http://www.jmwa.demon.co.uk/> www.jmwa.demon.co.uk J M Woodgate and Associates Rayleigh England

Sylvae in aeternum manent.

From: LTspice@... [mailto:LTspice@...]
Sent: Friday, October 14, 2016 6:06 PM
To: [LTspice] group <LTspice@...>
Subject: Re: [LTspice] Re: Working model of LM4040


rickard.steffensen wrote:

"The LM4040 do indeed work on its own"

The problem is that T.I. has SO many SPICE models for the part. 76 models! Only the models listed in that list specifically as "unencrypted" are generic enough to be usable in other SPICE programs such as LTspice.

Even so, mistakes are easy to make.

"but once adding the LM4040 it simulates forever and my computer-fans go amok."

Many SPICE models are poorly made and can cause a simulation to get very slow and run a simulator at "full tilt". Occasionally it's not just the added model that's at fault, but rather a combination of models. But more likely it is one badly-made model.

"Already destroyed one of my RAM-memories by letting it behave like this for a long time"

Now that's a big concern, even though it isn't an LTspice problem. Why was something in your computer so close to failing? Are you over-clocking to such an extent that you can overheat a component, and your computer's thermal design is incapable of keeping up? If so, I'd suggest not over-clocking. Overclocking helps most on long, compute-intensive tasks, the kind that would raise the temperature to max, where the difference between 6 hours and 5 hours matters. Overclocking doesn't make sense (to me) if it only benefits tasks that run for a second or two. (Does it really matter if it finishes 0.2 seconds faster?)

If it was flash memory that failed, then you may not have your computer configured correctly. Flash RAM shouldn't be used where there are very frequent write operations. Even so, I think you are not likely to "wear it out" unless you ran the simulation for several months straight. (But I didn't run the numbers.)

"I guess i'll take my chances trusting the embedded 10V for preference is close enough."

Yes, it's key to know what parts of a circuit really need to be simulated. There's no need to blindly simulate every component when some of them don't need to be simulated.

Andy
 

Noted and agreed upon the models.
My computer isn't overclocked or anything. I did find it had turbo-mode however, so i disabled that and now it behaves like you say it would. Awesome.  :)
 

There have never been any dust in my computers thank you. :)
 

At 12:41 AM (-0700) 10/13/2016, td35p3j...5bx6ry wrote:


---------- Original Message ----------
As TI never seem to fix their model I'm wondering if anyone else have a working model of this?
---------- End of Original Message ----------

See LM4040.zip on the Device Models & Subcircuits page of my website.

It models the whole TI LM4040 Precision Micropower Shunt Voltage Reference Family. Simply set parameter VZ to the desired reference voltage.


     ...Jim Thompson
 

Hello Jim,

I looked at your file. It "only" contains the most often used 2.5V part. 

.SUBCKT LM4040 ANODE CATHODE PARAMS: VZ=2.5V
...

The original question was about the 10V version.

Best regards,
Helmut
 

At 11:32 PM (-0700) 10/26/2016, Helmut Sennewald wrote:

---------- Original Message ----------
Hello Jim,

I looked at your file. It "only" contains the most often used 2.5V part.

.SUBCKT LM4040 ANODE CATHODE PARAMS: VZ=2.5V

The original question was about the 10V version.

Best regards,
Helmut
---------- End of Original Message ----------

You missed, "Simply set parameter VZ to the desired reference voltage" ??

You "call" the 10V part with...

X1 Anode Cathode LM4040 PARAMS: VZ=10V


...Jim Thompson

Web Site: <http://www.analog-innovations.com/>
 

Hello Jim,

Sorry, I overlooked that your LM4040 model allows a parameter. 

Best regards,
Helmut
 

At 08:43 PM (-0700) 10/27/2016, Helmut Sennewald wrote:

---------- Original Message ----------
Hello Jim,

Sorry, I overlooked that your LM4040 model allows a parameter.

Best regards,
Helmut
---------- End of Original Message ----------

LM4040.zip, on the Device Models & Subcircuits page of my website, has been updated to include subcircuit call/instantiation by part number... separately setting VZ is no longer required 8-)


...Jim Thompson

Web Site: <http://www.analog-innovations.com/>
 

Hello Jim,

Thanks for this enhancement. 
Now its reference voltage will be directly visible by the model name in the schematic.

Best regards,
Helmut
 

Thank you so much for this model!
Also for the other great models on your site. :)

- Rickard

1 - 20 з 25