LTSpice models for LM2907, LM2917 attached

Thread Starter

SgtWookie

Joined Jul 17, 2007
22,210
After struggling with SPICE models on and off for awhile (over a year :rolleyes:) here's something that might be useful for some folks that are attempting a frequency to voltage conversion; a tachometer for instance.

You can use the attached with the free LTSpice, a product of Linear Technology. Google is your friend here. I'm not going to be much help porting this to other Spice versions, so don't bother asking - you're on your own.

It's a macromodel, not a component model. However, it covers the following IC's:
LM2907-8
LM2907-14
LM2917-8
LM2917-14

It's based off Russ Kincaid's macromodel for the LM2907-14, which disappeared from the Internet not long after I downloaded it.

Your mileage may vary; however the model/simulations may prove useful as a starting place.

The .sub file belongs in:
\Program Files\LTC\LTSpiceIV\lib\sub
The .asc files go in:
\Program Files\LTC\LTSpiceIV
The .asy files go in:
\Program Files\LTC\LTSpiceIV\lib\sym\misc

The basic macromodel that Russ published is:



I rotated, re-sized and converted his original work to .PNG format.
 

Attachments

Last edited:

carl0s

Joined Apr 25, 2010
39
I know it's kind of obvious, but anyway, where you said .sym I think you meant .asy, also with LTSpiceIV (downloaded today), it's C:\Program Files\LTC\LTSpiceIV.

Just in case anybody can't figure it out.

Thanks very much for the models!
 

Thread Starter

SgtWookie

Joined Jul 17, 2007
22,210
Thanks for the update; my LTSpice installation has been updated numerous times, but I've had it installed for a couple of years (since SwitcherCadIII) - it doesn't update the directory name when getting the latest upgrades.

Good catch on the .sym vs .asy filename extension - can you tell that I was tired? ;)
 

carl0s

Joined Apr 25, 2010
39
Thanks for the update; my LTSpice installation has been updated numerous times, but I've had it installed for a couple of years (since SwitcherCadIII) - it doesn't update the directory name when getting the latest upgrades.

Good catch on the .sym vs .asy filename extension - can you tell that I was tired? ;)
LOL ;) Thanks again :)
 

carl0s

Joined Apr 25, 2010
39
Is it reasonable for me to use LTSpice (and these models) for component (R1,C1,C2) selection when using a 2917 ?
I am trying to use the part (lm2917) in a circuit which is to work in the range of 5Hz to 45Hz, and I am seeing that the output (in LMSpice) is very "ripply". I am using your resistor/capacitor output filter thingy to help with this, and that is helping smooth out the output somewhat, but it's still a ripple.

I have ended up with values of:
C1 = 20nF
C2 = 0.1uF
R1 = 470k
R2 = 10K

but no matter what, it looks as though low frequencies like 5Hz are not very friendly with this circuit ? Any ideas?

Also I am seeing that C2 has a huge effect on the peaks of the output voltage (e.g. with the above, the voltage at "Out" seems to peak at 820mv, whereas if C2 is changed to 0.01uF (1/10th previous value), then the voltage at Out peaks at 7v.., even though C2 is barely if ever mentioned in the specification of the part).
 
Last edited:

Thread Starter

SgtWookie

Joined Jul 17, 2007
22,210
Carl0s,
My original thought with creating these models is that people would want to use them for tachometer accessories. A normal auto engine might idle at 550 to 800 RPM. Even a 4-cylinder engine would likely have an idle tachometer output at around 20Hz or so.

The filter example between Out and Out2 - try making two filters in series by adding another resistor and capacitor. You will see the ripple decrease quite a bit more.

The model is not completely accurate. There is a note in the model Russ had added to it; the real part saturates at roughly Vcc/2 but this model does not.
 

carl0s

Joined Apr 25, 2010
39
Carl0s,
My original thought with creating these models is that people would want to use them for tachometer accessories. A normal auto engine might idle at 550 to 800 RPM. Even a 4-cylinder engine would likely have an idle tachometer output at around 20Hz or so.

The filter example between Out and Out2 - try making two filters in series by adding another resistor and capacitor. You will see the ripple decrease quite a bit more.

The model is not completely accurate. There is a note in the model Russ had added to it; the real part saturates at roughly Vcc/2 but this model does not.
Thanks very much for your advice again. It seems I had mis-understood the tach output frequency and you are right, it's 2Hz for 60rpm, so as you said 20Hz = 600rpm, so my circuit actually needs to deal with 20Hz to 160Hz, and everything looks fine now!

cheers,
Carl
 
Top