LT1013, LT1014 models still buggy

RC_opampST_osc.asc

Hi elektrouwe ,

I'll check your schematic and operation of the model. It is possible there is a newish model bug.

The warning on groups.io is related to legacy libraries that took advantage of undocumented LTspice features. If you are not using those libraries, you shouldn't see new problems. LTspice 24.1.6 is very stable, and all ADI models have been tested against it, including the LT1013/14.

Unfortunately, an accounting bug in the 1013/1014 symbols was introduced at the same time as the introduction of LTspice 24.1, giving the impression that they were related. They were not. 

If you do see a problem pop up with a certain device, do check Help > Model Change Log to see if there was a recent update to that model. To wit, the original introduction of the symbol bug and the symbol fixes:

02/17/25 LT1013D Corrected symbol.

02/17/25 LT1013A Corrected symbol.

02/17/25 LT1014 Corrected symbol.

02/17/25 LT1014A Corrected symbol.

02/17/25 LT1014D Corrected symbol.

01/13/25 LT1013 Corrected output spike in transient startup condition.

I will examine your particular schematic. I'm guessing it might also be related to that 1/13 update.

mike

HI elektrouwe ,

There indeed seems to be a bug introduced in the 1/13 update. I'll enter a bug report.

I do have a workaround for you, while the bug is fixed. I've set up a symbol that you can use in LTspice 24.1, with the symbol and (old) library attached in the zip. The library contains this sub:

.subckt LT1013 1 2 3 4 5
A1 2 1 0 0 0 0 0 0 OTA g=0 in=15f ink=60
B1 0 N003 I=10u*dnlim(uplim(V(1),V(3)-1.1,.1), V(4)-0.4, .1)+1n*V(1)
B2 N003 0 I=10u*dnlim(uplim(V(2),V(3)-1.09,.1), V(4)-0.41, .1)+1n*V(2)
C6 3 1 .1p Rpar=20G noiseless
C7 1 4 .1p noiseless Rpar=20G
C8 2 4 .1p Rpar=20G noiseless
C9 3 2 .1p Rpar=20G noiseless
C10 N003 0 1f Rpar=100K noiseless
D1 N005 3 XU
D4 4 N005 XD
M1 3 N004 5 5 N temp=27
M2 4 N004 5 5 P temp=27
C1 N004 0 12f Rpar=1Meg noiseless
C3 3 5 .5p
C4 5 4 .5p
D5 N004 5 YU
D6 5 N004 YD
G1 0 N004 N005 0 1µ
C2 2 1 1f Rpar=435Meg noiseless
A2 0 N006 0 0 0 0 N005 0 OTA g=160u iout=9.59u Cout=23.4p Vhigh=1e308 Vlow=-1e308
G2 0 N006 N007 0 1m
L2 N006 0 .457m Cpar=5.08p Rser=1.29k Rpar=4.448275862068965k noiseless
A3 0 N003 0 0 0 0 N007 0 OTA g=1m linear en=20.2n enk=2 Vhigh=1e308 Vlow=-1e308
C11 N007 0 532p Rpar=1k noiseless
D2 2 3 DBIAS
D3 1 3 DBIAS
.model XU D(Ron=1K Roff=100G Vfwd=-.72 epsilon=1.0 noiseless)
.model XD D(Ron=1K Roff=100G Vfwd=-.114 epsilon=1.0 noiseless)
.model YU D(Ron=500 Roff=1T Vfwd=.77 epsilon=.1 noiseless)
.model YD D(Ron=1k Roff=1T Vfwd=.74 epsilon=.1 noiseless)
.model DBIAS D(Ron=1k Roff=1T Vrev=0.6 revilimit=12n noiseless)
.model N VDMOS(Vto=-100m Kp=70m)
.model P VDMOS(Vto=100m Kp=70m pchan)
.ends LT1013
*

Note that all versions of the LT1013x and LT1014x use the same sub (above), so you can use the same symbol for all of them. This is an old library, so the convergence fixes will not apply.

Place all files in the same directory, and it should just work.

mike

LT1013-LT1014_Workaround.zip

thanks for the workaround,  Mike !

Hi elektrouwe ,

No problem. Hopefully you won't need it soon. 

mike

LT1413 model seems to be very closed to LT1013.