LTSpice MESFET models

Phil Hobbs

unread,

May 8, 2012, 2:19:39 PM5/8/12

to

Back on topic for a bit:

I'm getting back to work on the biochip preamp, and I'd really like a
more realistic SPICE model than the behavioural ones I cobbled together
previously.

There are various models in the Avago literature for their ATF3x143
pHEMTs, but nothing that I can persuade LTSpice to swallow. They
publish something called a Statz MESFETM1 model, as follows:

Statz Model MESFETM1 NFET=yes PFET=no Vto=–0.75
Beta=0.3 Lambda=0.07 Alpha=4 B=0.8 Tnom=27
Idstc= Vbi=0.7 Tau= Betatce= Delta1= Delta2=
Gscap=3 Cgs=0.997 pF Gdcap=3 Cgd=0.176 pF
Rgd=0.195 Tqm= Vmax= Fc= Rd=0.084 Rg=0.264
Rs=0.054 Ld=0.0014 nH Lg-0.0883 nH Ls=0.001 nH
Cds=0.0911 pF Crf=0.0936 Rc=137 Gsfwd=1 Gsrev=0
Gdfwd=1 Gdrev=0 Vjr=1 Is=1 nA Ir=1 nA Imax=0.1
Xti= N= Eg= Vbr= Vtotc= Rin= Taumd1=no Fnc=1E6
R=0.17 C=0.2 P=1 wVgfwd= wBvgs= wBvgd= wBvds=
wldsmax= wPmax= All Params=

Any ideas for how to get all this wonderful info into a form LTSpice can
digest?

Thanks

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

Joerg

unread,

May 8, 2012, 6:56:57 PM5/8/12

to

Phil Hobbs wrote:
> Back on topic for a bit:
>
> I'm getting back to work on the biochip preamp, and I'd really like a
> more realistic SPICE model than the behavioural ones I cobbled together
> previously.
>
> There are various models in the Avago literature for their ATF3x143
> pHEMTs, but nothing that I can persuade LTSpice to swallow. They
> publish something called a Statz MESFETM1 model, as follows:
>
> Statz Model MESFETM1 NFET=yes PFET=no Vto=–0.75
> Beta=0.3 Lambda=0.07 Alpha=4 B=0.8 Tnom=27
> Idstc= Vbi=0.7 Tau= Betatce= Delta1= Delta2=
> Gscap=3 Cgs=0.997 pF Gdcap=3 Cgd=0.176 pF
> Rgd=0.195 Tqm= Vmax= Fc= Rd=0.084 Rg=0.264
> Rs=0.054 Ld=0.0014 nH Lg-0.0883 nH Ls=0.001 nH
> Cds=0.0911 pF Crf=0.0936 Rc=137 Gsfwd=1 Gsrev=0
> Gdfwd=1 Gdrev=0 Vjr=1 Is=1 nA Ir=1 nA Imax=0.1
> Xti= N= Eg= Vbr= Vtotc= Rin= Taumd1=no Fnc=1E6
> R=0.17 C=0.2 P=1 wVgfwd= wBvgs= wBvgd= wBvds=
> wldsmax= wPmax= All Params=
>
> Any ideas for how to get all this wonderful info into a form LTSpice can
> digest?
>

Just plugged it in and LTSpice (rightfully ...) complains about all the
missing data. There is nothing behind Xti, N, Eg and so on.

--
Regards, Joerg

http://www.analogconsultants.com/

lang...@fonz.dk

unread,

May 8, 2012, 7:23:35 PM5/8/12

to

could try to delete the empty parameters, can't very well be important
if
they are empty

-Lasse

Phil Hobbs

unread,

May 9, 2012, 6:52:57 AM5/9/12

to

Thanks, gents. I did that already, as I should have mentioned. Even
with default parameters only, it doesn't seem to know about anything
resembling NMESFETM1 or STATZ_MESFETM1.

I saw what looked like a SPICE model for the ATF34143 (the big brother
of my device), which called out a device "GAASFET", but LTSpice doesn't
like that either. Using the NMF model, (which seems to be the vanilla
SPICE way of doing a MESFET) a simple simulation sort of runs, but just
about all the named parameters get rejected.

Cheers

Joerg

unread,

May 9, 2012, 11:08:43 AM5/9/12

to

One option would be to take a similar FET that LTSpice does eat and
compare the items line by line, parameter by parameter. If you get
totally stuck the best forum for this is the Yahoo Groups LTSpice group,
and there mainly Helmut Sennewald.

Gerhard Hoffmann

unread,

May 9, 2012, 5:00:11 PM5/9/12

to

Am 08.05.2012 20:19, schrieb Phil Hobbs:

> There are various models in the Avago literature for their ATF3x143
> pHEMTs, but nothing that I can persuade LTSpice to swallow. They publish
> something called a Statz MESFETM1 model, as follows:
>
> Statz Model MESFETM1 NFET=yes PFET=no Vto=–0.75
> Beta=0.3 Lambda=0.07 Alpha=4 B=0.8 Tnom=27
> Idstc= Vbi=0.7 Tau= Betatce= Delta1= Delta2=
> Gscap=3 Cgs=0.997 pF Gdcap=3 Cgd=0.176 pF
> Rgd=0.195 Tqm= Vmax= Fc= Rd=0.084 Rg=0.264
> Rs=0.054 Ld=0.0014 nH Lg-0.0883 nH Ls=0.001 nH
> Cds=0.0911 pF Crf=0.0936 Rc=137 Gsfwd=1 Gsrev=0
> Gdfwd=1 Gdrev=0 Vjr=1 Is=1 nA Ir=1 nA Imax=0.1
> Xti= N= Eg= Vbr= Vtotc= Rin= Taumd1=no Fnc=1E6
> R=0.17 C=0.2 P=1 wVgfwd= wBvgs= wBvgd= wBvds=
> wldsmax= wPmax= All Params=
>
> Any ideas for how to get all this wonderful info into a form LTSpice can
> digest?

in <http://cp.literature.agilent.com/litweb/pdf/ads15/spicent/snt0820.html>
is a cross reference. Spice2/3 and ADS Statz don't seem to have much in
common.

Agilent ADS, Genesys, Microwave Office and some others support the
Statz model. The handbooks apear to be on the net, so that should be
a place to search.

regards, Gerhard

Joerg

unread,

May 9, 2012, 5:15:53 PM5/9/12

to

Gerhard Hoffmann wrote:
> Am 08.05.2012 20:19, schrieb Phil Hobbs:
>
>> There are various models in the Avago literature for their ATF3x143
>> pHEMTs, but nothing that I can persuade LTSpice to swallow. They publish
>> something called a Statz MESFETM1 model, as follows:
>>
>> Statz Model MESFETM1 NFET=yes PFET=no Vto=–0.75
>> Beta=0.3 Lambda=0.07 Alpha=4 B=0.8 Tnom=27
>> Idstc= Vbi=0.7 Tau= Betatce= Delta1= Delta2=
>> Gscap=3 Cgs=0.997 pF Gdcap=3 Cgd=0.176 pF
>> Rgd=0.195 Tqm= Vmax= Fc= Rd=0.084 Rg=0.264
>> Rs=0.054 Ld=0.0014 nH Lg-0.0883 nH Ls=0.001 nH
>> Cds=0.0911 pF Crf=0.0936 Rc=137 Gsfwd=1 Gsrev=0
>> Gdfwd=1 Gdrev=0 Vjr=1 Is=1 nA Ir=1 nA Imax=0.1
>> Xti= N= Eg= Vbr= Vtotc= Rin= Taumd1=no Fnc=1E6
>> R=0.17 C=0.2 P=1 wVgfwd= wBvgs= wBvgd= wBvds=
>> wldsmax= wPmax= All Params=
>>
>> Any ideas for how to get all this wonderful info into a form LTSpice can
>> digest?
>
>
> in <http://cp.literature.agilent.com/litweb/pdf/ads15/spicent/snt0820.html>
> is a cross reference. Spice2/3 and ADS Statz don't seem to have much in
> common.
>

Normally it does, but I don't know about LTSpice:

http://www.ece.uci.edu/docs/hspice/hspice_2001_2-116.html

This App Note could help:

http://www.cel.com/pdf/appnotes/an1023.pdf

[...]

Jim Thompson

unread,

May 9, 2012, 7:12:24 PM5/9/12

to

On Tue, 08 May 2012 14:19:39 -0400, Phil Hobbs
<pcdhSpamM...@electrooptical.net> wrote:

>Back on topic for a bit:
>
>I'm getting back to work on the biochip preamp, and I'd really like a
>more realistic SPICE model than the behavioural ones I cobbled together
>previously.
>
>There are various models in the Avago literature for their ATF3x143
>pHEMTs, but nothing that I can persuade LTSpice to swallow. They
>publish something called a Statz MESFETM1 model, as follows:
>
>Statz Model MESFETM1 NFET=yes PFET=no Vto=–0.75
>Beta=0.3 Lambda=0.07 Alpha=4 B=0.8 Tnom=27
>Idstc= Vbi=0.7 Tau= Betatce= Delta1= Delta2=
>Gscap=3 Cgs=0.997 pF Gdcap=3 Cgd=0.176 pF
>Rgd=0.195 Tqm= Vmax= Fc= Rd=0.084 Rg=0.264
>Rs=0.054 Ld=0.0014 nH Lg-0.0883 nH Ls=0.001 nH
>Cds=0.0911 pF Crf=0.0936 Rc=137 Gsfwd=1 Gsrev=0
>Gdfwd=1 Gdrev=0 Vjr=1 Is=1 nA Ir=1 nA Imax=0.1
>Xti= N= Eg= Vbr= Vtotc= Rin= Taumd1=no Fnc=1E6
>R=0.17 C=0.2 P=1 wVgfwd= wBvgs= wBvgd= wBvds=
>wldsmax= wPmax= All Params=
>
>Any ideas for how to get all this wonderful info into a form LTSpice can
>digest?
>
>Thanks
>
>Phil Hobbs

Phil Hobbs

unread,

May 10, 2012, 12:37:32 PM5/10/12

to

Okay, after I gave up and resorted to Reading The Fabulous Manual, it
sort-of works. A bunch of the parameters are missing from the LTspice
version, but it claims to be based on the Statz model also, so it may be
better than the behavioural version.

Quite ridiculously, it leaves out the flicker noise parameters.

Thanks, all.

Jim Thompson

unread,

May 10, 2012, 12:46:33 PM5/10/12

to

On Tue, 08 May 2012 14:19:39 -0400, Phil Hobbs
<pcdhSpamM...@electrooptical.net> wrote:

>Back on topic for a bit:
>
>I'm getting back to work on the biochip preamp, and I'd really like a
>more realistic SPICE model than the behavioural ones I cobbled together
>previously.
>
>There are various models in the Avago literature for their ATF3x143
>pHEMTs, but nothing that I can persuade LTSpice to swallow. They
>publish something called a Statz MESFETM1 model, as follows:
>
>Statz Model MESFETM1 NFET=yes PFET=no Vto=–0.75
>Beta=0.3 Lambda=0.07 Alpha=4 B=0.8 Tnom=27
>Idstc= Vbi=0.7 Tau= Betatce= Delta1= Delta2=
>Gscap=3 Cgs=0.997 pF Gdcap=3 Cgd=0.176 pF
>Rgd=0.195 Tqm= Vmax= Fc= Rd=0.084 Rg=0.264
>Rs=0.054 Ld=0.0014 nH Lg-0.0883 nH Ls=0.001 nH
>Cds=0.0911 pF Crf=0.0936 Rc=137 Gsfwd=1 Gsrev=0
>Gdfwd=1 Gdrev=0 Vjr=1 Is=1 nA Ir=1 nA Imax=0.1
>Xti= N= Eg= Vbr= Vtotc= Rin= Taumd1=no Fnc=1E6
>R=0.17 C=0.2 P=1 wVgfwd= wBvgs= wBvgd= wBvds=
>wldsmax= wPmax= All Params=
>
>Any ideas for how to get all this wonderful info into a form LTSpice can
>digest?
>
>Thanks
>
>Phil Hobbs

Phil,

LTspice apparently addresses MESFET/GaAsFET, but no models (For a possible
model see below):

From LTspice Changelog:

02/07/04 Fixed a problem evaluating the MESFET saturation region.
02/17/02 Corrected some errors in MESFET current monitoring

From mesfet.asy:

Version 4
SymbolType CELL
LINE Normal 16 12 16 84
LINE Normal 48 80 48 96
LINE Normal 48 16 48 0
LINE Normal 16 16 48 16
LINE Normal 0 80 16 80
LINE Normal 16 80 48 80
WINDOW 0 56 32 Left 0
WINDOW 3 56 72 Left 0
SYMATTR Value NMF
SYMATTR Prefix Z
SYMATTR Description GaAs MESFET transistor Note: You must supply the .model
for this device
PIN 48 0 NONE 0
PINATTR PinName D
PINATTR SpiceOrder 1
PIN 0 80 NONE 0
PINATTR PinName G
PINATTR SpiceOrder 2
PIN 48 96 NONE 0
PINATTR PinName S
PINATTR SpiceOrder 3

Possibly useful link (maybe):

http://cp.literature.agilent.com/litweb/pdf/ads2005a/netlist/net0825.html

Here's an NMOS depletion-mode model (Fairchild, BSIM) which might SCALE to
match your needs:

***********************************************************************
* HIGH VOLTAGE DEPLETION NMOS TYPICAL MODEL CARD *
***********************************************************************
.MODEL NDP NMOS LEVEL = 6
+VTH0 = -1.7178513 K1 = 0.9981337 K2 = -0.0681299 +K3
= 37.6482304 K3B = -3.53199 W0 = -1.641794E-7 +NLX =
1E-9 DVT0 = 0.1073743 DVT1 = 3.534263E-3 +DVT2 =
0.0364547 DL = 3.2E-7 DW = 2.367287E-7 +U0 =
570.3449457 UA = 0 UB = 4E-18 +UC = 4.7804E-3
VSAT = 1.02031E7 A0 = 0.6869448 +KETA = -8.8323E-3 A1
= 1 A2 = 1 +RDS0 = 24.3914569 RDSW = 1.76728E3
VOFF = -0.06 +NFACTOR = 0.5281248 CDSC = 2.001847E-5 CDSCB = 0
+ETA0 = 1.063091E-3 ETAB = -0.036365 DSUB = 0.2415699 +PCLM
= 13 PDIBL1 = 0.01 PDIBL2 = 1E-5 +DROUT = 6.3088897
PSCBE1 = 1.040507E9 PSCBE2 = 1.003176E-7 +PVAG = 6.0679758 UTE
= -1.3 KT1 = -0.32 +KT1L = 0 KT2 = 0.015
UA1 = 4.422664E-9 +UB1 = -5.74133E-18 UC1 = -0.056 AT
= 3.3E4 +TOX = 3.13E-8 XJ = 4E-7 NPEAK = 1.7E17
+NSUB = 2E14 SUBTHMOD= 1 SATMOD = 2 +BULKMOD = 1
XPART = 1 XT = 1.55E-7 +VBM = -5 ETA =
1 VGHIGH = 0.15 +VGLOW = -0.15 CIT = 0
JS = 5.325E-5 +RSH = 0 CGDO = 6.383E-10 CGSO
= 6.383E-10 +CGBO = 1E-10 CJ = 2.818288E-4 PB = 0.88
+MJ = 0.5735524 CJSW = 5.415684E-10 PBSW = 0.65 +MJSW =
0.268275

HTH!

Phil Hobbs

unread,

May 10, 2012, 1:06:53 PM5/10/12

to

Thanks. Starting from where I am, scaling that model would be pretty
time-consuming. However, Reading the Fabulous Manual probably got me
there--I'm just putting the package parasitics in.

Cheers

Jim Thompson

unread,

May 10, 2012, 1:45:28 PM5/10/12

to

On Thu, 10 May 2012 13:06:53 -0400, Phil Hobbs

Maybe not, Just varying M and VTH0 works quite often for me.

>However, Reading the Fabulous Manual probably got me
>there--I'm just putting the package parasitics in.
>
>Cheers
>
>Phil Hobbs

Have fun ;-)