LTF good book(s) on Cadence Pspice

I'm puzzled by that comment. CMOS and BJT _inherently_ model temperature effects. The _standard_ passives requires you to enter a value (or values, if second order), otherwise zero is assumed. I/C passive are usually quite thoroughly modeled with substrate effects, etc.

"Spice, A Guide to Circuit Simulation & Analysis Using PSpice", Paul W. Tuinenga, Prentice Hall, 1995, ISBN 0-13-158775-7

is the classic.

Though it's out-of-print, you can obtain good quality copies from used book dealers... I think most people who bought it didn't read it cover-to-cover ;-) My copy came from a library and is in mint condition.

...Jim Thompson

Is your copy still in unused mint condition?

Just about. I bought it for my "shelf", to make sure I have copies for all the "classics" before they become truly unavailable.

I had a copy once before, but like lots of books I've owned, I loaned it to dishonorable persons unknown.

Same with my Motorola PLL booklet. But a very kind soul lurking on this newsgroup sent me a replacement, and wouldn't accept payment.

I now have a no-loan policy for books and tools, not even to relatives... particularly relatives ;-)

...Jim Thompson

Watch out, you'll grow old and die, while hogging all your electronics treasures, and your wife or children will have to sell them on eBay, saying "From the estate of an old IC designer... Unable to evaluate or test, sold as is, do we have any bidders?"

(1) Where did you get the 0.7X and 2.0X numbers... from some text book?

(2) The device Spice model is likely Level=3... not the most robust model in the world.

(3) Discrete device suppliers are notorious for providing sloppy models

(4) In other words, the problem is with model quality, not PSpice; or any other Spice for that matter.

...Jim Thompson

The best way to get a tool ruined is to loan to a relative ;-)

...Jim Thompson

The model parameters and equations are in the reference guide.

There's also "Semiconductor Device Modeling with Spice", Antognetti and Massobrio, McGraw-Hill, 1988, ISBN 0-07-002107-4

...Jim Thompson

Jim Thompson said

Perhaps I am confused. Take, for example, an IRF150 from the PSPICE "evaluation" library. I set it's Vgs@10V and Id@25A and perform a temperature parameter sweep of

-40C, 25C, and 125C.

The RDSons (V(m1:d)/I(m1:d)) is 52m at 25C. However, the -40C and 125C RDSons are 36m and 82m respectively.

The part specifies an RDSon of 55m@Vgs=10V@Id=25A.

So apparently the model is temperature sensitive but is not providing the 0.7x and 2.0x RDSons I hope to see at

-40C and 125C.

Thanks!

Jim Thompson said

From the RDSon verses temp curve of the part PDF. This is typical for discrete MOSFETS.

Indeed. This is why I'd like to understand the SPICE model. A buddy of mine has some old (DOS) PSPICE manuals that describe the various model parameters. I intend to add a temp coefficient to the Rd and Rs terms. I suppose the KP term is also a candidate.... but I don't fully understand what the hell it is. ;-) It's called a Transconductance Term but has units of A/(V^2).

BTW, I found a hardcopy of your "classic" book on AbeBooks for ~$30. As soon as I confirm the quality... it's mine. ;-D

Jim Thompson said

Excellent.

Thanks for all your guidance.

Wasn't it some old Chinese philosopher/warrior who said something to the effect that the best success is to see my enemy's corpse floating down the river? >;->

Cheers! Rich

This is just _soooo_ ripe, but I'll take the high road here. >;->

Thanks, Rich

I read in sci.electronics.design that Rich Grise wrote (in ) about 'LTF good book(s) on Cadence Pspice', on Fri, 29 Apr 2005:

..... but not if you've just drunk the water!