PDA

View Full Version : Spice Anyone?



Areal Person
09-08-2007, 08:47 AM
Hi,
ˇ
I got a program called B2 Spice.
http://www.beigebag.com/v5_features (http://www.beigebag.com/v5_features)
ˇ
I’m a total newbe to electronics, so please excuse my ignorance
about this stuff.
ˇ
I’m thinking it would be wonderfull to be able to use the Propeller
inside a program like this.
ˇ
Is this reasonable ? Or is the Propeller to advanced with features for it
to be easly modeled as a component in a program like Spice.
(might just be hard for me, because I don’t have the knowledge)
ˇ
Is there a spice program that uses the Propeller already?
Maybe I should change software ?
ˇ
I really don’t know why I picked this version of Spice, It looked advanced
and I thought I could try to learn it.
ˇ
Thanks,
-Areal http://forums.parallax.com/images/smilies/smile.gif

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I have a tree growing out of my head, but

what do you expect ? I'm a programmer.

Leon
09-08-2007, 11:38 AM
Spice is an analogue simulator.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

Areal Person
09-08-2007, 12:19 PM
I'm using B2 Spice.

B2 Spice v5 supports mixed-mode simulation, based partly on the Berkley Spice simulator and partly on the Georgia Tech XSpice simulator.

The mixed mode simulator allows you to simulate both analog and digital parts and signals in the same circuit. The B2Spice part library includes hundreds of Xspice models for digital simulation.

Will it not work ? Why ?

Look at this link & flash movie
http://www.beigebag.com/mixedMode.htm

QUESTION:

HasˇANYONE put the Propeller into any type Spice/electronic simulation program at all ?

What are your thoughts about this feature ?

-Areal




▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I have a tree growing out of my head, but

what do you expect ? I'm a programmer.

Post Edited (Areal Person) : 9/8/2007 4:32:42 AM GMT

Skogsgurra
09-08-2007, 12:50 PM
Hello Real,

It is absolutely doable. But, given the ease of use, low cost, demo board etcetera, I do not see myself using a simulator to find out the real behaviour of a circuit. I use simulation a lot where high voltages and power is involved. The simulator does not explode in your face. And it is a lot cheaper to build a MW power stage in a simulator than it is IRL - not to mention the problem getting enough power from the grid in a domestic area.

The fun with the Propeller is that it is almost a simulator per se. And, with ViewPort, even more so.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

Areal Person
09-08-2007, 01:22 PM
Interesting,

I'm new to electronics, so I don't know what I'm doing.
a sim seems to give me allot of data, quicker. And it's easy for me to make dumb mistakes
that I regret at this time. Also, it's quicker for me to wire things up on the spice program.

However, currently, it seems there's no Propeller in the model, that needs to change.

I'm hoping I can test some components with the Propeller before I buy them.
Heck, most of them, I've never hooked up, much less with a Propeller driving them.

I need to hook up fram memory, it's surface mount, it's expensive, and I don't know
how to do it. But... I MUST DO IT, and soon. (I like a challange).

It's interesting for me to try to learn what is going to happen before I really do it.
(Or, thats what I would like to do). I'm sure you understand.

Everyone starts at the back of the line.

But I'm concerned that it might be a monster of a project to get the Propeller model
into a spice program.

Whats the thought on the scale of difficuilty it would be to add the Propeller to
a good spice program ?

Thanks for the advise.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I have a tree growing out of my head, but

what do you expect ? I'm a programmer.

Post Edited (Areal Person) : 9/8/2007 5:28:17 AM GMT

Harrison.
09-08-2007, 01:52 PM
You would have to write an entire emulator for the Propeller. The level of difficulty also relies on how hard it is to add a digital model to the simulator. Some simulators use VHDL, others use C++ plugins. Even if you did manage to model the entire Propeller, it would likely run really slow.

No matter what, you would have to fully understand the Propeller architecture. I would recommend you go with real hardware just because it would be so much easier.

deSilva
09-08-2007, 03:10 PM
A Simulator is a device by which a very well described dynamic model is transfered to a different underlying physical reality, to profit from this different environment in terms of "learning" things.

A simulator can never "do" what the "original" does (without additional fancy interfaces). This might not matter when the "real thing" would have been also just build for "learning" something.

(BTW: When the underlaying reality of a simulator is similiar to the "original", the term "emulation" is sometimes used rather than "simulation". )

(Another BTW: It might in fact be cost effective to invest into those "additional interfaces" when the simulation is - say - "for free", in contrast to massiv costs of a built "original". This is what microcontrollers (as "electronics simulated by software") are all about!!)

The first advanced simulators funnily worked vice-versa to a modern "digital" simulator: They were called "Analogue Computers", consisting of cleverly aligned cogs (!) or (later) op-amps.

They were used to solve mathematical differential equations, by utilizing the fact that an EE's tool box contains amplifyiers, integrators, and differentiators. (Change of domain: Mathematics -> Electronics)

Digital Simulators use "Software" (it is of little interest that this software runs on a computer) as destination domain for any kind of "model". For this the "model" has to be defined exactly in terms of "formal requirements".
Using already well defined "building blocks" (e.g.idealistic electronic components) help a lot in reducing implicite complexity by using "abstraction levels".

A graphical IDE also helps for visualizing the "design" (and the results as well). Graphical manipulation for creating the model has pushed the popularity of simulators, simplifying the access to it for
- experts in the original "domain", with little knowledge of computers
- persons with neither lnowledge of the domain nor computers

It is extremely arguable to what respect a simulator can further the education within a specific "domain". It is "trial and error". Didactics knows that this is an important factor in early childhood experience and can even motivate adults. This is the cause for many "What If"-scenarios....

However electronics is no longer considered as something on a "What if" basis. The "underlying reality" is well understood (Yes, I know there are things like EMI...) and learning electronics (as well as programming, ecomomy, or politics) is learning the guiding rules and the working of established sub-assemblies.

Don't misunderstand me: Using a simulator is an indispensible tool nowadays for a "low cost validation" of your work. As some aspects of timing within large FPGAs or ASICs are too complex to understand, their working has to be tested by simulation before production!

Coming back to the beginning:
Simulation is at its strongest when the building blocks follow simple specifications. A programmed microcontroller can be such a building block, in terms of its program considered as a formal specification of its behaviour. This requires the simulator posesses specific knowlege of the internal working of the microcontroler to be efficient
This is how GEAR started. There are many very mature "simulators" around other microcontrollers with large libraries of peripheral components.

Using a model specification on transistor basis (as in FPGA design) seems phantastic overkill to me.
Modelling the functional units of the Propeller seems possible within some margins of accuracy (exact timing?)

Does any one know of a usable SPICE model of ANY micro processor?

Post Edited (deSilva) : 9/8/2007 8:17:48 AM GMT