Difference between code modeling and editing a componants model

[A-ron]

Hi, I recently submitted a question regarding the use the an avalanche transistor FMMT415. I wanted to simulate it but was having troubles with it. Someone replied and explained that multisim will not model avalanche breakdown, however, he then created a componant using a SPICE macro model for me. I now need to simulate another avalanche transistor the FMMT413, I have tried to use "editing a componats model" but gained unexpected results. I want to know if I should be using "code modeling" and if thats what using a "SPICE macro model" means. Finaly I was hoping someone might be able to do this for me, as I'm relativly new to multisim. As well as this i'm using vista and the explanation given in the multisim users guide seems to be for XP. The SPICE code can be found at

http://www.diodes.com/products/catalog/detail.php?item-id=823&popup=datasheet

Any help on this would be much appreciated, regards Aaron.

[GarretF]

I can't create the model for you (it isn't my area of expertise), however, I can answer some of your other questions. The short answer is that you probably don't want to be doing code modelling, instead, you need someone to modify the SPICE macro model.

In case your interested, I'll also give the long answer.

There are several levels of 'models' in SPICE, and unfortunately, we often use the word 'model' to mean several different things.

The core of SPICE is an equation solver. The equations SPICE is solving represent things, such as a resistor or a MOSFET.  However, these are only approximate representations of these devices. For example, you could describe a resistor as only having 'resistance' excluding other effects such a temperature. Similarly, there are a variety of MOSFETS, so one single set of equations does not describe all MOSFETS well. Consequently, there are several devices that represent MOSFETs to SPICE. This is the most basic level. When you create a code model, you are working at this level.

One step up from devices is SPICE macro models. These combine a number of primitive devices into a single unit. For example, an Opamp is constructed from resistors, capacitors, transistors, etc. Rather than create a primitive and generic Opamp, the model for an Opamp is a combination of these primitive units. The actual units may not be the same as the construction of the Opamp - for speed, it is often better to focus on specific characteristics of interest, rather than representing the topology of the Opamp exactly.

So, for your main question. When you edit a component's model, you are doing one of two things. (1) You are changing the values of parameters in the device - typically when you modify RLC, diodes, transistors. This is because usually these are often represented by primitive devices with different parameters. (2) Alternatively, you are changing the SPICE macro model - typically for things such as Opamps or more complex ICs, but it can also apply to describing components such as diodes where the device does not adequately represent the important characteristics.

For the avalance breakdown case, the primitive diode device does not include effects of avalance breakdown. To include this effect, there are two possible approaches (1) create a new SPICE diode device that includes this effect, or (2) create a SPICE macro model using some combination of devices to include this effect. Option 1 is very difficult. Researchers often spend years working at this level. Option 2, while not necessarily easy, is usually much easier (orders of magnitude easier).

The model from diodes.com for the FMMT413 is a SPICE macro model. You can see this if you look at the model, you see ".subckt" and there are a number of different primitives in the SPICE macro model. I think Max created a Multisim component for you that uses this SPICE macro model. The model seems to be from the manufacturer, so it is probably the best model that currently exists for this particular component. What you need is someone to modify the SPICE macro model to include this effect.

[A-ron]

Hi Garret, thanks a lot for taking the time to answer my questions, it was very useful. If anyone else out there thinks that they could modify the SPICE macro of the FMMT413, please let me know. This is for my undergrad thesis and is taking me much longer than it should, any help would be much appreciated. Regards Aaron.