[Ktechlab-devel] Some feature ideas

[Link]

On 28/05/09 21:00, Alan Grimes wrote:
> Link wrote:
>> Here are some ideas for features that I had, that you may or may not be
>> interested in. I thought I'd let you know anyway :).
>
>> 1. Field lines indicating the magnetic flux around inductors. Similar to
>> the animated wires indicating current flow, animated field lines could
>> show the changing magnetic fields around inductors.
>
> Before that we need transformers; 'nuff said.
>
I guess we do.

>> 2. Tooltips on components. For example, a capacitor's tooltip could show
>> the charge stored on it, and a resistive component could show the power
>> it dissipates.
>
> The engine does a lot of weird linear algebra stuff, there doesn't seem
> to be a variable for flux or charge... Computing power draw seems rather
> straightforward though...
>
Isn't the charge on a capacitor just C*V? The voltage is already 
calculated and the capacitance is a user-defined value.

>> 4. A "show power pins" option on opamps similar to the "show body pin"
>> on MOSFETs. Using the power pins, one could select the minimum and
>> maximum values of an opamp's output.
>
> That completely changes the nature of the device. It goes from being an
> idealized linear device to being a nonlinear device which becomes a good
> deal more complex on the implementation side. =\ It is definitely on the
> todo list though.
>
I didn't know that about the current opamp. In any case, a realistic 
opamp model is a must-have for me!

>> 5. Hidden wires or net ports. When you have more than a handful of
>> components, it can get extremely tricky to see different wires. A simple
>> solution would be "net ports" like gEDA's gschem has. Place a net port
>> on one pin, and another with the same identifier on another, and the
>> pins act as if they're connected by a wire, though none is drawn.
>
> Now that's a good one...
>
:)

>> 6. Transformers/mutual inductors, Zeners, regulators and other
>> components commonly found in real-life circuits that KTL lacks. Big one,
>> I know, but it would make KTechlab a lot better!
>
> You can use the existing diode part as a zener (when it works.) =P
>
I see. You'd need a separate model if you implemented idea #8, though; 
reverse-biasing a normal diode to its breakdown voltage IRL would be a 
bad thing.

>> 7. Enhanced potentiometers: log/antilog pots, and the ability to set the
>> fader position as a percentage in the properties window.
>
> Yeah, the slider needs to be longer to permit more precise setting...
>
Don't forget the percentage entry, though. Also, it would help if 
potentiometer/variable resistor tooltips showed the resistance at the 
current position.

>> 8. Overload indicators: if a component is subject to more
>> voltage/current/power than it is supposed to handle, it could display a
>> warning signal, informing the user to change the parameters or supplies.
>> This also involves addition of polar capacitors, capacitor voltage
>> ratings, resistor wattage ratings, and so on, but those values need not
>> be used by the actual simulator (i.e. a .25W resistor dissipating 1W
>> would still work, but would display a warning sign, as would a polar
>> capacitor that's reverse-biased), so it shouldn't be /that/ difficult to
>> implement.
>
> Yeah, I had something similar in mind, the bigger problem is that we
> need to completely overhaul the entire component library cuz the current
> one is unmaintainable long-term.
>
> Right now I'm just fixing bugs and trying to get the code to a state
> where it's feasible to implement such radical changes.
>
I was never under the impression it would be done any time soon. 
Unfortunate as it may be, even for an end-user it is often noticeable 
that KTL doesn't work perfectly.

On a totally unrelated note, if you're looking for more devs, I _may_ be 
able to help out in a few months when my exams are finished. I'm not 
making any promises yet, but once I get enough time and peace of mind, 
KTechlab is a project I'd love to help with.