Andy's bright ideas.

I am learning to use KiCad with increasing speed, and it is certainly good to have a FOSS cad system that is available for Windows and Linux.

So far so good. However, the mapping of schematic to layout has problems.

Taking the example of a diode, the schematic (drawing) symbol looks like an arrow with a bar across the sharp end, with a connection at each end. One of the possible physical shapes for this can be SOT23, a shape used by several different components, including transistors, mosfets, and double diodes. The standard numbering of the three connections on a SOT23 goes, unsurprisingly, 1 2 3. A normal single diode connects to numbers 1 and 3, while 2 is unused.

Mapping the connections from the symbol to the physical shape requires that the pin numbers match, so if a diode is to use a standard SOT23 shape, its connections would have to be numbered 1 and 3. However, if the diode is to have another physical shape, then pin numbers 1 and 2 might be needed, or maybe even 2 and 3. A single schematic symbol would not work for this approach – but one of the intended strong points of KiCad is that one can concentrate on the schematic until that level of theoretical design is dealt with, and only then consider the detailed physical shape of the parts used. If a different (but similar looking) schematic symbol had to be used for different shaped parts, that level of simplicity doesn’t exist.

The alternative is to have separate layout shapes called, for example, SOT23-diode, SOT23-mosfet, and so on. If there is ever a need to change the dimensions slightly, for example to accommodate a change in soldering process, then all these separate footprints will need to be edited. Meanwhile, back at the diode, with a system like that, the diode connections would be labelled A for anode and K for cathode (don’t ask), and the diode footprint library would have pads labelled A and K.

Most electronics CAD systems have a third type of element, often called a device. That is a file that maps which connections on a schematic symbol match with which connections on the layout footprint, so that for any type of component there is just one schematic symbol, and for any layout footprint shape, there is only one of them regardless of how many different types of part use it.

I am still trying to decide whether there is a logical way to use KiCad, or whether it is (very reluctantly) necessary to purchase an upgrade to my existing EDWIN software in the foreseeable future. Edwin is good, but CAD software is usually very costly, so I am using a version from 1999 which will only work on Windows 7 using the virtual machine, and which I have so far failed to make work under WINE.

Since the middle of last week, the main task I have been working on is to stabilise the voltage control loop of a switched mode battery charger. The design is slightly more complex than some, because the output voltage can be both above and below the input - so it uses a sepic converter. Not a big deal, but slightly more difficult to control. The control loop oscillates over a range of 0.2Hz to 500Hz, depending on the components fitted to attempt to stabilise the loop. The one thing it won't do is work in a stable and accurate manner.

( Read more... )

Since before the EU ROHS regulations came into force I have been wondering if the one that prohibits the use of lead in solder was a net benefit to the environment.

Yes, the solder in electronic equipment that is scrapped will be a bit less toxic, but:

1) It is all meant to be recycled anyway under regulations that came into force at around the same time (the WEEE regulations).

2) There will be more scrap because it will fail much sooner. 'Course, the councils will be able to charge us more to take the WEE to a recycling centre... and use more trucks and therefore fuel to do so.

I recently had an illustration of this when I tried to repair my niece's ipod. When I opened it I found that the problem appeared to be caused by a little of the solder paste not having melted, and having dried out and become a conductive dust in the works. The unused flux (I deduce) had caused corrosion when it got damp in the English weather, and so far repair has eluded me, despite repairing the obvious broken track and soldering the connections that were, previously, just touching.

With lead in the solder, there can (depending on the parts used) be tens of degrees between the temperature at which the solder melts, and the temperature that will rapidly destroy the components. Now, with lead-free solder, there is a wafer thin gap between not melting the solder and destroying everything rapidly by overheating. At best, some thermal damage is likely. At worst, a very few degrees of temperature difference between different areas being soldered can result in overheating in one place and not melting the solder elsewhere.

Today, while reading Electronics Weekly, I noticed this story about a lead-free reliability issue.

I think that the EU commissioners should be obliged to do all their air travel in aircraft with lead-free avionics for several years before anyone else does so. (There are currently temporary exemptions for avionics and military equipment.)