[plug] Call me crazy...

[Denis Brown]

At 10:37 29/10/2003 +0800, you wrote:
>In message <BBC5435B.2451%ahewitt at globaldial.com>
>on Wed, Oct 29, 2003 at 10:13:47AM +0800, Adam Hewitt wrote:
> > But does anyone with any significant electronics knowledge know if it would
> > be possible to somehow connect a kettle to a serial port on a PC

Okay, I'll bite...   I should probably prefix this with "Don't try this at 
home!" and other suitably legal-sounding warnings :-)  If serious about 
this, we ARE talking mains power here and that stuff Bites Real Good (tm).

First thought I had was to wire up an optically-isolated gated triac in 
series with the mains active lead to the kettle.   This will provide the 
mains switching capability.  Opto-isolated gated triacs with sufficiently 
sensitive control and rated for a kettle's 10-odd amps should be available 
from the likes of Farnell or RS Components.   You should be able to get 
enough current from the serial port output control line (DSR, RTS - I 
forget which) to trigger the triac directly.   Failing that you would need 
to have some form of contactor (or suitably mains-rated relay) controlled 
by a transistor switch and fed from a suitable power source.   Any 
contactor worth its name is unlikely to be powerable from the meagre output 
of a serial port control line.

To determine when the kettle has boiled, I would just do a crude current 
measurement trick - current flow = not yet boiled, no current flow = has 
boiled.   Assuming that you have an auto-cutout style kettle, I would wind 
a turn or two of the neutral lead to the kettle through a ferrite core, 
forming the primary of a current transformer.[1]   Wind some tens of turns 
of insulated copper wire around the same core, forming the secondary 
winding of the transformer and feed that to an op-amp configured as a 
Schmidt trigger.   When current flows to the kettle, the Schmidt trigger 
will have an output that could be read by one of the serial port's sensing 
control lines, CD for example.   There should be an integrator in the 
Schmidt output so that the 50 (100?) Hz ripple will not be problem for 
sensing.   When boiled and the kettle cutout has tripped, no more current 
flow = no more Schmidt output = "she's boiled, mate!"

The control and sensing electronics would need to be enclosed in a suitable 
IP65- or IP66-rated enclosure to protect it from moisture - after all we 
ARE talking about a kettle here :-)   Because this is involving mains 
power, a qualified electrician would be needed to at least give a final nod 
of approval, if not to construct it, otherwise the long arm of the law 
might get displeased if lives were damaged by such a gadget.   Not to 
mention the paperwork :-)

Hope this helps,
Denis

[1] Putting the current transformer in the neutral lead minimises the risk 
from damaged insulation on the active lead creating a hazardous 
situation.   Otherwise, technically, there is no reason why the active lead 
could not be used instead.



_______________________________________________
plug mailing list
plug at plug.linux.org.au
http://mail.plug.linux.org.au/cgi-bin/mailman/listinfo/plug