Posted by Christian Kaiser on June 10, 2007, 7:00 pm
Then use C or C++. No constraints that that type of programming has ;-)
No, I live in the 21st century. No need for that. But no problem to
include it if ever necessary.
We added 4 1-wire temp sensors (although the number is not a
limitation), several A/D converters, 8 bit digital I/O, a CAN bus, two
serial interfaces. That's enough for the moment, although experimenting
might add some additional requirements. ;)
Posted by Steve Cothran on June 11, 2007, 2:53 am
From a guy that spent a few K$ automating his building, MOV's and any
other transient protection you can put on a piece of wire will pay
itself back over and over. Zeners ad resistors are your frind,too.
God hates silicon attached to wires. At least he does in Liberty,TN.
Posted by Christian Kaiser on June 11, 2007, 12:53 pm
If you mean surge/lightning protection, you're correct, and this is used for
the controller ;-)
If you mean something else, would you please try to tell me what you mean?
English is not my mother tongue, so some meaning might get past me that
someone else would get immediately.
Posted by Steve Cothran on June 11, 2007, 9:09 pm
That was the meaning. Lightning surges burnt out most of my system
before I added surge suppression to all sensors,switches, and the
like.It only takes a short length of wire to have a larged induced
voltage on it from a nearby lightning strike.
I think to date I have lost 4 CCTV cameras, 3 LM35 temperature
sensors, several LED's, and 6 or so solid-state relays. All were
connected to some length of (shielded) wire or coax.
Posted by vern on June 9, 2007, 3:13 am
On 8-Jun-2007, firstname.lastname@example.org (David Williams) wrote:
Why didn't I use open loop tracking via astronomical data?
Many reasons. Open loop controllers on commercial heliostat fields require
very low backlash, a high degree of rigidity, perpendicular axes with
accurate sensors, accurate time and orientation, and a bit of setup on each
heliostat. Like I said, I'm going for simple and cheap hardware.
Software's another story but that scales up cheaply. If I'm one degree off
at 50' on a single axis, I lose more than half the light. At higher
distances, even more. I'd like an inexpensive heliostat that you put in a
field, point toward the target and walk away. When the woodchuck ate
through one of the 3 guide ropes holding the heliostat up, all I had to do
was pick it up, tie a knot in the rope and manually aim it at the target.
It tracked away and rewrote the log file as it did. I didn't have to align
anything or worry if it was up straight or if anything got bent. I do
however worry about that dam woodchuck!
As for the PC controlling it, I know it's a sledgehammer but hey, this is
prototype development. Plenty of time to burn it to a PIC or something. I
leave the computer on all the time and use it for other things - it just
runs the heliostat in the background. Using standard software (VPN, VNC)
and a webcam, I can play with the heliostat and keep an eye on it from work
- or anywhere else. Good to hear you mention the old VIC 20. I used to
have one controlling a robot arm. It could actually pour a bottle of beer.
That's how I got started in control theory.
I'm working on a new heliostat. This one will have 48 1'x1' mirrors - 4
times the size of the first one. I have wires running out to the first one
(no smarts on it) but I would like to make the second one completely
wireless with a small PV panel powering an onboard battery. 48 sq ft
focused to about 18"x18" will be too hot to use the shop window for a
target so I have to start thinking of making a mini "Power Tower".