Subject:
|
Re: MOC: Working solar powered rover and geiger counter
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Wed, 8 Feb 2006 17:42:31 GMT
|
Viewed:
|
1973 times
|
| |
|
|
In lugnet.robotics, Philippe Hurbain wrote:
> In lugnet.robotics, Justin Fisher wrote:
> > Hiyall. I just got my Lugnet posting access working again, so I thought I'd post
> > a here's-my-Lego-webpage link. Ignore as you see fit :-)
> >
> > One of my areas of interest is building solar-powered robots (and the necessary
> > custom mindstorms parts to make them work). I recently put up a page with some
> > pics of the solar rover I took to NWbrickCon, and some RCX geiger counter
> > sensors:
> >
> > http://home.earthlink.net/~stuff.tm/solarlego/nwbrickcon05/index.html
> >
> > Some various related stuff at the main page:
> > http://home.earthlink.net/~stuff.tm/solarlego/index.html
> >
> > Also at NWBrickCon, the tip of one of my models just barely enters the frame of
> > a photo of something else in Brick Journal issue 3. That's close to getting a
> > model in print, but no cigar. Maybe next year :-)
> >
> > - Justin
>
> Whooaaa!!! very neat stuff here, very good documentation of your projects. I
> especially like the solar powered RCX with auto-shutdown/resume...
>
> Any chance to get the schematics of you Geiger counter interface?
>
> Congratulations,
>
> Philo
Thank you :)
I don't have a schematic (more like a few pages of illegible notes :) and the
gieger counter sensor isn't something that's easy to make a useful schematic for
anyway because of unusual parts that vary in key properties, and the values of
the rest of the components is a real balancing act in some areas and depends on
those properties. But I can describe how to do it, and will probably add a page
at some point with this kind of info and maybe draw up an example schematic.
The heart is a low input voltage (5V) CCFL inverter and a low voltage geiger
tube. Geiger tubes normally operate at 900V +/- 20V, but some are designed for
500V, and most of these will start to operate in the 350V range. The low voltage
geiger tube is essential because the big killer is the 15mA power limit from the
RCX sensor port (which I imagine will not be a problem with NXT :), which for
the inverter I used isn't enough to create a regulated supply AND reach 400V.
One or the other. Since the 15mA is regulated, and the high voltage is well
below the threshold that can damage the tube (which is only slightly higher than
the voltage it is designed for) the high voltage can be unregulated and rely on
the RCX to constrict the power supply sufficiently to keep the voltage down.
However, if the RCX 15mA at 5V is enough to keep the voltage at 400V, then
accidentally plugging the sensor into a motor port is going to deliver 8V 100mA
and the HV will shoot right up and potentially damage the tube. Surprsingly, a
pot to split the load between inverter and the signal amplifier allowed an
adjustment that barely met the 350V at 15mA, but would reduce the rise to about
500V when plugged into a battery box. Additionally, I put a 500V diode chain on
the HV output as an additional precaution. The inverter tube output was
rectified with 1000V diodes and buffered with some caps for a 350V DC output,
this was connected to a 1 meg resister (or whatever the geiger tube specs
require) which is connected to the annode of the tube, and also to a HV cap, the
other side of which is where you tap the signal into the amplifier (which is
powered off part of the 5V feed via that pot).
The amplifier was just a chain of transistors to make a weak pulse into a signal
large enough to light an LED and momentarily change the RCX reading from 0 to
100%. The length of the event pulse from the geiger tube exceeds the 3ms gap
between RCX sensor readings, so you don't need to lengthen the pulse. Or at
least not with the tube and capacitor that I used. (The interface with the RCX
is the standard bridge rectifier analogue interface one, though schottky diodes
are needed because the power limit issues are so critical, the voltage drop of
normal diodes takes the inverter out of the geiger tube's operational range).
How to do it: Hook the inverter to the HV rectifier and buffer, and to the
geiger tube and the HV cap for the signal tap. Then use an oscilloscope to find
the signal from the tube (you'll want something radioactive by the tube so you
know the tube is active). Once you can see the signal coming from the tube, you
can juggle the rest until everything works.
Also, you'll want to make a 990 to 1000 megaOhm resistor to put in series with
the multimeter probes when measuring the high voltage, since otherwise the 10meg
resistance of the meter is going to make a significant dent in your high voltage
readings. Using the resistor means multiplying the meter reading by 100. Eg a
3.56V reading means 356V
I would prefer to be able to use a scintillator instead of a geiger tube, but
geiger tubes only draw power when an event occurs, so their power draw is lower.
I don't know if a useful scintillator is even possible on 5V 15mA, but I imagine
NXT will offer more power to the sensors, so who knows :-)
OTOH, geiger tubes are smaller, and need much less amplification, both of which
are very useful for Lego sensors :)
- Justin
|
|
Message has 1 Reply:
Message is in Reply To:
13 Messages in This Thread:
   
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
