Our target is to make this circuit (click on it to open a better version as PDF file):
Parts:
- Piece of perforated board
- Male DB 25 connector
- 3 pin screw terminal
- Some resistors, capacitors, diodes and related junk
- TLC 272 or LM 358
- TLC 0832 or ADC 0832
- 74HC243 buffer
- LM 2931 LDO
- A lot of imagination
More details will follow.
The circuit is simple so a short explanation will be enough.
- The power supply will be a battery pack of 4 Alkaline AA cells in a quad cell holder
- The 6 Volts is fed into the barrel connector J2
- (failsafe) capacitors are added to safeguard the batteries from rapid surges
-
Im the drawing I used a 78L05 voltage regulator. This was wrong, but it's a hell of a job to make new parts
in Eagle, so I took the liberty to take something with a similar footprint. The actual voltage regulator is
an LM 2931 low drop out regulator with extremely low quiescent current.
- The LM 2931 requires an electrolytic capacitor of 100 uF (or more) in the output in order to work.
- The 47 nF caps are filter capacitors which are mounted close to each IC
- The analog voltages enter in the right hand part: the X2 connector.
-
Each analogue signal is buffered by its own opamp. The resistors and diodes protect the input against too
high voltages.
- To extra safeguard the rest of the 5 Volt circuit, the fat TransZorb is mounted across the 5 Volt supply
- Each opamp buffers the input voltage and sends it through a filter circuit (time constant less than 1 us)
- The analogue signals enter CH0 and CH1 of the 0832 ADC.
- On the left side, the LPT port signals enter the circuit
- All relevant signals are buffered and (schmitt trigger) filtered
- D3/PE (pins 5 and 11) present a presence detect circuit and a visual indicator for debugging purposes
- The buffered signals control CS, CLK and DI to get the ADC to work for us
- The R5 pull up resistor is present to keep the ADC inactive when the circuit is disconnected from a PC
The power supply is everything around the LM 2931 regulator. From right to left we have:
-
Two filter capacitors. The small one (47 nF) will take out the higher frequencies, whereas the 100 uF
electrolytical capacitor (which in fact was a 470 uF one in the prototype) will smooth out any ripple and it
will take care of rapid power surges. This capacitor will give the battery time to start the chemical
reactions within a reasonable time. Batteries like this.
-
The LM 2931 is a very special voltage regulator:
- Low dropout voltage (typical 0.3 Volts)
- Very low quiescent current (less than 1 mA)
- Short circuit protected
- Thermally protected
- Can supply in excess of 100 mA
-
Shuts down when:
- the battery is reverse connected (upto 50 Volts!)
- the input voltage exceeds 26 Volts (upto 60 Volts!)
- the die gets too hot
- Will tolerate mirror image insertion in the PCB
which makes it ideal for our application. The shut down mechanism enables us to run without a reverse
voltage protection. The LM will do this for us free of charge.
-
The LM 2931 requires a 100 uF (or more) output capacitor. In the prototype I mounted a 470 uF capacitor. One
of the 47 nF ceramic caps was for the LM 2931 as well. In the prototype I used a 100 nF version.
-
The ICTE 5 Transzorb is for filtering out all voltages that exceed 6 Volts. It is mounted for the case a
very much too high voltage is applied to the analogue inputs and the resistor does not die fast enough (see
below)
- The 47 nF capacitors are to be mountd as close as possible to every integrated circuit.
The analog inputs may be exposed to a foolish action. Suppose someone puts the mains voltage (230 VAC in
Europe) between In1 and GND. This should be attacked by the built-in safety measures:
-
The 10k resistor combined with the 5 Volt zener diode will reduce the voltage on the input of the opamp to
5.1 Volts.
-
The resistor will take the full 230 Volts at 30 mA. The resulting 7 Watts of dissipated power will soon
evaporate it. In that case, the circuit is safe.
-
But if the zener diode dies prior to the resistor, the circuit remains attached to 230 Volts AC and the ADC
becomes very dangerous to the fool who connected it in the first place, but also to the controlling PC.
-
A second safety was built in: the ICTE5 device, also known as a 'TransZorb' or a 'Suppressor Diode'. These
diodes will take currents upto severak kilo-Ampere for a short time. In any way: they will relay the excess
voltage to GND, thereby overheating the input resistor, causing it to fail as a second chance
But it might as well, that the mains power will not only evaporate the resistor but the full PicoProbe circuit
and possibly (part of) the PC as well. Nothing beats safety and knowing what to measure.
This is a simple project with simple COTS material. It is meant to be built by knowledgable people. Higher
safety measures (like in the SOAP project) are not part of this topic. Both the hard- and the software for
this topic are released under the rules of the GNU GPL.
The 10k/5.1V resistor/zener combination will safely take the beating for input voltages exceeding 5 Volts, but
by not more than 25 Volts extra. So I guess the circuit is safe until input voltages of less than 30 Volts. At
30 Volts input, the resistors need to eat 25 Volts and with P = U^2 / R this gives us less than 100 mWatts.
Which is a piece of cake for the resistor (rated at 500 mWatt).
As with all projects on this site: you make this circuit out of your own will and desire. The circuit is safe
if the voltages are lower than 5 Volts DC. Do not expect the circuit to accept more than 6 Volts of DC or 3
Volts of AC. If it does: good for you, but it was not designed in.
Please read the rules of the
GNU GPL!
These rules apply to hardware, software and operation. They also apply to all software made by contributors.
Neither I nor any of my fellow developers will accept any liability for any loss or damage as a result of
abusing this project.
But then again: you still take part of traffic every day and those dangers exceed the ones mentioned here by
several magnitudes... Stay awake. Use your sense.
Of course you can go and fetch the datasheets of the most important parts from the manufacturers websites. But
you can also download them from a central place on the web:
http://groups.google.com/group/wisclub/files
Here you can find datasheets and information about:
- The LM 2931 LDO voltage regulator
- The TLC0832 analog to digital converter
- The 74xx243 buffer
- The TLC 272 or LM 358 opamp
For prototyping I use either of two methods:
- Wire wrapping (special wires on special sockets on perforated (Vero) board
- Point to point wiring on perforated (Vero) board with pieces of scrap wire
In the online photo album are several pictures, in high resolution (3000 x 2000 pixels) taken in Macro mode
with a first class digital camera (see Multimedia section below).
This project was constructed with the scrap wire technique, as you can see in the picture on the right. I use
as much of a resistor as possible... The resistor is placed on the board, one wire is bent towards the first
point to solder to and next the other wire is bent, guided and soldered in place. If the distance is too long,
or the wire would consume too much space, a separate wire is placed on the top side. This wire can be a bare
wire, or it can be a piece of wire wrap wire (the small red wire stubs in the picture).
This construction method requires a lot of concentration. It is also rather slow, since you need to study the
real estate of the Vero board every step of the way. But, although it is much slower than populating a custom
PCB, it is exgtremely fast, when compared to having one PCB prototype being made in a factory. And it is
virtually free: the 'PCB tracks' are made in spot with leftovers from resistors, rectifier bridges, plus a few
inches of solid hookup wire.
During conctruction and operation of the pSam 0832, I made several high resolution ppictures wit my Olympus
FE-200 digital camera (218-140 mm lens, 6 megapixel) and I put them online in a professional photo album:
http://www.mijnalbum.nl/Album=SF8DVVJX.
When visiting the album, you see thumbnails on the right. And bigger pictures in the center. If you click on
the latter, the original 3000 x 2000 pixel image is opened in a new window. These pictures reveal all details
you ever would want to see...
The first prototype was ready on April 12, 2007. There are no shorts, the power supply works. All sockets have
power and ground where they expect it.
A first test showed that the power consumption is very low:
| Condition |
Current (mA) |
No computer connected.
No ADC activity
LED off
|
2 |
No computer connected.
No ADC activity
LED on
|
8 |
Project halted
After some second and third thoughts I decided to call it a day and abandon ship. So I made a new design :
picoSam 2 saw the light of day. It is better in all respects.
Page created on 1 April 2007 and
Page equipped with FroogleBuster technology