From: Andreas Robinson (sleeper75se_at_yahoo.se)
Date: 2002-02-27 14:29:47
Hi all, long time no see...
I've been working on my project a bit. No, it is not
finished yet. It turned out to be a bit more of a
hassle than I expected as I sort of forgot that my PCB
lacks a solder mask. :-) One must be very careful not
to create unwanted connections between a trace or via
and the ground plane when soldering. A factory made
PCB will not have this problem, saving a lot of time
and effort. It will look prettier too.
I'll be using this PCB-manufacturer eventually,
located in Bulgaria:
http://www.olimex.com/pcb/index.html
unless someone shows me a cheaper manufacturer with
similar capabilities.
They claim to have as good quality as anyone else but
one double sided, euro-size PCB starts at $26.
Thinner board material and so on costs extra. Shipping
is $5 within europe and $8 for the rest of the world.
Since you guys have been throwing screenshots and
photos around maybe I should do the same. I don't have
a camera unfortunately, but I've put the assembled
amplifier stage in the scanner...see the attached jpeg
file.
I have made some preliminary tests, described below.
it may sound straightforward, but I had to hunt down a
few bad soldering points, and was about to deliberatly
*cut* away a $10 chip (the LT1114) in pure frustration
at one time. Fortunately I came to my senses. :-)
Test setup:
The amplifier PCB (1mm thickness FR-4, 35u copper) was
just laid out on the workbench without any shielding
whatsoever. Power (+/- 5 volts) was supplied from a
benchtop linear power supply using 60 cm (2 ft)
unshielded cables.
An oscilloscope was connected to the output of the
amplifier board using 60 cm coaxial cable. About 10 cm
from the end the coaxial cable is connected to two
unshielded, but twisted, wires, in turn connected to
the PCB.
1. Testing the filter - without an instrumentation
amplifier mounted.
A function generator was connected to the filter stage
using the same kind of cable used for the
oscilloscope.
Results: Ok, the filter works - no surprise there.
Turning down the signal amplitude on the generator to
zero (or at least very low) I saw repeated bursts of
HF-noise, 6mV peak to peak with quiet areas in
between. It probably comes from the computer nearby.
2. Testing the whole amplifier - floating inputs
The inamp was mounted and the gain was set to 50 with
two 470 ohm resistors. The total gain was then around
1200.
The inputs were left to float, without any wires
connected to the PCB.
The 100pF HF-chokes were floating as the input stage
shield was not connected to system ground.
Results: Fairly random 50mV(p-p) LF-noise at the
output (7.5 bits resolution, wee! :-)
Connecting the shield plane to true ground in a single
point on a direct line between the 100pF caps and the
amplifier ground input halves this noise (8.3 bits).
By the way: When I touch the amplifier inputs once,
the noise is replaced by a high level (200mV) 50Hz hum
that remains even when I remove my hand. The 50Hz
signal stays until I power-cycle the amplifier.
Attaching 3cm (around 1 inch) leads to the amplifier
inputs causes the hum to appear immediately at
power-up.
3. Testing the whole amplifier - shorted inputs
The inamp-gain was increased to around 100, so the
total gain is now about 2500.
Simply shorting the inputs to ground leaves a very low
level (1mV) 50Hz signal and some HF-noise of lower
amplitude (maybe 250-500uV) on the output. This means
the amplifier has an internal dynamic range of at
least 12.4 bits in an unshielded environment and with
standard lab-power.
That's how it stands at the moment. Testing with
electodes will have to wait until I can run the whole
thing on batteries and see the results on my computer
instead of an oscilloscope.
The ADC+MCU is assembled, but has only been
"smoke-tested" so far. (That is, it didn't catch
fire.)
The board is not flat enough to fit in the scanner or
I would have attached a picture.
The IR-receiver is about 50% done, because I made a
big mistake, using the narrow SO footprint for the
MAX232 when I should have used the wide, so I'm saving
that particular mess for later.
Power is not done yet either. A group member named
Scott Elofson was kind enough (thanks again Scott!) to
send me two +/-12V (500V isolation) DC/DC converters
to play with. That sort of got in the way of the
battery approach. I have built a working board (with
patches. :-p) with one of the converters, but I'm not
sure I've managed to get the noise levels down
sufficiently. 2mV HF noise is very close to being
unacceptable. The RLC-filters are not nearly as
efficient as calculated for some reason...
Ok, questions, comments, suggestions? Are there any
other tests I should do? Jim-M? Joerg? Anybody?
Regards,
Andreas
_____________________________________________________
Hitta snörapporter...
från 500 olika skidorter i Europa
på http://se.snow.yahoo.com
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:39 BST