Functional
Description

faceplate.cdr:
CorelDRAW 9.0 file used to make faceplates.
This
module is used to condition an EMG signal with a high pass-filter followed
by a full-wave rectifier. The high-pass filter is used to reject low-frequency
motion artefacts. The rectified output may be selectively inverted. In our
lab, the Tibialis Anterior EMG is positively rectified, whereas the Gastrocnemius
EMG is negatively rectified. (The 15-volt output once available to power an
EMG pre-amplifier has been discontinued in this module.
The
output signal from the EMG amplifier is applied to the input BNC connector
on the front panel.
If the signal
is to be high-pass filtered only, the output should be connected to the NORMAL
output BNC and the filter switch placed to ON. If the input is to be rectified
only, the ouput has to be connected to the RECTIFIER output BNC and the filter
switch must be set to OFF. The rectifier switch can be set to positive or
negative rectification. If the output
is to be rectified and filtered, the output has to connected to the RECTIFIED
output BNC with the filter switch ON. You may also have an output of 15V at
the 15V output BNC which can be used to power the EMG amplifier.
The
following table summarized the possible outputs:
Output
|
Filter
switch
|
Output
BNC
|
Rectifier
switch
|
HP
filtered only
|
ON
|
Normal
|
Not used |
Rectified
(+/-) only
|
OFF
|
Rectified
|
+
or -
|
HP
filtered and rectified
|
ON
|
Rectified
|
+
or -
|
15
volts
|
Doesn’t
matter
|
15
V
|
Not
used
|
Description
of rectifier circuit
R1,
R3, D1, D2 and IC2 form an inverting, half-wave precision rectifier with positive-polarity
selection. The circuit operates by providing two gains. For a positive input,
D1 is reversed biased and D2 is forward biased; resulting in a gain of -R3/R1
= -1. For an input with negative polarity, D1 is forward biased and D2 reversed
biased; resulting in a gain of 0/R1 = 0.
R2,
R4, R5 and IC3 form an summing amplifier with differing gains along each input
branch. A signal travelling along branch R2 is amplified with a gain of -R5/R2
= -1, whereas a signal travelling along branch R4 is amplified with a gain
of -R5/R4 = -2.
IC2
and IC3 together form a non-inverting, full-wave precision rectifier. The
circuit operates by adding the raw signal provided by SW1 to twice the output
of IC2, and inverting it. As such, positive portions of a signal, say, x are
processed as follows: -(x + 2(-x)) = x. Negative portions of the signal, on
the other hand, are processed as follows: -(x + 2(0*x)) = |x|. Hence, the
output of IC3 is always positive.
R7,
R8 and IC4 form an inverting amplifier with negative unity gain. It is used
to provide an inverted rectified output via SW2. C1, R6, C2 and R9 compensate
their respective OpAmps so that large capacitive loads may be driven (e.g,
long cables). This form of compensation does not reduce phase margin but may
only be used on OpAmps that are stable at unity gain.
Printed
Circuit Board

In
order to make new printed cricuit boards, use this template: pcb.pdf.
Alternatively, this diagram can be found in the filt_rect.cdr file above (see
Schematic).
Frequency
and noise analysis
Filter:
The
high pass filter is a fourth order ( 80 dB roll-off/decade) Butterworth filter
from Frequency Devices (711H4B) with a corner frequency at 5Hz. The gain in
the pass band region is 0 dB. The phase is 0 degrees in the stop-band region,
-90 degrees at the corner frequency and –180 degrees in the pass-band region.

The
output noise of the filter was very low.

Rectifier:
The
output noise of the rectifier was very low.

Last modified: November 30, 2001 Laura Galiana