In a message dated 09/01/2006, DSaum@............ writes:

System  Bandpass: 0.01 to 4 Hz.
High-Pass filter: ~100 sec or 0.01 Hz using  volume/capillary tube.
Low pass filter: 4 Hz anti-alias, 8 pole Bessel  filter (MAX4701)
    This gives a drop of 40 dB at 4 Hz, an average  signal delay of 0.2 sec 
and 0.35 sec peak. 

Differential pressure sensor: silicon, bridge type, 1"wc  (DUXL01D)
Pres Resolution: ~0.004 Pa - possibly lower depending  offsets
Range: +/- 120 Pa (32768 times pressure  resolution)
    For a 5 V supply, this sensor has an output of 6 mV  per inch water and a 
span of 6 mV. You might do better with a DUXL05D. This  has an output of 4.5 
mV per inch water and a span of 22.5 mV.
 
    1" water gauge corresponds to a pitot speed of  about 45 mph. This does 
not leave much, if any, safety margin for storm  conditions.

Amp: instrumentation,1000x,single supply,micropower,RTR out(INA122P) 
 
    While the INA122 has rail to rail output, it also  has a 0.1 to 10 Hz 
noise of 2 micro V pp. This is poor in comparison with an  INA118 which has a 
noise level of 0.28 micro V pp - roughly 1/7th. There are  other low noise opamps 
available.

Power: ~2ma 5VDC, derived from PC serial port, no exernal power  supply
Regulator: 5VDC, low dropout micropower, precision (LP2950-5.0)   
Firmware: ADC sampling at ~35 SPS, zero drift tracking ~5 min TC.
Serial  Out: RS232, 9600 baud, 8N1, ascii records, -32767 to +32767
Wind Spatial  Averaging: array of micro-perforated garden tubing 
Data Log/Display:  freeware PC AMASEIS seismic log/analysis software

Improvements? Offset  voltages in the instrumentation amp and pressure
sensor seem to limit the  maximum gain.  I may try a variable amp reference
voltage to eliminate  these offsets.
 

Use the average signal level as your  zero? 

Regards,
 
    Chris Chapman

 





In a message dated 09/01/2006, DSaum@............ writes:
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>System=20
  Bandpass: 0.01 to 4 Hz.
High-Pass filter: ~100 sec or 0.01 Hz using=20
  volume/capillary tube.
Low pass filter: 4 Hz anti-alias, 8 pole Bessel=20
  filter (MAX4701)
    This gives a drop of 40 dB at 4 Hz, an average=20
signal delay of 0.2 sec and 0.35 sec peak. 
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000=20
  size=3D2>Differential pressure sensor: silicon, bridge type, 1"wc=20
  (DUXL01D)
Pres Resolution: ~0.004 Pa - possibly lower depending=20
  offsets
Range: +/- 120 Pa (32768 times pressure=20
resolution)
    For a 5 V supply, this sensor has an output of=20=
6 mV=20
per inch water and a span of 6 mV. You might do better with a DUXL05D.=20=
This=20
has an output of 4.5 mV per inch water and a span of 22.5 mV.
 
    1" water gauge corresponds to a pitot speed of=20
about 45 mph. This does not leave much, if any, safety margin for storm=
=20
conditions.

Amp: instrumentation,1000x,single supply,micropower,RTR out(INA=
122P)=20

 
    While the INA122 has rail to rail output, it al=
so=20
has a 0.1 to 10 Hz noise of 2 micro V pp. This is poor in comparison with an=
=20
INA118 which has a noise level of 0.28 micro V pp - roughly 1/7th. There are=
=20
other low noise opamps available.

Power: ~2ma 5VDC, derived from PC serial port, no exernal power=20
supply
Regulator: 5VDC, low dropout micropower, precision (LP2950-5.0)&nb=
sp;=20

Firmware: ADC sampling at ~35 SPS, zero drift tracking ~5 min TC.
Ser=
ial=20
Out: RS232, 9600 baud, 8N1, ascii records, -32767 to +32767
Wind Spatial=20
Averaging: array of micro-perforated garden tubing 
Data Log/Display:=20
freeware PC AMASEIS seismic log/analysis software

Improvements? Offse=
t=20
voltages in the instrumentation amp and pressure
sensor seem to limit the=
=20
maximum gain.  I may try a variable amp reference
voltage to elimina=
te=20
these offsets.
 
    Use the average signal level as your=20
zero?
 
    Regards,
 
    Chris Chapman