In a message dated 17/09/02, charles.r.patton@........ writes:

Hi Charles,

> A sensor solution for the SG seimo, it was proposed to use "=E2=80=A6a $1.=
83

> A3515LUA sensor=E2=80=A6"

>=20

> So this brings up this thought: Most standard IC frames are magnetic -

> consisting of a Kovar frame. =20

       This one is non magnetic.

And magnetic sensors in particular are o
> ften constructed with "magnetic concentrators" to focus magnetic flux on t=
o=20
> the sensitive area of the chip. =20

       No ferrite or other concentrator is used.

Both make the chip physically=20
> attracted to ambient fields. =20

       There is NO PROBLEM in this case. Didn't you check the data sheet?=20
=20
If the chip is not mounted on the SG arm,=20
> then the magnets must be, which would be even worse about ambient=20
> sensitivity.  So if one goes for the higher sensitivity of the SG, doesn't=
=20
> this create problems in terms of an actual physical sensitivity
> to the surrounding changing magnetic fields? =20

       No. You can get problems if you fail to take sensible design=20
precautions, or do something real dumb. The field generated by these magnets=
=20
is basically quadrupole, not dipole. It is quite easy to shield the magnet(s=
)=20
and the sensor with a U of mild steel and even easier to use two flat plates=
=20
and two separators. Allegro show suitable shielded sensor housings in their=20
applications sheet. The high sensitivity arises from the intense fields clos=
e=20
to the surface of the NdBFe magnets, usually mounted within a shield housing=
..=20
The A3515 has a typical sensitivity of 5 mV / gauss, so stray earth fields=20
are not generally a problem and definitely not when it is shielded. =20

I'm thinking back to some=20
> experiments that Roger Baker did on magnetometers in which his magnetomete=
r=20
>=20

       ????? I am definitely NOT Roger Baker and this is definitely NOT a=20
magnetometer application!=20

>=20

       While this may be very interesting, how is it relevant to an A3515=20
magnetic sensor, please?=20

>=20

       This ancient design uses ferrite tuned circuits and diode=20
rectification, both of which are inherently temperature sensitive, very much=
=20
so in the case of the diodes. This type of sensor was dropped by=20
seismologists well over 20 years ago for output drift problems, amongst=20
others. Can you provide any experimental noise and thermal stability=20
measurements for such a system, to back up your 'vote' decision, please?
=20
Today you can buy ICs which do all the detection.  For one source see=20
"Universal=20
> Capacitive ReadoutTM IC (MS3110)" at:=20
> http://www.microsensors.com/products.html
> I don't know how much it costs, but in volume this can't be that expensive=
..=20
>  I also know there is at least a couple of European companies that makes a=
=20
>=20

       The MS3110 is supplied in 16 pin surface mount SOIC housing with a 5=20=
V=20
supply.
       You can either program the chip ROM of you can input serial data into=
=20
the chip, otherwise you can't use it.=20
       The minimum output filter frequency is 500 Hz, two pole.
       There is no drive provided for a screen for an input sensor wire /=20
line.
       The dynamic range does not seem to be quoted. It is very often 10^-4=20
in commercial devices which use this type of sensor, like the ADXL05. While=20
this is fine for airbag control on cars, seismic sensor applications require=
=20
a greater dynamic range.=20

       You can buy a MS3110 development board with 5 SOIC devices for $199.=20
This enables you to programme SOIC chips. The minimum order quantity for the=
=20
MS3110 is 25 off at $13-64 each.

       With these limitations, how do you suggest that the MS3110 could be=20
applied to our amateur seismic equipment, please? I fully appreciate that=20
capacitative sensors can give superb results, but this device does not quite=
=20
look a 'first choice' for amateur use.

       What other companies are producing devices similar to the MS3110,=20
please?

       Regards,

       Chris Chapman
      =20
> Aa a sensor solution for the SG seimo, it was proposed to use "=E2=80=A6a=20=
$1.83

> A3515LUA sensor=E2=80=A6"

>=20

> So this brings up this thought:  Most standard IC frames are magnetic -=20
> consisting of a Kovar frame.  And magnetic sensors in particular are often=
=20
> constructed with "magnetic concentrators" to focus magnetic flux on to the=
=20
> sensitive area of the chip.  Both make the chip physically attracted to=20
> ambient fields.  If the chip is not mounted on the SG arm, then the magnet=
s=20
> must be, which would be even worse about ambient sensitivity. So if one=20
> goes for the higher sensitivity of the SG,
> doesn't this create problems in terms of an actual physical sensitivity to=
=20
> the surrounding changing magnetic fields? I'm thinking back to some=20
> experiments that Roger Baker did on magnetometers in which his magnetomete=
r=20
> is actually a magnet on a torsion balance. He said he could detect a small=
=20
> magnet many feet away and I know he could sense geomagnetic changes. It=20
> would seem to me that would turn the SG seismo into a geomagnetic field=20
> detector. Roger's detectors were based on photodetectors and LED or Laser=20
> LED sources.
>=20
> I think I'd vote for the original SG variable capacitance detector.=20
> Today you can buy ICs which do all the detection.  For one source see
> "Universal Capacitive ReadoutTM IC (MS3110)" at:
> http://www.microsensors.com/products.html
> I don't know how much it costs, but in volume this can't be that expensive=
..=20
>  I also know there is at least a couple of European companies that makes a=
=20
> similar IC.=20
> Regards,
>=20

In a message dated 17/09/=
02, charles.r.patton@........ writes:



Hi Charles,



A sensor solution for the S=
G seimo, it was proposed to use "=E2=80=A6a $1.83
A3515LUA sensor=E2=80=
=A6"

So this brings up this thought: Most standard IC frames are magn=
etic -
consisting of a Kovar frame.  



       This one is non magnetic.



And magnetic sensors in particular are o

ften constructed with "magn=
etic concentrators" to focus magnetic flux on to the sensitive area of the c=
hip.  



       No ferrite or other concentrator is=
 used.



Both make the chip physically=20

attracted to ambient fields=
..  



       There is NO PROBLEM in this case. D=
idn't you check the data sheet?=20

=20

If the chip is not mounted on the SG arm,=20

then the magnets must be, w=
hich would be even worse about ambient sensitivity.  So if one goes for=
 the higher sensitivity of the SG, doesn't this create problems in terms of=20=
an actual physical sensitivity

to the surrounding changing magnetic fields?  



       No. You can get problems if you fai=
l to take sensible design precautions, or do something real dumb. The field=20=
generated by these magnets is basically quadrupole, not dipole. It is quite=20=
easy to shield the magnet(s) and the sensor with a U of mild steel and even=20=
easier to use two flat plates and two separators. Allegro show suitable shie=
lded sensor housings in their applications sheet. The high sensitivity arise=
s from the intense fields close to the surface of the NdBFe magnets, usually=
 mounted within a shield housing. The A3515 has a typical sensitivity of 5 m=
V / gauss, so stray earth fields are not generally a problem and definitely=20=
not when it is shielded.  



I'm thinking back to some=20

experiments that Roger Bake=
r did on magnetometers in which his magnetometer is actually a magnet on a t=
orsion balance. 



       ????? I am definitely NOT Roger Bak=
er and this is definitely NOT a magnetometer application!=20



 Roger's detectors were bas=
ed on photodetectors and LED or Laser LED sources.



       While this may be very interesting,=
 how is it relevant to an A3515 magnetic sensor, please?=20



I think I'd vote for the or=
iginal SG variable capacitance detector. 



       This ancient design uses ferrite tu=
ned circuits and diode rectification, both of which are inherently temperatu=
re sensitive, very much so in the case of the diodes. This type of sensor wa=
s dropped by seismologists well over 20 years ago for output drift problems,=
 amongst others. Can you provide any experimental noise and thermal stabilit=
y measurements for such a system, to back up your 'vote' decision, please?

=20

Today you can buy ICs which do all the detection.  For one source s=
ee "Universal=20

Capacitive ReadoutTM IC (MS=
3110)" at: http://www.microsensors.com/products.html

I don't know how much it costs, but in volume this can't be that expensi=
ve.  I also know there is at least a couple of European companies that=20=
makes a similar IC. 



       The MS3110 is supplied in 16 pin su=
rface mount SOIC housing with a 5 V supply.

       You can either program the chip ROM=
 of you can input serial data into the chip, otherwise you can't use it.=20

       The minimum output filter frequency=
 is 500 Hz, two pole.

       There is no drive provided for a sc=
reen for an input sensor wire / line.

       The dynamic range does not seem to=20=
be quoted. It is very often 10^-4 in commercial devices which use this type=20=
of sensor, like the ADXL05. While this is fine for airbag control on cars, s=
eismic sensor applications require a greater dynamic range.=20



       You can buy a MS3110 development bo=
ard with 5 SOIC devices for $199. This enables you to programme SOIC chips.=20=
The minimum order quantity for the MS3110 is 25 off at $13-64 each.



       With these limitations, how do you=20=
suggest that the MS3110 could be applied to our amateur seismic equipment, p=
lease? I fully appreciate that capacitative sensors can give superb results,=
 but this device does not quite look a 'first choice' for amateur use.



       What other companies are producing=20=
devices similar to the MS3110, please?



       Regards,



       Chris Chapman

       

Aa a sensor solution for th=
e SG seimo, it was proposed to use "=E2=80=A6a $1.83
A3515LUA sensor=E2=
=80=A6"

So this brings up this thought:  Most standard IC frames=
 are magnetic - consisting of a Kovar frame.  And magnetic sensors in p=
articular are often constructed with "magnetic concentrators" to focus magne=
tic flux on to the sensitive area of the chip.  Both make the chip phys=
ically attracted to ambient fields.  If the chip is not mounted on the=20=
SG arm, then the magnets must be, which would be even worse about ambient se=
nsitivity. So if one goes for the higher sensitivity of the SG,

doesn't this create problems in terms of an actual physical sensitivity=20=
to the surrounding changing magnetic fields? I'm thinking back to some exper=
iments that Roger Baker did on magnetometers in which his magnetometer is ac=
tually a magnet on a torsion balance. He said he could detect a small magnet=
 many feet away and I know he could sense geomagnetic changes. It would seem=
 to me that would turn the SG seismo into a geomagnetic field detector. Roge=
r's detectors were based on photodetectors and LED or Laser LED sources.



I think I'd vote for the original SG variable capacitance detector.=20

Today you can buy ICs which do all the detection.  For one source s=
ee

"Universal Capacitive ReadoutTM IC (MS3110)" at:

http://www.microsensors.com/products.html

I don't know how much it costs, but in volume this can't be that expensi=
ve.  I also know there is at least a couple of European companies that=20=
makes a similar IC.=20

Regards,

Charles R. Patton