In a message dated 2008/03/17, tchannel1@............ writes:

> Hi Chris,  Is this the type of Hall Effect Magnetic field detector chip  
> http://www.allegromicro.com/en/Products/Design/an/an27701.pdf

Hi Ted,

       This is the general applications sheet for the older linear devices. 
See 
http://www.allegromicro.com/en/Products/Part_Numbers/1321/index.asp for 
current sensors. You do need linear chopper stabilised devices.

       The A1321 looks suitable. I have only used the older A3515 and A3516s. 
You need a 5 milli H choke followed by a 0.1 muF ceramic capacitor to ground 
on the output, otherwise they are very noisy. I got up to ~ +/-15 nano metres 
resolution, but I only had about +/- 0.5 mm range. They have a fixed maximum 
gauss level at which they saturate, so the sharper you make the NS field 
junction, the greater the sensitivity, but the lower the range. I used INA118 
differential opamps.

       There is a seismometer circuit on Meredith's website. See 
http://www.geocities.com/meredithlamb/index.html 

       See if there are any online sources for the soft iron backed quad 
NdFeB magnet systems used in computer hard drives? These butterfly magnets are 
shown being used on http://www.geocities.com/meredithlamb/page008.html

       You can also make up a square of four square NdFeB magnets and place 
four sensors over the joints, connected in differential pairs.

>  I see one chip can determined direction. X and Y? If this is so, is there 
> a way to apply that directional information to trip a relay or otherwise 
> light a series of LED's
> They look inexpensive, $2, so if not one, we could use several, (8) to form 
> a compass.

       You can ~use them for a rough compass, but they don't really have 
sufficient sensitivity or stability on their own. 5 mV / gauss is not very high, 
when the Earth's field may be only ~ 0.18 gauss! You can use T shaped ferrite 
flux concentrators, or mu metal, but ferrite fluxgate ring sensors are very 
much better - different technology.
       I did try using lengths of  T shaped 1/8" galvanised mild steel 
fencing wire with fair results for monitoring noise / changes in the Earth's field 
and for detecting vehicle movements. You need several yards of wire.

       Once you get out a low noise amplified signal, you can do what you 
like with it.
       You can get LM3914, 3915 and 3916 IC drivers for LED bargraphs and 
also DIL LED ICs in various colors.

       You can get dedicated compass IC modules, but I seem to remember that 
they were not cheap when I last looked about a year ago.

       Hope that this helps.

       Regards,

       Chris Chapman   
In a me=
ssage dated 2008/03/17, tchannel1@............ writes:



Hi Chris,  Is this the typ=
e of Hall Effect Magnetic field detector chip  http://www.allegromicro.co=
m/en/Products/Design/an/an27701.pdf




Hi Ted,



       This is the general applications sheet=20=
for the older linear devices. See 

http://www.allegromicro.com/en/Products/Part_Numbers/1321/index.asp for curr=
ent sensors. You do need linear chopper stabilised devices.



       The A1321 looks suitable. I have only u=
sed the older A3515 and A3516s. You need a 5 milli H choke followed by a 0.1=
 muF ceramic capacitor to ground on the output, otherwise they are very nois=
y. I got up to ~ +/-15 nano metres resolution, but I only had about +/- 0.5=20=
mm range. They have a fixed maximum gauss level at which they saturate, so t=
he sharper you make the NS field junction, the greater the sensitivity, but=20=
the lower the range. I used INA118 differential opamps.



       There is a seismometer circuit on Mered=
ith's website. See http://www.geocities.com/meredithlamb/index.html 



       See if there are any online sources for=
 the soft iron backed quad NdFeB magnet systems used in computer hard drives=
? These butterfly magnets are shown being used on http://www.geocities.com/m=
eredithlamb/page008.html



       You can also make up a square of four s=
quare NdFeB magnets and place four sensors over the joints, connected in dif=
ferential pairs.



 I see one chip can determined=
 direction. X and Y? If this is so, is there a way to apply that directional=
 information to trip a relay or otherwise light a series of LED's

They look inexpensive, $2, so if not one, we could use several, (8) to form=
 a compass.




       You can ~use them for a rough compass,=20=
but they don't really have sufficient sensitivity or stability on their own.=
 5 mV / gauss is not very high, when the Earth's field may be only ~ 0.18 ga=
uss! You can use T shaped ferrite flux concentrators, or mu metal, but ferri=
te fluxgate ring sensors are very much better - different technology.

       I did try using lengths of  T shap=
ed 1/8" galvanised mild steel fencing wire with fair results for monitoring=20=
noise / changes in the Earth's field and for detecting vehicle movements. Yo=
u need several yards of wire.



       Once you get out a low noise amplified=20=
signal, you can do what you like with it.

       You can get LM3914, 3915 and 3916 IC dr=
ivers for LED bargraphs and also DIL LED ICs in various colors.



       You can get dedicated compass IC module=
s, but I seem to remember that they were not cheap when I last looked about=20=
a year ago.



       Hope that this helps.



       Regards,



       Chris Chapman