In a message dated 2007/03/01, mckimzey@........... writes:

> You wrote:
> >        An alternative is to use a flat rectangular coil with quad NdFeB
> >magnets on mild steel backing plates. You might find this construction 
> >easier? See
> >drawings at http://jclahr.com/science/psn/chapman/lehman/index.html The 
> >more
> >powerful NdFeB magnets can give larger magnetic fields in a wider gap.
> 
> I built a damping assembly for my AS1-copy and used this type of dampening. 
> For it's use here, I guess I would use a coil (maybe 
> like a  220v relay coil?) between the magnets and instead of using it as the 
> 
> actual sensor (like on the AS1), it would rather function only as a 
> damper...do I understand correct?

Hi Mike,

       That is not totally correct.

       For a STM-8 you need a active force feedback coil. Placing a flat coil 
in the central high field position and putting a current through it allows 
you to generate a linear force / current relationship parallel to the magnet 
faces. But you are much better off using a thin rectangular former and a fairly 
narrow magnet gap, rather than a round relay coil. Make the former using fairly 
thin sheet metal with a radial slot + PCB which overlaps the ends of the 
joint in the magnets. If you are putting much current through the coil, you will 
likely need to cool it. There is a plan of a suitable sensor coil somewhere in 
my diagrams, but for force feedback you want thicker wire and a resistance of 
maybe 100 Ohms or less. If you try to use a high impedance / high inductance 
sensor type coil, you are likely to have problems making the circuit stable. I 
seem to remember that coil systems and forces were discussed, but using 
speaker magnets and cylindrical coils. 

> >        You need to be able to measure the force / current characteristic 
> and
> >the resistance to calculate the feedback required in a triple loop system.
> 
> It is this part that scares me. I don't have a really good backround in 
> electronic theory to understand some of the fine points of this stuff. If I 
> went ahead with this some time in the future, are there instructions 
> somewhare that could tell me how to measure and calculate these values?  I 
> know that in reading though STM's emails that there is mathlab formula's and 
> 
> other such stuff....but I don't even know what mathlab IS! 

    Matlab is a fairly large and complicated mathematics spreadsheet software 
program. It is fairly old now, but I have not used it. It allows you to plot 
graphical system responses.

       You need a multimeter, a power supply and a wirewound variable 
resistor. You weigh the coil + former, suspend it freely in the centre of the magnet 
assembly and put enough DC current through it to take it's own weight. You 
measure the current and the coil resistance with the multimeter. I have some 
small electronic digital pocket scales which allow me to measure to 0.1 gm. Do you 
have access to a chemical lab balance / know anyone who has? You don't need 
1% precision!

       Regards,

       Chris Chapman   
In a me=
ssage dated 2007/03/01, mckimzey@........... writes:



You wrote:

>        An alternative is to use a fl=
at rectangular coil with quad NdFeB

>magnets on mild steel backing plates. You might find this construction <=
BR>
>easier? See

>drawings at http://jclahr.com/science/psn/chapman/lehman/index.html The=20=


>more

>powerful NdFeB magnets can give larger magnetic fields in a wider gap.


I built a damping assembly for my AS1-copy and used this type of dampening.=20=


For it's use here, I guess I would use a coil (maybe 

like a  220v relay coil?) between the magnets and instead of using it a=
s the 

actual sensor (like on the AS1), it would rather function only as a 

damper...do I understand correct?




Hi Mike,



       That is not totally correct.



       For a STM-8 you need a active force fee=
dback coil. Placing a flat coil in the central high field position and putti=
ng a current through it allows you to generate a linear force / current rela=
tionship parallel to the magnet faces. But you are much better off using a t=
hin rectangular former and a fairly narrow magnet gap, rather than a round r=
elay coil. Make the former using fairly thin sheet metal with a radial slot=20=
+ PCB which overlaps the ends of the joint in the magnets. If you are puttin=
g much current through the coil, you will likely need to cool it. There is a=
 plan of a suitable sensor coil somewhere in my diagrams, but for force feed=
back you want thicker wire and a resistance of maybe 100 Ohms or less. If yo=
u try to use a high impedance / high inductance sensor type coil, you are li=
kely to have problems making the circuit stable. I seem to remember that coi=
l systems and forces were discussed, but using speaker magnets and cylindric=
al coils. 



>    &nb=
sp;   You need to be able to measure the force / current character=
istic and

>the resistance to calculate the feedback required in a triple loop syste=
m.



It is this part that scares me. I don't have a really good backround in 

electronic theory to understand some of the fine points of this stuff. If I=20=


went ahead with this some time in the future, are there instructions 

somewhare that could tell me how to measure and calculate these values? =
; I 

know that in reading though STM's emails that there is mathlab formula's and=
 

other such stuff....but I don't even know what mathlab IS! 




    Matlab is a fairly large and complicated mathematics spre=
adsheet software program. It is fairly old now, but I have not used it. It a=
llows you to plot graphical system responses.



       You need a multimeter, a power supply a=
nd a wirewound variable resistor. You weigh the coil + former, suspend it fr=
eely in the centre of the magnet assembly and put enough DC current through=20=
it to take it's own weight. You measure the current and the coil resistance=20=
with the multimeter. I have some small electronic digital pocket scales whic=
h allow me to measure to 0.1 gm. Do you have access to a chemical lab balanc=
e / know anyone who has? You don't need 1% precision!



       Regards,



       Chris Chapman  =20