In a message dated 07/12/2007, paleoartifact@......... writes:

The "sticky" terminology in particular is that yes; it is normally a very  
real problem.  It "seems", to be most visually prevalent when only a  single 
piece of pyrolytic graphite is used.  By "sticky" I mean it seems  to sometimes 
be anchored in one or more spots and even with some light induced  tilt where 
one would surely expect a movement response; they occasionally do  not get any. 
 Of course, I'am not referring to debris (small hair, etc.)  that may 
actually be the cause sometimes.  With PG, its possible to  impart small magnetic 
particles thereon which will react in the magnetic  fields of course.

Hi Meredith,
 
    I suspect that this is due to slightly differing  field strengths along 
the magnets. It is very troublesome when you are using  several magnets joined 
in line.

Actually the ONLY reason, I'am trying again is that with two separated  but 
interconnected pieces of
PG, I know I'am seeing quite a improvement in overall response.  I  can't 
really presently describe the "why"; except perhaps the joint sum of the  two 
pieces are kind of like a differential diamagnetic
response....

    The sum of two pieces should help smooth out local  magnetic 
fluctuations. 

 I've also looked at totally switching out the PG, with other  different 
pieces, and the results
look the same; so its obviously not just specific to acouple PG  pieces.  
Also, the PG pieces have
been on and off the magnets numerious time for various test trials; so  the 
results are real.  The PG
pieces are also fairly close to each other in dimensions; which might be  
necessary (?).  The PG
pieces I have are not precision machined so they vary in flatness and in  
other dimensions by around
~ 0.010".

    You might try using emery paper on a flat surface  to give a flat under 
surface to the PG pieces? 

Another feature I like is the variable period with the variable PG  spacing.  
How I'am going to do that
with flimsy aluminum is yet to be attempted.  One may have to settle  for 
whatever period they select before any "glueing" down of the  aluminum.  Even 
being able to do so, is unique.

    Can you buy some copper foil from K&S Metals  and use that? They stock 3 
and 5 thou sheet. The Aluminum usually supplied is a  Si alloy, not pure and 
it is paramagnetic. Copper is diamagnetic but  far weaker. I ran up against 
these differences / problems when using NdFeB  quad magnets for damping the 
Lehman.   

I've even put a large optical magnifying lens next to the model; to check  
for minute lockups, and
of course whether it physically oscillates around a reference "zero" mark  I 
put on the magnet.  I can't say I've definitely seen any sticking yet  on this 
model; but if it happens, I will.

    If you use a small lens to give a ~parallel light  beam from a small bulb 
at rignt angles to the slip direction, small side  movements probably won't 
be too obvious. 

One good indication that it works better is that I do see increased wider  
span tilt oscillation motion
effects with the 2 PG pieces than I've ever seen with one piece of  PG.  It 
seems to be ~ 3X over
that of just using 1 piece of PG.
 
I don't yet really know the result of adding on steel/iron atop the  magnets; 
as I've not yet done so.
I think I'll try 1/2" width X 1/8" thick iron initially.  The iron  is 
commercially common size stuff, but
it does have dual rounded outer width edges; which might not be totally  
ideal.  The size is common;
but its different, as it was used as a desk drawer guide and was coated  with 
a thin layer of copper
and then chrome plated.  Its probably (?) bright steel and not the  black 
structural variety.
 
I've seen thin ~ 1/16" thick iron/steel work very well on kind of a "U"  
channel type setup...where the
(then) spectrographite rod levitated lower down on the deliberately lower  
middle channel magnet.
That was Chris's recommendation and it did dramatically improve its  
levitation height. 

    I bought some 1/32" galvanised steel sheet, which  seemed to work quite 
well. I suspect that 1/8" will be much too thick and it  will allow the field 
to be linked sideways, instead of looped over the top of  the magnets and going 
through the graphite. Iron will carry about twice the  field that the magnet 
can supply. You might consider filing the edges of the  strips at 45 degrees, 
rather than leaving it at a right angle?

Yes; the aluminum paramagnetism has been seen in the past also; and  its 
going to show up here
also.  The aluminum could be responsible for some of the sticking,  with 
their various impurities.
I am using K&S Engineering aluminum; which seems to be purer  stuff.  With 
the severe weight
limitations; its hard to consider using other metal/s.  I suppose I  can hang 
up a piece of the stuff
on a long thread, and bring a magnet near it; to just see a rough  indication 
of its general paramagnetism.

    K&S supply mostly 6061 Al alloy
    See _www.ksmetals.com_ (http://www.ksmetals.com)  and 
_http://en.wikipedia.org/wiki/6061_aluminum_ (http://en.wikipedia.org/wiki/6061_aluminum)    

I am also presently using 2 plastic spacers in between the magnets; which  
are 0.035" thick X ~0.475" in width.  The spacer is a idea from John  Lahr; and 
it does help a little for some slightly increased levitation.   I may try to 
find some thin plastic or non-ferrious metal that is 5/8" in  width and 
somewhere around 0.020 to 0.0315" thick, that could aid in setting  up the iron on 
the magnet tops.  The iron will be magnetically  normally forced against the 
joints for alignment.
 
    Regards,
    
    Chris Chapman




   





In a message dated 07/12/2007, paleoartifact@......... writes:
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
The "sticky" terminology in particular is that yes; it is normally a=20=
very=20
  real problem.  It "seems", to be most visually prevalent when only a=20
  single piece of pyrolytic graphite is used.  By "sticky" I mean it se=
ems=20
  to sometimes be anchored in one or more spots and even with some light ind=
uced=20
  tilt where one would surely expect a movement response; they occasionally=20=
do=20
  not get any.  Of course, I'am not referring to debris (small hair, et=
c.)=20
  that may actually be the cause sometimes.  With PG, its possible to=20
  impart small magnetic particles thereon which will react in the magnetic=20
  fields of course.
Hi Meredith,
 
    I suspect that this is due to slightly differin=
g=20
field strengths along the magnets. It is very troublesome when you are using=
=20
several magnets joined in line.
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
Actually the ONLY reason, I'am trying again is that with two separate=
d=20
  but interconnected pieces of
  
PG, I know I'am seeing quite a improvement in overall response. =
 I=20
  can't really presently describe the "why"; except perhaps the joint sum of=
 the=20
  two pieces are kind of like a differential diamagnetic
  
response....
    The sum of two pieces should help smooth out lo=
cal=20
magnetic fluctuations. 
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
 I've also looked at totally switching out the PG, with other=20
  different pieces, and the results
  
look the same; so its obviously not just specific to acouple PG=20
  pieces.  Also, the PG pieces have
  
been on and off the magnets numerious time for various test trials; s=
o=20
  the results are real.  The PG
  
pieces are also fairly close to each other in dimensions; which might=
 be=20
  necessary (?).  The PG
  
pieces I have are not precision machined so they vary in flatness and=
 in=20
  other dimensions by around
  
~ 0.010".
    You might try using emery paper on a flat surfa=
ce=20
to give a flat under surface to the PG pieces? 
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
Another feature I like is the variable period with the variable PG=20
  spacing.  How I'am going to do that
  
with flimsy aluminum is yet to be attempted.  One may have to se=
ttle=20
  for whatever period they select before any "glueing" down of the=20
  aluminum.  Even being able to do so, is unique.
    Can you buy some copper foil from K&S Metal=
s=20
and use that? They stock 3 and 5 thou sheet. The Aluminum usually supplied i=
s a=20
Si alloy, not pure and it is paramagnetic. Copper is diamagnetic but=20
far weaker. I ran up against these differences / problems when using Nd=
FeB=20
quad magnets for damping the Lehman.   
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
I've even put a large optical magnifying lens next to the model; to c=
heck=20
  for minute lockups, and
  
of course whether it physically oscillates around a reference "zero"=20=
mark=20
  I put on the magnet.  I can't say I've definitely seen any sticking y=
et=20
  on this model; but if it happens, I will.
    If you use a small lens to give a ~parallel lig=
ht=20
beam from a small bulb at rignt angles to the slip direction, small sid=
e=20
movements probably won't be too obvious. 
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
One good indication that it works better is that I do see increased w=
ider=20
  span tilt oscillation motion
  
effects with the 2 PG pieces than I've ever seen with one piece of=20
  PG.  It seems to be ~ 3X over
  
that of just using 1 piece of PG.
  
 
  
I don't yet really know the result of adding on steel/iron atop=20=
the=20
  magnets; as I've not yet done so.
  
I think I'll try 1/2" width X 1/8" thick iron initially.  The ir=
on=20
  is commercially common size stuff, but
  
it does have dual rounded outer width edges; which might not be total=
ly=20
  ideal.  The size is common;
  
but its different, as it was used as a desk drawer guide and was coat=
ed=20
  with a thin layer of copper
  
and then chrome plated.  Its probably (?) bright steel and not t=
he=20
  black structural variety.
  
 
  
I've seen thin ~ 1/16" thick iron/steel work very well on kind of a "=
U"=20
  channel type setup...where the
  
(then) spectrographite rod levitated lower down on the deliberately l=
ower=20
  middle channel magnet.
  
That was Chris's recommendation and it did dramatically improve its=20
  levitation height. 
    I bought some 1/32" galvanised steel sheet, whi=
ch=20
seemed to work quite well. I suspect that 1/8" will be much too thick and it=
=20
will allow the field to be linked sideways, instead of looped over the top o=
f=20
the magnets and going through the graphite. Iron will carry about twice the=20
field that the magnet can supply. You might consider filing the edges of the=
=20
strips at 45 degrees, rather than leaving it at a right angle?
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
Yes; the aluminum paramagnetism has been seen in the past also; and=20
  its going to show up here
  
also.  The aluminum could be responsible for some of the stickin=
g,=20
  with their various impurities.
  
I am using K&S Engineering aluminum; which seems to be purer=20
  stuff.  With the severe weight
  
limitations; its hard to consider using other metal/s.  I suppos=
e I=20
  can hang up a piece of the stuff
  
on a long thread, and bring a magnet near it; to just see a rough=20
  indication of its general paramagnetism.
    K&S supply mostly 6061 Al alloy
    See www.ksmetals.com and http://en.wikipedia.org/=
wiki/6061_aluminum=20

    
I am also presently using 2 plastic spacers in between the magnets;=
 which=20
are 0.035" thick X ~0.475" in width.  The spacer is a idea from Jo=
hn=20
Lahr; and it does help a little for some slightly increased levitation. =
;=20
I may try to find some thin plastic or non-ferrious metal that is 5/8"=20=
in=20
width and somewhere around 0.020 to 0.0315" thick, that could aid in setting=
=20
up the iron on the magnet tops.  The iron will be magnetically=20
normally forced against the joints for alignment.
 
    Regards,
    
    Chris Chapman