Starting at the beginning with Mirror Screws

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Postby Viewmaster » Fri Sep 16, 2011 7:39 pm

gary wrote:In Steve's case I would have thought that the hardening process would also be the hardest part (no pun intended).

Steve, this is probably teaching grandmother to suck eggs, but when specifying the holes make sure either your specified size or the laser cutter operator (via a specification) takes the laser kerf into account otherwise the pins will drop straight through or will wobble in the hole.


Oh............ There is NO NEED for any hardening at ALL.

My sole reason in reccomending silver steel was it's very accurate dimensional diameters and fine surface finish.
Steve could give the lazer cutter a sample of the shaft too, to make sure of a tight fit in the slat holes asd well as the pins.
But he knows this anyway.
Just to recap....No hardening of SS.
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby gary » Fri Sep 16, 2011 7:54 pm

Sure, I was thinking of hardening the bearing surface, but that is probably not necessary.
gary
 

Postby Panrock » Fri Sep 16, 2011 8:10 pm

Thanks for the tips chaps - I'll proceed with care.

However, I have now hit another snag.

I was still a little unclear about how to predict the final width of the picture so I thought it would be a good idea to set up a test rig. This morning I have set up two polished sections of aluminium angle, 240mm long, fixed together by a central pivot (held in a vice). I have set them at mutual angles of 3 and then 6 degrees.

To my horror, I found the 'width would be filled' at twice the viewing distance I had expected! 2 metres for 6 degrees/60 lines and 4 metres for 3 degrees/120 lines.

Of course, using a smaller mirror screw I could 'fill' it at a reduced viewing distance. But the angle subtended by the picture (visual size) would be the same.

The deal seems to be that with mirror screws, for high line counts you always see a small picture - period. It won't be possible to examine the exquisite line structure of my picture because it'll be right across the room. Maybe I should include a telescope in the kit... :)

The only solution I can think of is to use not flat, but convex mirror surfaces. Presumably the curve would be parabolic? For the very shallow curve expected - I presume a simple circular profile would be fine.

Back to the drawing board!

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby Viewmaster » Fri Sep 16, 2011 9:00 pm

Panrock wrote:The only solution I can think of is to use not flat, but convex mirror surfaces. Presumably the curve would be parabolic? For the very shallow curve expected - I presume a simple circular profile would be fine.

Back to the drawing board!

Steve O


I don't have my own miror screw running at present but seem to recall that the viewing distance was also changed depending on the distance from the screw to the LED's too. My memory may be wrong here.

I would certainly get back to a BIG drawing board before embarking on making 120 'curved' slats, be they circular or parabolic.

Hey ho, stamp collecting was always much easier than NBTV. :-)
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby Panrock » Fri Sep 16, 2011 9:18 pm

Viewmaster wrote: I don't have my own miror screw running at present but seem to recall that the viewing distance was also changed depending on the distance from the screw to the LED's too. My memory may be wrong here.

Yes, from my test rig - the further away the line of light, the more the width (though it's a law of diminishing returns). BUT the closer the line of light, the less the problem with 'differential focus'. However I would imagine a close line of light to one side might compromise the line linearity... Bit of a nightmare these mirror screws!

Having said that, I am unable to use my 'test rig' now. After attempting to bend the strips of angle inward to mimic a convex surface, I simply buckled them and the whole thing is now useless for anything but the bin! :oops:

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby Panrock » Sat Sep 17, 2011 7:31 am

Never was good at maths but (see if you agree)...

...as it rotates, a convex surface 240mm across, that has a radius of curvature of (say) 2000mm should provide the effect of (240 / 2000 x 2 x Π) x 360 degrees of additional rotation. = 6.87 degrees)

Likewise a 1500mm radius of curvature surface should provide the effect of (240 / 1500 x 2 x Π ) x 360 degrees of extra rotation = 9.16 degrees.

The current flat surface scans over 3 degrees and has to be viewed 4 metres away to fill the 240mm line. I need a threefold increase.

Anyway, I've produced two more CAD pieces now, with curved surfaces of radius 1.5 and 2.0 metres, and I propose to have just a couple of each of them made next week, to try out the effect. Since frankly, I can't get my head round these optics - with all this doubling or halving of the expected effect going on - the only thing is to make the d*mned things and see what really happens...

My ideal would be a 12-inch (diagonal) 120 line picture viewable at a comfortable distance of about 1½ metres. But to start, I must make a couple of 2-line mirror screws! :D

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby Viewmaster » Sat Sep 17, 2011 6:44 pm

Yes, two accurately made test slats is the way to go, as I too, do not understand the optics of rotating curved surfaces and do not see what your 6.87 degrees as an arc of the large curved circumference of a 2000mm rad proves.

Another thing that might give me some thought is the distance from the curved surface to the shaft hole.
In the sketch would there be any differences in the picture obtained from these two extreme examples I wonder ?
Would there be pixel elongation at the top and bottom of a line scan for example?
Attachments
slat.JPG
slat.JPG (6.46 KiB) Viewed 15300 times
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby kareno » Sat Sep 17, 2011 8:14 pm

Hi Steve,

It is an often overlooked fact that, if a mirror rotates 3 degrees, the reflection of the room inside the mirror rotates 6 degrees. This is verified by periscopes - the mirrors are at 45 degrees yet turn the reflection through 90 degrees.

On the width of the slats - the mirrors are going to be more or less facing you when it is their turn for a line scan. So choose the width according to your desired aspect ratio, plus a bit more.

For instance, imagine the slats are all on their shaft bit aligned flat, as if you were going to polish them all. Suppose the stack is 120mm high and you want a 3:4 aspect ratio (i.e. a picture width of 160mm). I would go for slats of around 200mm.

This seems to invite overscan but you'll need that because the apparent width of the image area shrinks because each mirror is rotated away from you slightly when reflecting the light source in your direction.

If you give me the thickness of your slats (I hope I don't get barred for making such a request!) and the desired aspect ratio of your picture, I will try to work out the geometry for you.
kareno
 

Postby Panrock » Sat Sep 17, 2011 8:18 pm

Albert, (Karen. our posts crossed - I'll look at your comments next)...One of my concerns when setting the position of the shaft hole is of course - balance. Ideally it should be balanced fore-and-aft as well as side-to-side - for a trouble free ride later. Yes, the bunch of slats as a whole may (in other words I don't really know) cancel out each other's imbalances, but let's keep things as predictable as we can.

Don't know if your diagram is simply showing a cutaway or is depicting the reflecting face curve. If so, the curve is convex not concave.

I had previously wondered about your point concerning the distance of the fulcrum of the reflecting surface from the shaft. This would mean its distance to the observer would vary as the mirror surface rotated, in theory causing non linearity? But the active point on the mirror slat varies its distance anyway when it rotates! Because this variation is only a small proportion of the viewing distance, I presume its effect will be minimal.

The usual warnings about the danger of me 'talking through my hat' apply! :lol:

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby Panrock » Sat Sep 17, 2011 8:40 pm

kareno wrote:It is an often overlooked fact that, if a mirror rotates 3 degrees, the reflection of the room inside the mirror rotates 6 degrees. This is verified by periscopes - the mirrors are at 45 degrees yet turn the reflection through 90 degrees.

I imagine this might have something to do with the measured viewing distance in my tests turning out to be twice what I had expected... :oops:

kareno wrote:On the width of the slats - the mirrors are going to be more or less facing you when it is their turn for a line scan. So choose the width according to your desired aspect ratio, plus a bit more.

Useful tip, I hadn't considered this.

kareno wrote:For instance, imagine the slats are all on their shaft bit aligned flat, as if you were going to polish them all. Suppose the stack is 120mm high and you want a 3:4 aspect ratio (i.e. a picture width of 160mm). I would go for slats of around 200mm.

My stack is 180mm high and the slat lengths currently 240mm. This was based on the 4:3 ratio.

kareno wrote:If you give me the thickness of your slats (I hope I don't get barred for making such a request!) and the desired aspect ratio of your picture, I will try to work out the geometry for you.

Very kind - yes I'd like to take you up on this. Thanks! The work I've so far done on this is at
http://www.taswegian.com/NBTV/forum/vie ... c&start=18

See whether you think it makes sense. :wink:

Steve O

P.S. I thought it would be helpful to add a .gif showing the CAD design. Of course, the original is in vector format, so doesn't have the jagged edges of this representation.
Attachments
slat.gif
(3 KiB) Downloaded 267 times
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby Harry Dalek » Sun Sep 18, 2011 12:23 am

Hi i came across these patents some time ago ,i rather like the Triangle one..
may be of some help.
Attachments
2112390_SCANNING_DRUM_CONSTRUCTION.pdf
(297.79 KiB) Downloaded 791 times
2192376_HIGH_DEFINITION_TELEVISION_APPAR.pdf
(375.46 KiB) Downloaded 763 times
2202795_MULTIPLE_SCANNING_SYSTEM.pdf
(249.84 KiB) Downloaded 766 times
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Postby Viewmaster » Sun Sep 18, 2011 12:40 am

Panrock wrote:Don't know if your diagram is simply showing a cutaway or is depicting the reflecting face curve. If so, the curve is convex not concave.


O dear, is my face red? That's through hurrying. It's supposed to be a plan view of one slat......er concavity notwithstanding. :-)

I can just see now that 12" x9" 120 line colour TV of yours purring away at 25RPS in the corner, with the gentle whistle of wind from the slats.
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby Viewmaster » Sun Sep 18, 2011 1:20 am

Steve, looking at Harry's interesting sketches, the first sketch (2112390)shows the shaft (No 24) which has a lock nut threaded onto the end as many mirror screw shafts have.
As you seem to have settled on a very large diameter shaft of 1 inch I was wondering what slat clamping method you might have considered?

Are you planning to have the shaft turned down to a smaller diameter with smaller lock nuts, I wonder? It would need quite a large lathe to 'chuck' 1 inch dia to do this.
Maybe you have such a beast.
No doubt you have considered alternative ways to clamp the slats other than lock nuts anyway.
Just a thought.
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby Panrock » Sun Sep 18, 2011 1:41 am

Albert, It's good you've put me on the spot about this. I admit I haven't as yet looked at the clamping method (like a number of other things). I kind of imagined the slats would fit tightly on the shaft anyway; moreover for 120 of them there would be 120 times as much friction, and because they would be all locked together "they wouldn't go anywhere". So, to answer your question, I was going to do nothing! Am I stupid? (No - don't answer that! :oops: )

Steve O

PS. Harry, thanks for the pics!
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby Viewmaster » Sun Sep 18, 2011 7:03 pm

Panrock wrote:I kind of imagined the slats would fit tightly on the shaft anyway; moreover for 120 of them there would be 120 times as much friction, and because they would be all locked together "they wouldn't go anywhere". So, to answer your question, I was going to do nothing!
Steve O
PS. Harry, thanks for the pics!


You may well be right if slats are a near force fit onto the shaft. Although it is difficult to see what others have done in their photos there does seem to be some clamping at the top of vertical screws that I have seen on the net.

With, maybe, some slight vibration present and being in the vertical plane, there is an awful temptation on Murphy's law to let the top slats work themselves upward :-) I assume there will be, at least, some bottom stop.

However, if it were mine with zero clamping, I would have a back up position in that some clamping method could be made and easily be added if required, without interfering with the existing design at all.

(such as cross bars through the shaft with clamp screws or shaft collars and clamp screws etc.)
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

PreviousNext

Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 31 guests

cron