The Mihaly-Traub scanner
Hello
So if I wanted to make a Mihaly-Traub scanner, how many sides would my polygone need in order to keep the speed requirments down if I wanted to use it for PAL television? How big would the mirrors need to be, and in how big an arc would the outer mirrors be? Is a polygone in this sence like this shape:
-
/ \
\ - /
well sort of, if you can get the "picture"
Hello
So if I wanted to make a Mihaly-Traub scanner, how many sides would my polygone need in order to keep the speed requirments down if I wanted to use it for PAL television? How big would the mirrors need to be, and in how big an arc would the outer mirrors be? Is a polygone in this sence like this shape:
-
/ \
\ - /
well sort of, if you can get the "picture"
Greetings
Fiat1 A Mihaly-Traub scanner uses a double sided mirror mounted on a motor surrounded by a ring of stationary mirrors. You need one mirror per tv line. For PAL television thats 625 mirrors. PAL tv needs 25 frames per second your motor speed has to be 25 times 60= 1500 RPM. If you use 1/4 inch wide mirrors they will fit on a circle 4 feet in diameter. This system was used at that time tv lines numbers were low so not a great number of mirrors were needed. With PAL tv it is probably not practical to buy and mount and accurately position 625 mirrors.
One tv frame of 625 lines consists of two almost identical fields of 312.5 lines. Using 312 lines you need a polygon with say 8 sides and 39 mirrors. That gives you 312 lines in one revolution. or using 10 sided polygon you need 31 mirrors or 20 sided polygon you need 16 mirrors or a 52 sided polygon you need only 6 stationary mirrors. This is at 3000 RPM.
The number of polygon facets multiplied by the number of stationary mirrors should be 312
So as you increase the number of facets on the polygon you decrease the number of stationary mirrors. Drawing 312 lines in one revolution requires a motor speed of 3000 RPM. As you can see the motor speed is not high. People ask me all the time why not use a polygon with more facets because that reduces the number of staionary mirrors. The answer is simple. Cost. 8 sided polygons are available on the surplus market quite cheap. I have seen them several places for under $100. Getting a polygon manufactured is very expensive in the range of U.S $3000-$5000. I have a quote here for a 52 facet polygon complete with motor and controller from a Japanese company that is just over U.S. $3000. With a 52 facet polygon and 6 stationary mirrors at 3000 RPM I can get one tv field or at 18000 RPM it can draw one tv field with no stationary mirrors at a scan angle of 14 degrees. This would be excellent for laser scanning but in New Zealand U.S. $3000 comes to $8000. I can buy a good second hand car for that with air conditioning and all extras. So I have to go the cheap way which is using ex printer polygon with 8 facets and 39 stationary mirrors at 3000 RPM
Fiat1 A Mihaly-Traub scanner uses a double sided mirror mounted on a motor surrounded by a ring of stationary mirrors. You need one mirror per tv line. For PAL television thats 625 mirrors. PAL tv needs 25 frames per second your motor speed has to be 25 times 60= 1500 RPM. If you use 1/4 inch wide mirrors they will fit on a circle 4 feet in diameter. This system was used at that time tv lines numbers were low so not a great number of mirrors were needed. With PAL tv it is probably not practical to buy and mount and accurately position 625 mirrors.
One tv frame of 625 lines consists of two almost identical fields of 312.5 lines. Using 312 lines you need a polygon with say 8 sides and 39 mirrors. That gives you 312 lines in one revolution. or using 10 sided polygon you need 31 mirrors or 20 sided polygon you need 16 mirrors or a 52 sided polygon you need only 6 stationary mirrors. This is at 3000 RPM.
The number of polygon facets multiplied by the number of stationary mirrors should be 312
So as you increase the number of facets on the polygon you decrease the number of stationary mirrors. Drawing 312 lines in one revolution requires a motor speed of 3000 RPM. As you can see the motor speed is not high. People ask me all the time why not use a polygon with more facets because that reduces the number of staionary mirrors. The answer is simple. Cost. 8 sided polygons are available on the surplus market quite cheap. I have seen them several places for under $100. Getting a polygon manufactured is very expensive in the range of U.S $3000-$5000. I have a quote here for a 52 facet polygon complete with motor and controller from a Japanese company that is just over U.S. $3000. With a 52 facet polygon and 6 stationary mirrors at 3000 RPM I can get one tv field or at 18000 RPM it can draw one tv field with no stationary mirrors at a scan angle of 14 degrees. This would be excellent for laser scanning but in New Zealand U.S. $3000 comes to $8000. I can buy a good second hand car for that with air conditioning and all extras. So I have to go the cheap way which is using ex printer polygon with 8 facets and 39 stationary mirrors at 3000 RPM
Mihaly-Traub
A Mihaly-Traub also used a mirror with mulitpul sides, not just a double sided miror, that was used originally for a 60 line system. What is the shape of a polygone in this instance? With colour, if you didi not use a laser but a highly focused beam of light you an use a colour wheel. To reduce the speed of the colour wheel add say 3 sets of primary colours so one rotation takes care of 9 frames of colour! Gee you have got it worse then me and I am in Australia with the American dollar thing.
A Mihaly-Traub also used a mirror with mulitpul sides, not just a double sided miror, that was used originally for a 60 line system. What is the shape of a polygone in this instance? With colour, if you didi not use a laser but a highly focused beam of light you an use a colour wheel. To reduce the speed of the colour wheel add say 3 sets of primary colours so one rotation takes care of 9 frames of colour! Gee you have got it worse then me and I am in Australia with the American dollar thing.
hi,
I think, you cannot have a light focused as well as a laser.. and the size of the spot will increase with the distance.
And when you are using optical colour filters there is a problem with power losses.
It should be cheaper, easier but not a technological advance like Laser projection, will be.
pa
I think, you cannot have a light focused as well as a laser.. and the size of the spot will increase with the distance.
And when you are using optical colour filters there is a problem with power losses.
It should be cheaper, easier but not a technological advance like Laser projection, will be.
pa
Greetings
This will be my last message for two or three weeks. I am starting to gather firewood for winter. It is supposed to go in the wood shed. The wood shed is full of several motor mowers, rotary hoe, a large machine I built to cut up tree prunings, engine parts, gearboxes, garden tools etc etc. These should be in the storage shed which was demolished when a tree fell on it. So I have to rebuild the shed.
The reason I started this thread was to show my experiments with a laser projector. The idea was to use inexpensive parts which can be obtained at small cost. I bought a surplus polygon, a 5 milliwatt 670 nm red laser diode and some front surface mirrors from OATLEY ELECTRONICS in Australia, and some new laser diodes 650 and 635 nm from the LASER GUY AND DEHARPPORTE TRADING Co in USA. Total cost less than 200 dollars NZ. Those of you who have followed the thread from the start will see the emphasis was on getting a fast scan. Television needs a very fast scan to move the beam from left to right across the screen. Using a polygon to do the scanning needs a very high speed around 36000 - 45000 RPM which needs a new polygon made to operate at that speed and they are very expensive. A way had to be found to use lower speed or the system became too expensive and out of reach for DIY people.
For PAL tv 625 interlace line system the tv lines are transmitted 312.5 at a time. This is a tv field. The first 312.5 lines are transmitted, then a second set of 312.5 lines are transmitted which fit in between the first set just like two combs fit together. These two fields make up a complete frame. The fields have to be drawn on screen in a 50th of a second. Looking at that I thought if I could get 312.5 lines in one motor revolution the motor only has to run at 50 times 60 = 3000 RPM which is well within the speed range of surplus polygon motors.
Many surplus polygons have 8 facets which in one revolution would only give 8 tv lines however by using a system of stationary mirrors surrounding the polygon it is possible to increase the number of lines in one motor revolution. This can be done by aiming the laser at the mirror facet, reflecting onto a stationary mirror, back from the stationary mirror to the polygon facet and out to a frame scanner. As the polygon rotates each stationary mirror gives 8 tv lines. Using 39 stationary mirrors we have 39 times 8 = 312 tv lines. You can test this idea at home using one stationary mirror by rotating a polygon by hand and see that each time a new polygon facet comes around it causes a new scan line. Mirror width should be 1/2 inch wide and 8 inches away from the polygon. These are approximate mesurements but will prove the point that one stationary mirror gives 8 tv lines, two stationary mirrors gives 16 tv lines etc. Ordinary mirrors work alright. Buy a glass cutter my one is a German one model Diamantor 200.1 $7.50 NZ. To cut mirrors buy cheap thin ones from the 2 dollar shop, lay the mirror on several peices of newspaper on a flat base, put a LOT of weight into the glass cutter and it will score first time then a slight tap and it will snap off. Be very careful to clean up after cutting with a vacuum cleaner otherwise you get little pieces of glass everywhere and they can be dangerous especially if children are about. The mirror width and distance from the polygon should be such that the reflection back from the from the stationary mirror should just fill the facet. Too long or too short will give incorrect operation. If you double the speed to 6000 RPM you only need half the number of stationary mirrors. Cheap mirrors are often not flat but you can get better mirrors if you proceed. Always be careful with laser beams.
So now we have a system to scan tv at low speed. What are the advantages. Uses cheap surplus parts, if you burn out a motor its not the end of the world because they are cheap, interesting DIY project, could be developed into a full scale laser projector. Disadvantages You will find it very difficult to position mirrors with accuracy. The slightest misadjustment will give poor results. You need to make a mirror table with adjustable legs for height adjustment. Same for laser and motor mounting. Then you need a system to adjust the mirrors. This will require considerable ingenuity. But it is possible and you can see adjustment/misadjustment on the screen. The most serious disadvantage is using a low power laser diode you will find the scanned output is very low in light. This is not surprising considering the low power. If you want a brighter scan you need more light probably 100 - 1000 times as powerfull. For this small scale experiment I just wanted to see if it is possible to scan at high speed using low motor speed and surplus components. I think the answer is yes it is possible to do the scanning and the next investigation will be how to get a lot of light for a small amount of money.
My email is remp@ps.gen.nz and I am always happy to answer questions but because I get so much mail answers have to be short. I cannot tell you where to get parts or how to wire up polygon motor controllers. Surplus parts supply changes often and motor connections best to get from your supplier.
This will be my last message for two or three weeks. I am starting to gather firewood for winter. It is supposed to go in the wood shed. The wood shed is full of several motor mowers, rotary hoe, a large machine I built to cut up tree prunings, engine parts, gearboxes, garden tools etc etc. These should be in the storage shed which was demolished when a tree fell on it. So I have to rebuild the shed.
The reason I started this thread was to show my experiments with a laser projector. The idea was to use inexpensive parts which can be obtained at small cost. I bought a surplus polygon, a 5 milliwatt 670 nm red laser diode and some front surface mirrors from OATLEY ELECTRONICS in Australia, and some new laser diodes 650 and 635 nm from the LASER GUY AND DEHARPPORTE TRADING Co in USA. Total cost less than 200 dollars NZ. Those of you who have followed the thread from the start will see the emphasis was on getting a fast scan. Television needs a very fast scan to move the beam from left to right across the screen. Using a polygon to do the scanning needs a very high speed around 36000 - 45000 RPM which needs a new polygon made to operate at that speed and they are very expensive. A way had to be found to use lower speed or the system became too expensive and out of reach for DIY people.
For PAL tv 625 interlace line system the tv lines are transmitted 312.5 at a time. This is a tv field. The first 312.5 lines are transmitted, then a second set of 312.5 lines are transmitted which fit in between the first set just like two combs fit together. These two fields make up a complete frame. The fields have to be drawn on screen in a 50th of a second. Looking at that I thought if I could get 312.5 lines in one motor revolution the motor only has to run at 50 times 60 = 3000 RPM which is well within the speed range of surplus polygon motors.
Many surplus polygons have 8 facets which in one revolution would only give 8 tv lines however by using a system of stationary mirrors surrounding the polygon it is possible to increase the number of lines in one motor revolution. This can be done by aiming the laser at the mirror facet, reflecting onto a stationary mirror, back from the stationary mirror to the polygon facet and out to a frame scanner. As the polygon rotates each stationary mirror gives 8 tv lines. Using 39 stationary mirrors we have 39 times 8 = 312 tv lines. You can test this idea at home using one stationary mirror by rotating a polygon by hand and see that each time a new polygon facet comes around it causes a new scan line. Mirror width should be 1/2 inch wide and 8 inches away from the polygon. These are approximate mesurements but will prove the point that one stationary mirror gives 8 tv lines, two stationary mirrors gives 16 tv lines etc. Ordinary mirrors work alright. Buy a glass cutter my one is a German one model Diamantor 200.1 $7.50 NZ. To cut mirrors buy cheap thin ones from the 2 dollar shop, lay the mirror on several peices of newspaper on a flat base, put a LOT of weight into the glass cutter and it will score first time then a slight tap and it will snap off. Be very careful to clean up after cutting with a vacuum cleaner otherwise you get little pieces of glass everywhere and they can be dangerous especially if children are about. The mirror width and distance from the polygon should be such that the reflection back from the from the stationary mirror should just fill the facet. Too long or too short will give incorrect operation. If you double the speed to 6000 RPM you only need half the number of stationary mirrors. Cheap mirrors are often not flat but you can get better mirrors if you proceed. Always be careful with laser beams.
So now we have a system to scan tv at low speed. What are the advantages. Uses cheap surplus parts, if you burn out a motor its not the end of the world because they are cheap, interesting DIY project, could be developed into a full scale laser projector. Disadvantages You will find it very difficult to position mirrors with accuracy. The slightest misadjustment will give poor results. You need to make a mirror table with adjustable legs for height adjustment. Same for laser and motor mounting. Then you need a system to adjust the mirrors. This will require considerable ingenuity. But it is possible and you can see adjustment/misadjustment on the screen. The most serious disadvantage is using a low power laser diode you will find the scanned output is very low in light. This is not surprising considering the low power. If you want a brighter scan you need more light probably 100 - 1000 times as powerfull. For this small scale experiment I just wanted to see if it is possible to scan at high speed using low motor speed and surplus components. I think the answer is yes it is possible to do the scanning and the next investigation will be how to get a lot of light for a small amount of money.
My email is remp@ps.gen.nz and I am always happy to answer questions but because I get so much mail answers have to be short. I cannot tell you where to get parts or how to wire up polygon motor controllers. Surplus parts supply changes often and motor connections best to get from your supplier.
Have you considered using Front Surface Mirrors?
How about a polygon made of a mirror polish metal?
As I understand it, front surface mirrors are better for optical use since you do not get loss from the light passing through the glass layer as well as some "ghosting". In the other thread I see mention that front surface mirrors can be made/bough fairly inexpensively.
What part of the Polygon is the weak point to faster rotation (what is the danger/risk)?
How about a polygon made of a mirror polish metal?
As I understand it, front surface mirrors are better for optical use since you do not get loss from the light passing through the glass layer as well as some "ghosting". In the other thread I see mention that front surface mirrors can be made/bough fairly inexpensively.
What part of the Polygon is the weak point to faster rotation (what is the danger/risk)?
Dragonhalf,
10,000 dollars for a DLP starter kit plus 5000 dollars for a controller pack is somewhat outside most DIY budget. At least it is mine.
Yes I use front surface mirrors often. They have the reflective coating on the front surface so radiation does not have to pass through the glass, reflect from the back surface and travel back through the glass again. They are a bit hard to find but many surplus shops have front surface or "first surface" mirrors available at reasonable cost.
Regarding strain on polygons at speeds up to 20,000 RPM there is normally no problem. As speed goes up centrifugal force starts to act on the polygon and initially the facet expands slightly until at extreme speed it can burst. No commercial polygons have this problem because they are well engineered for safety. It could be a problem for home constructors perhaps building a polygon wheel without knowledge of safety constraints. I believe that would be very unlikely because no home constructor would have the diamond turning equipment needed to finish a polygon facet to useful accuracy.
10,000 dollars for a DLP starter kit plus 5000 dollars for a controller pack is somewhat outside most DIY budget. At least it is mine.
Yes I use front surface mirrors often. They have the reflective coating on the front surface so radiation does not have to pass through the glass, reflect from the back surface and travel back through the glass again. They are a bit hard to find but many surplus shops have front surface or "first surface" mirrors available at reasonable cost.
Regarding strain on polygons at speeds up to 20,000 RPM there is normally no problem. As speed goes up centrifugal force starts to act on the polygon and initially the facet expands slightly until at extreme speed it can burst. No commercial polygons have this problem because they are well engineered for safety. It could be a problem for home constructors perhaps building a polygon wheel without knowledge of safety constraints. I believe that would be very unlikely because no home constructor would have the diamond turning equipment needed to finish a polygon facet to useful accuracy.
Greetings,
Last time I wrote for this thread I said I was getting some wood for winter and rebuilding the shed and a lot of other jobs round the house. I also said I would look into how to get a lot of light for not too much money. Well I have been doing that. The obvious question is since this is a laser projector thread will I be using lasers. No fraid not. At present it is simply too expensive to buy either new or second hand a red a green and a blue laser of several watts power. You would need to spend in the region of 5-10 thousand dollars for lasers at the very minimum plus you would have to rewire your house for three phase power pay the 1000 dollars or so power bill every month and lay in a quality water cooling system to keep the lasers cool. For a small diy project this seems out of scale. No doubt with rapid advances in technology new lasers for green and blue will come on the market as red semiconductor laser diodes have but could be 10 years away. So are there alternatives to lasers for a laser projector. What is needed is a source of light similar to a laser. Very bright, very small source size able to give red green and blue and affordable. My research has until recently come to a sticky end. It seems that suitable point source lights such as xenon short arcs are still very expensive. Metal halide short arc lights are available in the 100-200 dollar range but these are not really short arcs they are 5-10mm arcs. There are one or two very short arc about 1mm MH lights I found starting at $500 but only 25 watts power. We need something like 500 - 1000 watts power because if you take a 1000 watt light, extract the red green and blue components process the beams into even illumination run it through a modulator you will be very lucky to have 1 watt of optical energy left to power the projector. A question that is often asked is "why do we need an expensive point source light".
It has to do with turning the light into a useful form. An ordinary light bulb radiates light in all directions. To gather that the light into say a beam of light needs a reflector of some sort. The most common reflector that people see are reflectors in torches and domestic lights. The reflector takes some of the overall radiation and directs the light into a beam either a tight beam as with a torch or a less tight beam as with domestic lighting. That is quite satisfactor for those purposes but suppose you wanted an extremely bright torch with a very tight beam. Reflectors behind the bulb can be manufactured to better tolerances but you will start to run into the real problem that with a light source with a large size source such as a tungsten filament there is no further improvement in light gathering capacity until the size of the light source is reduced. Not just a small reduction either, you have to start thinking of your light source as having a source size of only a few millimeters. Very small. That is too small for tungsten filaments to produce any real power so you have to go to an arc light where the light is produced by the discharge of electricity between two electrodes. This problem of making the light source smaller to make it brighter is compounded by several other matters. Arc lights need controllers you cannot just plug them into the mains power. Short arc lights often contain a mixture of gases at high pressure and can be regarded as highly dangerous with the possibility of explosion. In addition even short arc lights do not give all their energy to a beam because of the lenght of the arc. Reflectors are based on a mathematical shape which assumes the light source is what is called a point source no larger than the point on a pin. At present it is not possible to make lights that small except lasers have a very small source size. The final problem is an extremely good reflector (expensive) with a very small light source (expensive) will normally only gather at most half the light anyway and the rest has to be wasted. I found all this information very depressing. I cannot afford to buy and run a laser system and it seems I cannot afford to buy the quality of short arc light to give simulation of laser light with any chance of success
It seemed the reflectors were a big stumbling block. For good light gathering you need a good reflector. A good reflector demands a point source light. Its a circular argument with no way out. For the DIY person the only options seemed to be pay a very high price for lasers or pay a high price for point source arc lights. Either way you are still way over a 1000 dollars and you still have to modulate the beams which will not be cheap.
There are plenty of cheap light sources that are long. I can buy a 500 watt 4 inch long halogen lamp for $6. That is very cheap and there are lots of those sorts of lights for sale. The problem is how to get the light from a long light into a usable form. We have already said that it is not possible, you need to have a short arc lamp but I found a type of reflector not much different from ordinary reflectors which can take the light from a long light and turn it into a point source. This remarkable reflector is called an ORTHOGONAL PARABOLIC REFLECTOR. It is similar in shape to an ordinary parabolic reflector but acts like an elliptical reflector where all the light gathered ends up at one point. The person who developed this reflector also claims that at the focal point the light behaves as if it came from a light source of infinitely small size. Apparently it makes no difference if the long light source is fat or thin. This would seem to be a remarkable breakthough in that one can take an inexpensive long light source and turn it into an extremely small point source. If anyone wants to see the particulars of this reflector go to the US patent database and look at #5037191. As always I mention you cannot use patent information for comercial purposes without discusion with the patent holder.
I will have a short break here now because I have bought an lcd panel and am trying to get it going. Hope to be back in a month or so.
Last time I wrote for this thread I said I was getting some wood for winter and rebuilding the shed and a lot of other jobs round the house. I also said I would look into how to get a lot of light for not too much money. Well I have been doing that. The obvious question is since this is a laser projector thread will I be using lasers. No fraid not. At present it is simply too expensive to buy either new or second hand a red a green and a blue laser of several watts power. You would need to spend in the region of 5-10 thousand dollars for lasers at the very minimum plus you would have to rewire your house for three phase power pay the 1000 dollars or so power bill every month and lay in a quality water cooling system to keep the lasers cool. For a small diy project this seems out of scale. No doubt with rapid advances in technology new lasers for green and blue will come on the market as red semiconductor laser diodes have but could be 10 years away. So are there alternatives to lasers for a laser projector. What is needed is a source of light similar to a laser. Very bright, very small source size able to give red green and blue and affordable. My research has until recently come to a sticky end. It seems that suitable point source lights such as xenon short arcs are still very expensive. Metal halide short arc lights are available in the 100-200 dollar range but these are not really short arcs they are 5-10mm arcs. There are one or two very short arc about 1mm MH lights I found starting at $500 but only 25 watts power. We need something like 500 - 1000 watts power because if you take a 1000 watt light, extract the red green and blue components process the beams into even illumination run it through a modulator you will be very lucky to have 1 watt of optical energy left to power the projector. A question that is often asked is "why do we need an expensive point source light".
It has to do with turning the light into a useful form. An ordinary light bulb radiates light in all directions. To gather that the light into say a beam of light needs a reflector of some sort. The most common reflector that people see are reflectors in torches and domestic lights. The reflector takes some of the overall radiation and directs the light into a beam either a tight beam as with a torch or a less tight beam as with domestic lighting. That is quite satisfactor for those purposes but suppose you wanted an extremely bright torch with a very tight beam. Reflectors behind the bulb can be manufactured to better tolerances but you will start to run into the real problem that with a light source with a large size source such as a tungsten filament there is no further improvement in light gathering capacity until the size of the light source is reduced. Not just a small reduction either, you have to start thinking of your light source as having a source size of only a few millimeters. Very small. That is too small for tungsten filaments to produce any real power so you have to go to an arc light where the light is produced by the discharge of electricity between two electrodes. This problem of making the light source smaller to make it brighter is compounded by several other matters. Arc lights need controllers you cannot just plug them into the mains power. Short arc lights often contain a mixture of gases at high pressure and can be regarded as highly dangerous with the possibility of explosion. In addition even short arc lights do not give all their energy to a beam because of the lenght of the arc. Reflectors are based on a mathematical shape which assumes the light source is what is called a point source no larger than the point on a pin. At present it is not possible to make lights that small except lasers have a very small source size. The final problem is an extremely good reflector (expensive) with a very small light source (expensive) will normally only gather at most half the light anyway and the rest has to be wasted. I found all this information very depressing. I cannot afford to buy and run a laser system and it seems I cannot afford to buy the quality of short arc light to give simulation of laser light with any chance of success
It seemed the reflectors were a big stumbling block. For good light gathering you need a good reflector. A good reflector demands a point source light. Its a circular argument with no way out. For the DIY person the only options seemed to be pay a very high price for lasers or pay a high price for point source arc lights. Either way you are still way over a 1000 dollars and you still have to modulate the beams which will not be cheap.
There are plenty of cheap light sources that are long. I can buy a 500 watt 4 inch long halogen lamp for $6. That is very cheap and there are lots of those sorts of lights for sale. The problem is how to get the light from a long light into a usable form. We have already said that it is not possible, you need to have a short arc lamp but I found a type of reflector not much different from ordinary reflectors which can take the light from a long light and turn it into a point source. This remarkable reflector is called an ORTHOGONAL PARABOLIC REFLECTOR. It is similar in shape to an ordinary parabolic reflector but acts like an elliptical reflector where all the light gathered ends up at one point. The person who developed this reflector also claims that at the focal point the light behaves as if it came from a light source of infinitely small size. Apparently it makes no difference if the long light source is fat or thin. This would seem to be a remarkable breakthough in that one can take an inexpensive long light source and turn it into an extremely small point source. If anyone wants to see the particulars of this reflector go to the US patent database and look at #5037191. As always I mention you cannot use patent information for comercial purposes without discusion with the patent holder.
I will have a short break here now because I have bought an lcd panel and am trying to get it going. Hope to be back in a month or so.
question for REMP
Hi there, I have just been having a quick browse of the info given with the aspect of different ways of doing this project. I note that you are using spinning mirrors to scan the beam into an image. How is it working do you have trouble with locking the rotational speed with the incomming video freq, thus stopping the picture from rolling. I see that Fiat1 was looking into moving/vaibrating mirror mounts, this I am more in favour with. I was starting to play with two voice coils with a mirror on each to do the vertical and horizontal scanning these are driven off an old TV chassis. But alas trying to find time between work and house rebuilds is hard.
Have many others had much luck with the moving/viabrating mirror design?
cheers.
Hi there, I have just been having a quick browse of the info given with the aspect of different ways of doing this project. I note that you are using spinning mirrors to scan the beam into an image. How is it working do you have trouble with locking the rotational speed with the incomming video freq, thus stopping the picture from rolling. I see that Fiat1 was looking into moving/vaibrating mirror mounts, this I am more in favour with. I was starting to play with two voice coils with a mirror on each to do the vertical and horizontal scanning these are driven off an old TV chassis. But alas trying to find time between work and house rebuilds is hard.
Have many others had much luck with the moving/viabrating mirror design?
cheers.
Too much mirror/suspension mass to achieve tv H scan speed. Too little movement to scan a decent size picture. Always have non linear scan whatever amplitude because mirror/suspension always has to slow down,stop and re-accelerate opposite direction. Best results I have heard of 5KHz. If could solve these problems it would be way to go because only one mirror signature to worry about but so far no one has offered working unit at full tv H scan rate. Problem gets worse because USA going to HDTV with higher scan rate. DLP is best working example of moving mirror system not cheap or available for DIY. A company called Printworks claims a very high speed moving mirror scanner which is used for scanning documents. Its on the net. If you believe in it, have a go. Someone will crack it.
Mountain
Sorry did not answer question re motor stability. No problem. Motor presently driven by frequency gen. If frequency stable, motor stable due to rotating mass. My system is present free running and sits there 5 minutes without generator readjustment. Speed can only be refreshed during frame interval since one field scanned per motor rev. Seems stable. Speed is 3000 RPM which is H sync/312.5. Oatley electronics in Australia used to have plenty polygons but seem to have dropped out of that.
Sorry did not answer question re motor stability. No problem. Motor presently driven by frequency gen. If frequency stable, motor stable due to rotating mass. My system is present free running and sits there 5 minutes without generator readjustment. Speed can only be refreshed during frame interval since one field scanned per motor rev. Seems stable. Speed is 3000 RPM which is H sync/312.5. Oatley electronics in Australia used to have plenty polygons but seem to have dropped out of that.
Hi, I was just curious about the vibrating mirror method. How big would the mirror be? How massive would it be? Would there be any way to reduce the mass? Could there be some sort of way to use a powerful driver to increase the throw and/or speed? Sorry for the questions, this idea is fairly foreign to me, but it's more interesting than LCD + OHP.
piGuy,
i suppose you mean resonant scanner systems. Material and size depend of scanning frequency. The higher the frequency, the smaller and lighter the mass of mirror. DMDs are an example of micro mirros, which can moved up to 100000 times a second. For laser resonant scanners, as far as i can see, the frequency limit seems to be at about 20khz.
http://www.eopc.com/scanners.html
xblocker
i suppose you mean resonant scanner systems. Material and size depend of scanning frequency. The higher the frequency, the smaller and lighter the mass of mirror. DMDs are an example of micro mirros, which can moved up to 100000 times a second. For laser resonant scanners, as far as i can see, the frequency limit seems to be at about 20khz.
http://www.eopc.com/scanners.html
xblocker
Hmm. I was thinking of some sort of setup using linear displacement of the mirror as opposed to angular displacement like in the link you posted. At least that's how I thought mountain_nz's setup was. I know he said he used a voice coil from a speaker or something. Correct me if I'm wrong, but wouldn't a mirror attached to a voice coil produce a linear displacement? I was thinking that if the light hit the mirror at 45 degrees, and the mirror was oscillating in a single linear axis, the reflected light would deflect at 90 degrees and move along the axis of the mirror. With two mirrors, you would be able to produce any position in a 2 dimensional system. I guess it would still be hard to produce any kind of displacement > 1mm at ~30kHz though. I hope I'm not sounding too stupid, I'm just trying to understand the basics of a scanning beam display. Anyone have any links that might explain this stuff better?
Also, on the subject of lasers: I was reading up on dye lasers, which sound pretty cool. I was wondering if there might be some way to utilize a dye laser. If it would be possible to use a pulsed light to produce the display (i.e. one pulse per pixel) instead of a continuous scanning beam, perhaps there might be some way to use a dye laser to produce the light. I don't know enough about them yet, but I imagine the main problems would be rapid decay of the dye, providing a powerful enough light source for the laser, and modulating the light at a high enough frequency. Okay, maybe it's not such a great idea. It would take care of the need for red, green, and blue laser light though.
Also, on the subject of lasers: I was reading up on dye lasers, which sound pretty cool. I was wondering if there might be some way to utilize a dye laser. If it would be possible to use a pulsed light to produce the display (i.e. one pulse per pixel) instead of a continuous scanning beam, perhaps there might be some way to use a dye laser to produce the light. I don't know enough about them yet, but I imagine the main problems would be rapid decay of the dye, providing a powerful enough light source for the laser, and modulating the light at a high enough frequency. Okay, maybe it's not such a great idea. It would take care of the need for red, green, and blue laser light though.
piGuy,
in theory, the basics of imaging systems is very simple. Throw a lightinformation of a video signal point after point onto a room coordinate in a given time. Practically it isn't that simple as it sounds. It could be done:
1. with a single laser, an X/Y scanner, giving every lightpoint his room destination. Problems: scan speed of horizontal sync can't' work with conventional moving coils, due to inertia of material.
2. with a horizontal laser array, scanning only Y-values. Advantage:
vertical sync 50hz can be easily done with galvanic or other scanners. Disadvantage: for every hor. pixel a laser (led) is needed.
Look at this thread. It's an example of problems that occur!
xblocker
in theory, the basics of imaging systems is very simple. Throw a lightinformation of a video signal point after point onto a room coordinate in a given time. Practically it isn't that simple as it sounds. It could be done:
1. with a single laser, an X/Y scanner, giving every lightpoint his room destination. Problems: scan speed of horizontal sync can't' work with conventional moving coils, due to inertia of material.
2. with a horizontal laser array, scanning only Y-values. Advantage:
vertical sync 50hz can be easily done with galvanic or other scanners. Disadvantage: for every hor. pixel a laser (led) is needed.
Look at this thread. It's an example of problems that occur!
xblocker
Piguy
There are two types of mirror scanners (usually called galvanometers) One type is a driven mirror assembly where the driving motor controls the mirror position more or less exactly at all points of the scan similar to a stepping motor and these have a linear scan from left to right then a quick return to start position, or the mirror can be positioned at any required location and held there. These rarely go past 5khz scan speed so are useless for TV scanning but used all the time in laser light shows and work very well. The mirror unit and driving motor and driving amplifier assemblies are not all that cheap starting at US 500 for medium quality and over US 1500 for top quality units. Then there is a completely different type of unit which is almost self oscillating and only requires a small electric drive at the start to get it going. These units are not controllable during the scan and rely on very light weight stiff but light suspension system and have several problems you need to think about if using. For a start because of the mass of the mirror/suspension assy its hard to push the speed much past 5khz and those that quote speeds like 16khz have a very small beam angle. You can see this with the link Xblocker posted and the model SC30 at its highest scan rate only gives a beam angle of 5 degrees. You can externally increase deflection angle by a mirror tunnel as was used by people using very fast small angle piezo-electric mirror actuators. They increase their apparent speed by displaying a scan in the left/right movement of the mirror and also in the right to left movement of the mirror which means you need a memory store for your scan and output left to right for one tv line and right to left for the next tv line. Because the mirror/suspension mass has to slow down/stop/reaccelerate at end of mirror movement you get a type of scan similar to a sine wave which means you need to electronically deliver your pixel output to the mirror which matches the sine wave deflection otherwise your picture will be grossly distorted. These units usually need a vacuum chamber and a controlled temperature to operate correctly and start at US 2500 and upwards. Many people I have discussed this with started out thinking if you have a very very small mirror and apply a very strong energising force you must be able to get high speed mirror scan but this is not so because the mirror suspension also has to be small and lightweight because it moves as well and does not have the strength to convert a powerful moving force from a driver to the mirror and inertia of the assy limits how quickly it will move. The idea is to use a very small light weight mirror not much larger than your laser beam diameter, a very strong lightweight mirror suspension and stator unit then apply strong energising to that. Not easy but you come up with a solution that works and you won't have to worry about working the rest of your life. I wrote to one company who claimed they were working on a virtual zero mass mirror generated by hologram. They did reply but said their research results were not available.
Search for galvanometer. This will show a large number of pages many detailing how they are made, what performance you can get and an idea of prices.
There are two types of mirror scanners (usually called galvanometers) One type is a driven mirror assembly where the driving motor controls the mirror position more or less exactly at all points of the scan similar to a stepping motor and these have a linear scan from left to right then a quick return to start position, or the mirror can be positioned at any required location and held there. These rarely go past 5khz scan speed so are useless for TV scanning but used all the time in laser light shows and work very well. The mirror unit and driving motor and driving amplifier assemblies are not all that cheap starting at US 500 for medium quality and over US 1500 for top quality units. Then there is a completely different type of unit which is almost self oscillating and only requires a small electric drive at the start to get it going. These units are not controllable during the scan and rely on very light weight stiff but light suspension system and have several problems you need to think about if using. For a start because of the mass of the mirror/suspension assy its hard to push the speed much past 5khz and those that quote speeds like 16khz have a very small beam angle. You can see this with the link Xblocker posted and the model SC30 at its highest scan rate only gives a beam angle of 5 degrees. You can externally increase deflection angle by a mirror tunnel as was used by people using very fast small angle piezo-electric mirror actuators. They increase their apparent speed by displaying a scan in the left/right movement of the mirror and also in the right to left movement of the mirror which means you need a memory store for your scan and output left to right for one tv line and right to left for the next tv line. Because the mirror/suspension mass has to slow down/stop/reaccelerate at end of mirror movement you get a type of scan similar to a sine wave which means you need to electronically deliver your pixel output to the mirror which matches the sine wave deflection otherwise your picture will be grossly distorted. These units usually need a vacuum chamber and a controlled temperature to operate correctly and start at US 2500 and upwards. Many people I have discussed this with started out thinking if you have a very very small mirror and apply a very strong energising force you must be able to get high speed mirror scan but this is not so because the mirror suspension also has to be small and lightweight because it moves as well and does not have the strength to convert a powerful moving force from a driver to the mirror and inertia of the assy limits how quickly it will move. The idea is to use a very small light weight mirror not much larger than your laser beam diameter, a very strong lightweight mirror suspension and stator unit then apply strong energising to that. Not easy but you come up with a solution that works and you won't have to worry about working the rest of your life. I wrote to one company who claimed they were working on a virtual zero mass mirror generated by hologram. They did reply but said their research results were not available.
Search for galvanometer. This will show a large number of pages many detailing how they are made, what performance you can get and an idea of prices.
Piguy
We all face difficulties trying to get a project going. The idea is to increase your knowledge base. Study as much as you can on the subject . We cannot give you more encouraging information just like that because it is not an easy problem. Hundreds of people have lodged patents for scanning TV rasters with moving mirrors but none have been succesfully marketed except for Texas instruments with their DLP moving mirror system and they took twenty years from first patent to a working unit on the market. All we can do is tell you what we know and try and point you in the right direction to gain more knowledge. I always try to give as much information as I possibly can especially DIY people having a go at a problem. Is there anything specifically we can help you with. I will do the best to answer.
We all face difficulties trying to get a project going. The idea is to increase your knowledge base. Study as much as you can on the subject . We cannot give you more encouraging information just like that because it is not an easy problem. Hundreds of people have lodged patents for scanning TV rasters with moving mirrors but none have been succesfully marketed except for Texas instruments with their DLP moving mirror system and they took twenty years from first patent to a working unit on the market. All we can do is tell you what we know and try and point you in the right direction to gain more knowledge. I always try to give as much information as I possibly can especially DIY people having a go at a problem. Is there anything specifically we can help you with. I will do the best to answer.
- Status
- Not open for further replies.