AVS › AVS Forum › Display Devices › Screens › DIY Screen Section › DIY Vertical Masking Design......Need Design Advice
New Posts  All Forums:Forum Nav:

DIY Vertical Masking Design......Need Design Advice

post #1 of 42
1/14/10 at 9:23am
Thread Starter 
I've posted this in the Screens forum, but it started out more as a general using curtains as a mask thing and then led into a DIY masking system.

Here's the link: http://www.avsforum.com/avs-vb/showthread.php?t=1215344

I'll reattach the sketch (shown at the bottom).

One of my biggest issues is not knowing where to go to get the parts, like motor, encoders, etc.. I know where to get discrete electronic parts. I am designing the circuitry for it now, but it's possible that it is already available as a complete circuit. For instance, can you buy an integrated motor, encoder and processing circuitry to take an IR command and set preset positions for differenent aspect ratios. I have an idea for the circuitry if I have to go the discrete route.
As far as IR or even RS232 go, I'm pretty clueless as to what is available or how to integrate it.

I would appreciate any guidance that you can give. And if you can give me some good sources for the components that would be great.

Thanks,
Greg

 

Screen Masking Idea.pdf 15.1396484375k . file
Reply
Reply

AVS Top Picks

post #2 of 42
1/16/10 at 8:29am
Quote:
Originally Posted by Gregory View Post

... For instance, can you buy an integrated motor, encoder and processing circuitry to take an IR command and set preset positions for differenent aspect ratios. ...

Greg,
Finding an off-the-shelf integrated motor and control unit has been one of the primary stopping blocks for many of us who have wanted to attempt DIY masking systems. Over in my thread on masking, discussions on this very topic went on for months.

I had thought I found a solution using rotator units, but even this was not an elegant solution since the motor was not powerful enough. Then on top of that, the manufacturer for the rotator I was using stopped making them.

Probably the best option that I am aware of right now are the motors and control systems by Somfy. They will do exactly what you are asking for, but they are pretty pricey.

You might try to contact an AVSforum user named "Shady_". See THIS POST that he posted in my thread. He seems to know a lot about the Somfy product line and offered to be of assistance.

- Scott
Reply
Reply
post #3 of 42
1/16/10 at 10:04am
Thread Starter 
Scott,

Thanks for the quick reply and the advice on the Somfy motor.

Do you know how much torque, horsepower, etc. that you need? In my case I think the power that is need is what is already provided by a power drapery motor.

If I can't find something that is affordable I was thinking of just getting a motor and adding my own decoder and circuitry to it. I just need to find the right motor and then design the decoder circuitry for it. It will probably be an optical decoder............I have a basic design that I think will work. If interested, I will post the details after I iron it out a little more.

Thanks,
Greg
Reply
Reply
post #4 of 42
1/16/10 at 9:25pm
HI Gregg,

ScottJ007's thread has a vast amount of info from many posters regarding the same goal. A former AVS member Don (HTIQ company once produced multi pre-set rmt ctl masking systems) once posted on here that he thought a torque value of 175 oz-in was good (that may depend on how smoothly your system works).

Below is a link to company that offers 3-12Vdc gearmotors with a wide range of torque values, but no encoder options.
http://servocity.com/html/3-12v_gear_motors.html

Somewhere I have a link to another robotics web vender that offers a couple of dual shaft DC geamotors that do have optional encoders, but I think their torque values were a bit low. Will post back when I find that info.

John

Edit: Found it: Lynxmotion has a couple of motors that have a rear shaft on which a optional USDigital encoder may be attached. The rpm's may be too high on these (One poster an RPM of 40):

http://www.lynxmotion.com/Search.aspx?txtSearch=encoder

Another vendor has 3 dc gearmotors with back shaft (two are Lynxmotion models), but all too fast in rpms:

http://www.robotshop.us/gear-motors-...k-shaft-2.html
Reply
Reply
post #5 of 42
1/17/10 at 3:32pm
I ran across a bunch of motors used to operate windows. See attached picture. The motor output shaft has a hole that I passed a piece of 1/2" all-tread rod through. I used copper pipe with jam nuts to transfer torque to the all-thread. Also used copper pipe to roll the blackout fabric on.

Here is info on the whole motor system. I just used the motor, not any of the other stuff.
LL
LL
Reply
Reply
post #6 of 42
1/17/10 at 3:32pm
Thread Starter 
Hi John,

Thanks for those links..............that is very helpful.

Could gear those high RPM motors down, I suppose. If I can't find the right combination of motor and encoder I could always attempt to make my own encoder.

The robotshop had a lot of motor controllers, so at least I should be able to find one that will work and won't have to design that part of it. I came up with a basic design (up/down pulse counter with a jog/fine tune function) and now have to determine how I will interface it with the motor controller.

Thanks again for the help.

Greg
Reply
Reply
post #7 of 42
1/17/10 at 3:42pm
Thread Starter 
Quote:
Originally Posted by petew View Post

I ran across a bunch of motors used to operate windows. See attached picture. The motor output shaft has a hole that I passed a piece of 1/2" all-tread rod through. I used copper pipe with jam nuts to transfer torque to the all-thread. Also used copper pipe to roll the blackout fabric on.

Here is info on the whole motor system. I just used the motor, not any of the other stuff.

Thanks for the advice. Interesting adaptation of that type of motor.

Do you use any type of control system or is it just fully extended and fully retracted?

Greg
Reply
Reply
post #8 of 42
1/17/10 at 4:14pm
I'm just using a DPDT center off toggle switch for now. Plan to rig some type of control system with limit switches at some point. The motor is 24VDC.

I've got a bunch more motors and am planning to build another set of blinds and an automated masking that slides horizontally. Still working out details on that.
Reply
Reply
post #9 of 42
1/17/10 at 9:35pm
Greg,

Wow had alot written & lost it all. Will go for a shorter post this time.
Here is a link to a motorized curtain track from PowerCurtain that offers IR & wall mt ctls. It only offers full open & full close with stop at any position in between, but no presets.
This site has very good photos of whats included:

http://www.powercurtain.com/product1_800.html
Reply
Reply
post #10 of 42
1/17/10 at 9:41pm
Thread Starter 
I hate when that happens!

Thanks for the link. I really want presets for various aspect ratios.

Greg
Reply
Reply
post #11 of 42
1/18/10 at 12:51am
Thread Starter 
I've been working on a schematic for a control system. To finalize it I'll have to pick a motor control circuit, unless I design that too...........but, I really don't want too. The robotshop link that John posted has a good variety. The one issue that I need to consider is when the motor hits it target stop point..........it may oscillate about that point. I'll be using a comparator to compare the actual position (pulse count) to the target position (pulse count). When the values are equal the motor will stop. But, if the motor overshoots then it will back-up and it may get into this hunting condition. I have a couple of ideas, but really need to think it out better.

I still need to add the jog/fine tune circuit. Basically, for the jog/fine tuning circuit I add another set of counters with a astable multivibrator that is switched in when I want to tweak the aspect ratio slightly. I preload the counter with the aspect ratio that I want and the motor positions the masks for that particular aspect ratio. If I need to jog the mask left or right because the movie is not centered I then switch in the astable multivibrator, which will provide a pulse train to the counter. The counter will either add or subtract these pulses from the target pulse count, giving me a new target. The motor now repositions for this new tweaked (jogged) aspect ratio.

The real fun part of this will be getting it to respond to IR commans for the various aspect ratios and jogging left or right.

As soon as I get a pdf of the circuit I'll post it for comments.

This could be a fun project.

Greg
Reply
Reply
post #12 of 42
1/18/10 at 7:19am
This sounds like a project that I would attempt that spins out of control (as in complexity!)

A few questions. What type of encoder are you using? There are two kinds, absolute and incremental. Absolute always know where they are on power up. Incremental need to move around until an "index" or "home" pulse is found so it knows where it is.

Most incremental encoders are two digital pulses offset by 90 degrees. Therefore you need a counting circuit which figures out positive and negative counts. It's a basic state machine. Absolute encoders are generally serial and have some communication protocol.

What type of circuit were you thinking to decode this? Were you planning on using a microprocessor?

As long as there is enough friction in the system, I really don't think you'll need a control servo. I think the overshoot would be very minimal. Worst case if you have a microprocessor you can implement a PID servo. You can research it but its fairly easy to implement. Take the control error (actual position - set point) and multiply by a fixed gain constant. You also integrate the error and multiply that by a fixed integral gain constant. You probbaly won't need the derivative portion of the PID. Again, I doubt you'll need a servo in the first place.

Instead I would use a DC motor with a H-bridge driver. Then use reed limit switchs that detect the stop positions. If you want, you could come up with some mechanical sliding mechanism to change the reed switch position.

I assume your using a 12V motor. You can use a H-bridge with embedded controller such as an Infineon TLE5206 for $7.57. This H-bridge can short the positive and negative leads of the motor together which makes a great motor brake. It will stop almost instantly which is why I don't think you need a servo.

I have a Jetski and for the trim to make the jet go up and down, they use a 12V windshield wiper motor. It is geared down and turns maybe a few times a second. I'm sure you could get those on ebay or your local junkyard. Then connect this to a lead screw (even a long machine thread lead screw you can buy at Home Depot) that' running along the top of the screen that hangs your mask.

Now if you could make the mechanics small and cheap enough, I might give it a go!
Reply
Reply
post #13 of 42
1/18/10 at 11:46am
Thread Starter 
Quote:
Originally Posted by Pyro2 View Post

This sounds like a project that I would attempt that spins out of control (as in complexity!)

A few questions. What type of encoder are you using? There are two kinds, absolute and incremental. Absolute always know where they are on power up. Incremental need to move around until an "index" or "home" pulse is found so it knows where it is.

Most incremental encoders are two digital pulses offset by 90 degrees. Therefore you need a counting circuit which figures out positive and negative counts. It's a basic state machine. Absolute encoders are generally serial and have some communication protocol.

What type of circuit were you thinking to decode this? Were you planning on using a microprocessor?

As long as there is enough friction in the system, I really don't think you'll need a control servo. I think the overshoot would be very minimal. Worst case if you have a microprocessor you can implement a PID servo. You can research it but its fairly easy to implement. Take the control error (actual position - set point) and multiply by a fixed gain constant. You also integrate the error and multiply that by a fixed integral gain constant. You probbaly won't need the derivative portion of the PID. Again, I doubt you'll need a servo in the first place.

Instead I would use a DC motor with a H-bridge driver. Then use reed limit switchs that detect the stop positions. If you want, you could come up with some mechanical sliding mechanism to change the reed switch position.

I assume your using a 12V motor. You can use a H-bridge with embedded controller such as an Infineon TLE5206 for $7.57. This H-bridge can short the positive and negative leads of the motor together which makes a great motor brake. It will stop almost instantly which is why I don't think you need a servo.

I have a Jetski and for the trim to make the jet go up and down, they use a 12V windshield wiper motor. It is geared down and turns maybe a few times a second. I'm sure you could get those on ebay or your local junkyard. Then connect this to a lead screw (even a long machine thread lead screw you can buy at Home Depot) that' running along the top of the screen that hangs your mask.

Now if you could make the mechanics small and cheap enough, I might give it a go!

I'm not exactly sure what type of encoder it is............maybe from my description you will know or at least tell me where I'm going to have issues. I was planning on using a single pulse type encoder (not quadrature). I will need to provide a reset pulse that sets the system either, to it's home position (everything retracted) or to it's last position. I plan to always keep power on the circuitry, but in case of power failure or whatever, it needs to find it's way home. I still need to work this out.

The circuitry uses CMOS up/down counters to keep track of the pulse count. The target pulse counts that correspond to the various aspect ratios will be determined by trial and error. At a push of a button, say for a 1.78 movie, the corresponding pulse count is parallel loaded into a CMOS magnitude comparator. The comparator compares the actual pulse count from the encoder to the target pulse count and outputs a high on one of 3 output lines (AB, A=B) when the two pulse counts match. The counter has two clocks, an up clock and a down clock. The output lines will be routed to the counter clock lines via logic gates, which will logic AND the encoder output. This will determine which clock line to use. If I used a quadrature encoder I could probably eliminate this logic.

I just started researching motors and where I stand right now I would go with a DC motor with an H-bridge controller. I set-up the control logic to work with one of the motor controllers that I saw at the robotshop. When the two pulse counts match, the control logic puts the motor into high brake mode. Depending on the actual motor controller that I go with the logic may have to change.

So, going full speed and then putting the brakes on stops the motor nearly instantly? If so, that is great. I was thinking that I may have to redue the speed as I approached the target pulse count to more precisely stop the motor..........that would be more logic to decode that condition.

The reed limit switches that you suggest is another option that I thought about, but I thought that this would be a cool design project............even if I never actually build it. You can even use some type of optical detection method, break a beam or something similar. It would be a whole lot easier! Nah..........what am I thinking!

As I mentioned in the previous post preventing hunting will need to be addressed.

I haven't looked into micoprocessors yet, but it would probably simplify a lot of the circuitry. One uP vs a bunch of IC's.

I still need to work out the mechanics..........thought about it some, but need to give it more thought.

In summary, here is what I need to consider:

- uP vs discrete IC's and logic
- Determine if I will change to a quadrature encoder
- System initialization when the power comes up
- Method to stop hunting (some type of deadband)
- Settle on a motor and controller
- Mechanical design

So, given the above description where do you see problems? Yes..........this could spin out of control!

Thanks,
Greg
Reply
Reply
post #14 of 42
1/18/10 at 5:29pm
My two cents on Somfy motors: I realize that Somfy Motors can be expensive, but at the same time I also have personal experience with projects where I went my own route and ended up spending much more. I really get a lot of satisfaction in coming up with my own solutions, so I understand where you guys are going. With a selection of pulleys, cords, and takeup reels, you could probaby get away with a single motor operating both side masks in unison. And another for any vertical masking. Somfy's ILT Motor has encoder based positioning, 16 intermediate positions, RS-485/RS-232 w adapter/IR/RTS/Z-Wave. They have a 24VDC version out now, but it is nearly as expensive as the AC motor and pulls less weight (quiet though).

It just seems easier to stick a tubular motor in a tube, throw fabric on it, figure out a take-up reel on the end of the tube, and mount it. I'll say again though that I do enjoy cobbling my own creation together. I'll do my best to check back and give any input I can, let me know if you have any questions about lift capacities. Fabric weight in horizontal applications is factored in but the tension put on the screen to keep it tight when calculating the correct motor size is really important. E.g. Torque = Load x Radius so a motor in a 2 inch dia tube lifiting a 21 lb shade (or 21lbs of tension) would need a motor with 21 in/lbs of torque.

edit: I've also thought about using panel track for the side masks: http://www.ikea.com/us/en/catalog/products/80079363 this track has three rails so the panels could cover in succession (overlapping eachother). For those with less space, the roller concept might be better.
Reply
Reply
post #15 of 42
1/18/10 at 7:38pm
Is a vertical masking system that moves the mask up and down, or left and right? This will be much easier for left/right masks.

Here are some cheap motors: http://www.surplustraders.net/a/0184.shtml

In fact, look at the "Beuhler DC Gear Motor with linear screw assembly...". That is a lead screw that converts rotation to linear motion. That's what I envision.

You can get long 3/8 threaded rod (can even buy at Home Depot, Lowes) at 16 threads per inch. Run a 330RPM (5.5 rev/sec) motor on that and you can move 20.6 inch per min (0.34 inch / sec). How far do you need to move? 12inch? That's 35 seconds. You have such a mechanical advantage the motor does not need to be strong. In fact, if something gets snagged, it will act as a trash compactor. Run the whole thing off an old computer power supply.

I think you should abandon the encoder all together. It makes the electronics more complex than I think they need to be...especially if there is any plan to ever build this thing...

Instead use one magnet and reed switches. Reed switch is $0.75 each. Have them mechanically adjustable to the setpoints of where you want the mask to be. You could even just hot glue it to the frame. The electronics will keep a state counter so it "remembers" which state its at. When it gets a new command, it will know which way it has to move. It moves forever until it hits the next reed switch. No encoder needed
Reply
Reply
post #16 of 42
1/18/10 at 8:07pm
Thread Starter 
Quote:
Originally Posted by Shady_ View Post

My two cents on Somfy motors: I realize that Somfy Motors can be expensive, but at the same time I also have personal experience with projects where I went my own route and ended up spending much more. I really get a lot of satisfaction in coming up with my own solutions, so I understand where you guys are going. With a selection of pulleys, cords, and takeup reels, you could probaby get away with a single motor operating both side masks in unison. And another for any vertical masking. Somfy's ILT Motor has encoder based positioning, 16 intermediate positions, RS-485/RS-232 w adapter/IR/RTS/Z-Wave. They have a 24VDC version out now, but it is nearly as expensive as the AC motor and pulls less weight (quiet though).

It just seems easier to stick a tubular motor in a tube, throw fabric on it, figure out a take-up reel on the end of the tube, and mount it. I'll say again though that I do enjoy cobbling my own creation together. I'll do my best to check back and give any input I can, let me know if you have any questions about lift capacities. Fabric weight in horizontal applications is factored in but the tension put on the screen to keep it tight when calculating the correct motor size is really important. E.g. Torque = Load x Radius so a motor in a 2 inch dia tube lifiting a 21 lb shade (or 21lbs of tension) would need a motor with 21 in/lbs of torque.

edit: I've also thought about using panel track for the side masks: http://www.ikea.com/us/en/catalog/products/80079363 this track has three rails so the panels could cover in succession (overlapping eachother). For those with less space, the roller concept might be better.

Thanks for the advice on the Somfly motor.........I'll check it out. I know others have used them.

I think all I will be lifting (rolling) is a shade type of material to be used as a mask, so I doubt I'll need much torque. The other thing that I'm not sure of is the RPM.

The triple track won't work in my application because I need the mask to be completely hidden when not used and I do not have much room on the sides.

Depending on how much adjustability I want I'll need either, 3 motors or 4 motors.

- 3 motors: separate rods, one on each side with one motor each to roll down the side masks. A third motor to slide the masks and the curtain.

- 4 motors: separate rods, one on each side with one motor each to roll down the side masks. Separate motors on each side to independently slide each of the masks with the curtain. This would be used for independent mask adjustability for off center movies.

For the roll down masks I'll probably just need limit switches because it is either, fully retracted or fully extended.

For sliding the curtains and the roll down masks I'll need a control system.

Thanks,
Greg
Reply
Reply
post #17 of 42
1/18/10 at 9:10pm
Thread Starter 
Quote:
Originally Posted by Pyro2 View Post

Is a vertical masking system that moves the mask up and down, or left and right? This will be much easier for left/right masks.

Here are some cheap motors: http://www.surplustraders.net/a/0184.shtml

In fact, look at the "Beuhler DC Gear Motor with linear screw assembly...". That is a lead screw that converts rotation to linear motion. That's what I envision.

You can get long 3/8 threaded rod (can even buy at Home Depot, Lowes) at 16 threads per inch. Run a 330RPM (5.5 rev/sec) motor on that and you can move 20.6 inch per min (0.34 inch / sec). How far do you need to move? 12inch? That's 35 seconds. You have such a mechanical advantage the motor does not need to be strong. In fact, if something gets snagged, it will act as a trash compactor. Run the whole thing off an old computer power supply.

I think you should abandon the encoder all together. It makes the electronics more complex than I think they need to be...especially if there is any plan to ever build this thing...

Instead use one magnet and reed switches. Reed switch is $0.75 each. Have them mechanically adjustable to the setpoints of where you want the mask to be. You could even just hot glue it to the frame. The electronics will keep a state counter so it "remembers" which state its at. When it gets a new command, it will know which way it has to move. It moves forever until it hits the next reed switch. No encoder needed


The actual masks are similar to a roller blind........well that's exactly what it is........a black blind. It just needs to unroll fully (vertically), so no need for a complex control system.

The masks, when extended, will slide left and right to the desired aspect ratio, pulling the curtains with it or maybe it's a motorized drapery rod pulling the masks.

How far I need to move the masks will depend on the aspct ratio I want to mask. If I'm at the home position (2.35 AR) I will need to move each side about 25" for a 1.33 AR. That would be the furthest. Most of the time it will be an 11" move (2.35 to 1.78).

The reed switch approach is a great idea. The switches can be placed on a slide rail to quickly fine tune the position. Ok, I'll need to think about this a little.

Thanks,
Greg
Reply
Reply
post #18 of 42
1/19/10 at 10:12pm
Gregg,
Regarding reed switches; someone back on ScottJ007's thread mentioned that there may be a gap in where the mask would stop -- depends from which direction the mask is coming from. From the left it will stop X amount away from the reed switch & coming from the right the mask will trigger the reed swt in a different spot. The mag's field is wider than the mag itself, so the field will trip the swt in a diff place depending on the direction of approach. The width may not be large enough to matter maybe?
--------------------------
I had thought about placing movable (for adjustable AR pre-sets) slotted opto sensors along a track (wood working miter track?). A small tab mounted on the mask edge top would trip the swt as it passes thru the slot, of course alignment would be critical (slots are are only only a few mm wide). Moving the opto along the track would position pre-set points. Wiring less of a problem this way, no worries about sensor wiring on a moving mask. See link below for two premounted slotted opto sensors; OSU-69 & OSU-40.

http://www.allelectronics.com/make-a...tronics/1.html

Another possible sensor: Hall Effect sensor also could be mounted along an adjustable track (again maybe a mitre track). Electroncally I wonder how you could tell the MP which sensor was which pre-set? No idea...
------------------------------
On the motorized curtain track I mentioned in an earlier post; I wondered about using a tracked sensor setup & some how using them to give a "stop command" at each point when tripped. The system has a STOP button on the rmt ctl & wall swt, would sorta dulication the swt press. But I guess the system would be hampered by stopping at each preset coming from either direction. Just a thought.
--------------
Here's a animation I found on this forum, shows a dual roll down mask. This approach yields two possible ARs, good idea if you have no room on either side for curtain stack back, but only need two ARs.

http://home.earthlink.net/~sushinut/dropmask.gif

If one could mount a motorized curtain track behind the top assembly, a connection between the L & R master carriers could pull the two rollers in & out along the shaft. Another connecting rod between the rollers to the curtains pulls them along as well.
-----------------------------
Here is a very basic drawing (from the British Film Club) using a top mounted single motorized roller that moves two masks towards & away from each other. The cables are wound in opposite directions around the top roller, as one lowers, the other is pulled up equally. Of course the image would have to be centered between the masks for this to work... Have seen several versions of this done on this forum.
http://www.bfcc.biz/masking.jpg

John
Reply
Reply
post #19 of 42
1/19/10 at 10:31pm
Here are a couple of links to companies offering IR & RF remote control RX & TX boards. Maybe this will be of help:

Rentron: http://www.rentron.com/remote.htm


Tauntek: http://www.tauntek.com/
Reply
Reply
post #20 of 42
1/19/10 at 11:19pm
Here is a link to a "Kits-R-Us" 4-chnl RF kit from Carl's Electronics (can be opuchased for many web vendors):

http://www.electronickits.com/remote_control/rf4.htm

The kit can be purchased assembled as well. The RX board has 4 on board relays, each chnl has a reset circuit to wire in a limit switch (snap action swt) to reset each relay & each relay has a dip swt to choose momentary or locked function.

"Kits-R-Us" once made an 8-chnl IR combo, www.alltronics.com still has one left. They also make an
10 & 12 chnl versions, but no reset function on any.


John
Reply
Reply
post #21 of 42
1/19/10 at 11:22pm
Thread Starter 
Quote:
Originally Posted by HDGTX View Post
Gregg,
Regarding reed switches; someone back on ScottJ007's thread mentioned that there may be a gap in where the mask would stop -- depends from which direction the mask is coming from. From the left it will stop X amount away from the reed switch & coming from the right the mask will trigger the reed swt in a different spot. The mag's field is wider than the mag itself, so the field will trip the swt in a diff place depending on the direction of approach. The width may not be large enough to matter maybe?
--------------------------
I had thought about placing movable (for adjustable AR pre-sets) slotted opto sensors along a track (wood working miter track?). A small tab mounted on the mask edge top would trip the swt as it passes thru the slot, of course alignment would be critical (slots are are only only a few mm wide). Moving the opto along the track would position pre-set points. Wiring less of a problem this way, no worries about sensor wiring on a moving mask. See link below for two premounted slotted opto sensors; OSU-69 & OSU-40.

http://www.allelectronics.com/make-a...tronics/1.html

Another possible sensor: Hall Effect sensor also could be mounted along an adjustable track (again maybe a mitre track). Electroncally I wonder how you could tell the MP which sensor was which pre-set? No idea...
------------------------------
On the motorized curtain track I mentioned in an earlier post; I wondered about using a tracked sensor setup & some how using them to give a "stop command" at each point when tripped. The system has a STOP button on the rmt ctl & wall swt, would sorta dulication the swt press. But I guess the system would be hampered by stopping at each preset coming from either direction. Just a thought.
--------------
Here's a animation I found on this forum, shows a dual roll down mask. This approach yields two possible ARs, good idea if you have no room on either side for curtain stack back, but only need two ARs.

http://home.earthlink.net/~sushinut/dropmask.gif

If one could mount a motorized curtain track behind the top assembly, a connection between the L & R master carriers could pull the two rollers in & out along the shaft. Another connecting rod between the rollers to the curtains pulls them along as well.
-----------------------------
Here is a very basic drawing (from the British Film Club) using a top mounted single motorized roller that moves two masks towards & away from each other. The cables are wound in opposite directions around the top roller, as one lowers, the other is pulled up equally. Of course the image would have to be centered between the masks for this to work... Have seen several versions of this done on this forum.
http://www.bfcc.biz/masking.jpg

John
Quote:
Originally Posted by HDGTX View Post
Here are a couple of links to companies offering IR & RF remote control RX & TX boards. Maybe this will be of help:

Rentron: http://www.rentron.com/remote.htm


Tauntek: http://www.tauntek.com/
Hi John,

Thanks for the links to the IR stuff.

I thought about that magnetic field issue as a possible problem. I think optical sensors (either, reflective or active) gives you more precise stop points. There may be ways to mitigate the magnetc field issues, but I understand optical better.........I can see it, unless it's infrared.

Miter track............you are talking about the T-slot track, I assume? This way you can lock your sensors down if you can mount them to a T-bolt or say a piece of wood or plastic that is held in place with the T-bolt. Good idea!

Per Pyro2's suggestion I am looking into an alternate (non-encoder) method, which may be simpler to implement. However, I may or may not use reed switches. The method would not be that much different between magnetic and optical, except for maybe the interface to the sensors/switches.

That being said, I have not totally abandoned the encoder method, either...........mainly because it's fun to design.

Attached below, is a rough schematic of what I envision.

The area that says Aspect Ratio Target Value still needs some work...........I didn't have enough room to show it all. Basically, I will parallel load the target aspect ratio pulse count into a counter (not shown). This will in turn be fed into the magnitude comparator (the unconnected lines hanging down). The reason for the additional counter is that I can fine tune or jog the aspect ratio by having a switch connected to a pulse generator that offsets the value loaded into the counter.

Let me know what you think.

Thanks,
Greg

 

Motor Control Circuit.pdf 12.017578125k . file
Reply
Reply
post #22 of 42
1/20/10 at 4:11pm
Gregory,
I used a Picaxe RK18Motor microcontroller board w/h-bridge to control a top/bottom mask system. I used the board without modification. Position feedback was from an optical encoder (modified PC-mouse). The Picaxe is a rather slow controller, but it has a lot of built in functions that is useful in this application. It has built in IR-deocding (Sony codes), PWM-control, analog and digital I/O, eeprom etc..

With this simple board I got 12 posistion presets that were programmable from the IR-remote (jog up/down to desired position, press save button then desired preset button). Presets were stored in eeprom, so they were not lost when system was powered off.

I wanted a 4-way system, and then I needed a more advanced controller, with more I/O and some other functions that I want. So I`m designing my own. It is still based on PICAXE, but I will use 3 controllers. Two small ones that will be used mainly as counters, and one larger that will handle IR, upload target positions to the counters, store positions, handle limit switches etc. It has a more powerful H-bridge chip (L298HN) than the first controller, and can handle at least 2x1A motors. It will have a 12V trigger input for power on, and can decode more IR-codes for presets and other functions.

I THINK I have included what I need, and if I`m lucky I get time to etch/drill the circuit board this week. Then some hours of programming..

I also did some mechanical changes to my system. Like many others I used a system where the lower mask was pulled up by the upper roller, and the upper mask moved down by gravity. This give a slight error, since the diameter of the upper roller increases when the fabric rolls on. So, the upper mask will move faster and faster as the fabric rolls on. The opposite happens with the lower mask, and I got problems with binding.

So I made a continous loop of cable with pullys where the masks are attached to the cable, see picture. I put springs from two old roller blinds into the rollers. This works very well as a manual system as well. Just lift/lower one of the masks to the desired posistion and it stays where you put it. The other mask will ofcourse move the same distance in the opposite direction..

Gunnar
LL
Reply
Reply
post #23 of 42
1/20/10 at 9:19pm
Thread Starter 
Quote:
Originally Posted by Gunnar View Post

Gregory,
I used a Picaxe RK18Motor microcontroller board w/h-bridge to control a top/bottom mask system. I used the board without modification. Position feedback was from an optical encoder (modified PC-mouse). The Picaxe is a rather slow controller, but it has a lot of built in functions that is useful in this application. It has built in IR-deocding (Sony codes), PWM-control, analog and digital I/O, eeprom etc..

With this simple board I got 12 posistion presets that were programmable from the IR-remote (jog up/down to desired position, press save button then desired preset button). Presets were stored in eeprom, so they were not lost when system was powered off.

I wanted a 4-way system, and then I needed a more advanced controller, with more I/O and some other functions that I want. So I`m designing my own. It is still based on PICAXE, but I will use 3 controllers. Two small ones that will be used mainly as counters, and one larger that will handle IR, upload target positions to the counters, store positions, handle limit switches etc. It has a more powerful H-bridge chip (L298HN) than the first controller, and can handle at least 2x1A motors. It will have a 12V trigger input for power on, and can decode more IR-codes for presets and other functions.

I THINK I have included what I need, and if I`m lucky I get time to etch/drill the circuit board this week. Then some hours of programming..

I also did some mechanical changes to my system. Like many others I used a system where the lower mask was pulled up by the upper roller, and the upper mask moved down by gravity. This give a slight error, since the diameter of the upper roller increases when the fabric rolls on. So, the upper mask will move faster and faster as the fabric rolls on. The opposite happens with the lower mask, and I got problems with binding.

So I made a continous loop of cable with pullys where the masks are attached to the cable, see picture. I put springs from two old roller blinds into the rollers. This works very well as a manual system as well. Just lift/lower one of the masks to the desired posistion and it stays where you put it. The other mask will ofcourse move the same distance in the opposite direction..

Gunnar

Thanks for the tip on the Picaxe............way more powerful than my circuit. I like the EEPROM for storing the pre-sets and the built in IR. I'll give it a look.

Sounds like you got quite a project going. Good luck to you.

Question about the IR with Sony codes: can other remotes learn the Sony codes or do you have to use a Sony remote?

Thanks,
Greg
Reply
Reply
post #24 of 42
1/21/10 at 1:36am
I use a Philips Pronto universal remote, and downloaded the Sony codes from Remote central. Only the codes from Sony televisions can be used, not other Sony equipment. The PICAXE 18X on the first board can decode 17 of the SOny television codes. My new controller will use the PICAXE 28X, which vil decode over 100 codes. All that is neede is a 3-pin IR-receiver chip of the type that demodulates the IR-signal to baseband. Cost about $2..

I will post the progress on the controller in my thread herehttp://www.avsforum.com/avs-vb/showt...ghlight=gunnar

Gunnar
Reply
Reply
post #25 of 42
1/21/10 at 9:58am
I've seen quite a few good/slick DIY masking designs here.

For "my turn", I'm going to combine an automotive sunroof mechanism with an automotive rear shade mechanism
This on each RH/LH side will pull shut, they even have variable stop mech.


while the shade mechanism (roller) will pull shut (with different dark/black fabric wrap on the roller, not sun shade fabric)


I know people who work at suppliers for both components.
Another summer 2010 job on my to-do list.

I'll put details in my screen build thread, http://www.avsforum.com/avs-vb/showt...3#post17972663
Reply
Reply
post #26 of 42
1/21/10 at 9:24pm
Thread Starter 
Quote:
Originally Posted by Gunnar View Post

I use a Philips Pronto universal remote, and downloaded the Sony codes from Remote central. Only the codes from Sony televisions can be used, not other Sony equipment. The PICAXE 18X on the first board can decode 17 of the SOny television codes. My new controller will use the PICAXE 28X, which vil decode over 100 codes. All that is neede is a 3-pin IR-receiver chip of the type that demodulates the IR-signal to baseband. Cost about $2..

I will post the progress on the controller in my thread herehttp://www.avsforum.com/avs-vb/showt...ghlight=gunnar

Gunnar

Thanks for the answer to my IR question. I saw that you recently updated your thread. Great thread!

I just came up with another design that doesn't use an encoder or a pulse counter. I'll post it after I review it a bit more to make sure I didn't miss anything.

I would like to use a microcontroller, but do not have any experience with them. I'm more used to discrete electronics. I should pick one up an play with it, because it would be easier to do more tasks and it can be changed when needed.



Quote:
Originally Posted by mtbdudex View Post

I've seen quite a few good/slick DIY masking designs here.

For "my turn", I'm going to combine an automotive sunroof mechanism with an automotive rear shade mechanism
This on each RH/LH side will pull shut, they even have variable stop mech.


while the shade mechanism (roller) will pull shut (with different dark/black fabric wrap on the roller, not sun shade fabric)


I know people who work at suppliers for both components.
Another summer 2010 job on my to-do list.

I'll put details in my screen build thread, http://www.avsforum.com/avs-vb/showt...3#post17972663

I just love the creativity and ingenuity in this forum!

Greg
Reply
Reply
post #27 of 42
1/22/10 at 9:41am
I agree! A lot of creativity here! Sunroof mechanism sounds promising.

I also thought of using counters like eg. http://www.web-books.com/eLibrary/En.../DIGI_11P3.htm see bottom circuit. But I don`t felt confident that I could make the rest of the needed electronics. I have almost no experience with microcontrollers, but when I came across the PICAXE it seemed like a chance to learn something, and hopefully make something useful. The PICAXE is a nice controller to start with, as it has a built in serial interface for program download, so you don`t need a programmer. The editing software is free, and there is a user forum where you can ask and find code examples. The code is written in a Basic like language... The downside is that it is slow, much slower than say a standard PIC. But it seems fast enough for this application. have a look at http://www.picaxeforum.co.uk/ You`ll even find manuals for interfacing, programming etc there.

Gunnar
Reply
Reply
post #28 of 42
1/22/10 at 9:19pm
Thread Starter 
Quote:
Originally Posted by Gunnar View Post

I agree! A lot of creativity here! Sunroof mechanism sounds promising.

I also thought of using counters like eg. http://www.web-books.com/eLibrary/En.../DIGI_11P3.htm see bottom circuit. But I don`t felt confident that I could make the rest of the needed electronics. I have almost no experience with microcontrollers, but when I came across the PICAXE it seemed like a chance to learn something, and hopefully make something useful. The PICAXE is a nice controller to start with, as it has a built in serial interface for program download, so you don`t need a programmer. The editing software is free, and there is a user forum where you can ask and find code examples. The code is written in a Basic like language... The downside is that it is slow, much slower than say a standard PIC. But it seems fast enough for this application. have a look at http://www.picaxeforum.co.uk/ You`ll even find manuals for interfacing, programming etc there.

Gunnar


That engineering site is great!

As far as counters go, you can buy various counter IC's, as well as other application specific IC's...............you don't have to design with the basic building blocks, like flip-flops and logic gates. My approach was to use the specifc IC's, such as counters for example, otherwise the real estate would get quite large with the amount of small-scale devices needed. I still need some logic gates to steer and decode signals properly. However, the microcontroller will give you the least real estate penalty of all.

Thanks,
Greg
Reply
Reply
post #29 of 42
1/25/10 at 9:47pm
Thread Starter 
Attached, is a rough sketch (very ugly, too) of the non-encoder based motor controller (no optical encoder or pulse counters).

The relays and the contacts at the bottom are outputs from an IR receiver module. Four commands will set 4 binary input codes (one for each aspect ratio). I may choose to add more, but for now this is adequate to describe the operation. The binary input codes are inputted into the magnitude comparator and the 1 of 4 decoder. The photodetectors, which will be attached to the sliding mask assembly, are shown as the circles at the very top of the page. As the masks move, the photodetector will eventually pass over a light beam (LED) and trigger the motor to stop. Each photodetector has a binary destination address that corresponds to one of the binary input codes. These addresses are set by the D flip-flop pairs. The D flip-flop pairs holds the last address as the motor is transitioning to the next address, as there are no continuous positioning data, as there would be if a counter was counting pulses from an encoder. In this design the position between aspect ratios is not known or needed, only where it came from and where its going. The magnitude comparator compares the two addresses and based on result moves the motor in one direction or the other. The AND gates feeding the D flip-flop pairs clock the destination address in when the photodetector is triggered.the photodetector for the destination address is enabled by the 1 of 4 decoder, based on the input address selected. Once the destination address is clocked in it is immediately compared to the input address and if they match the motor stops. At the same time the OR gates next to the AND gates resets all of the D flip-flop pairs, except for the one associated with the destination address.

Heres an example:

- Let's say the masks are set in the home position (binary address 00 = 2.35 aspect ratio).

- A binary input code of 11 is entered, which corresponds to an aspect ratio of 1.33.

- The decoder enables the address 11 AND gate, which is waiting for the photodetector to be triggered.

- The magnitude comparator compares the input address (B = 11) to the initial address (destination address A = 00). The result is A

- As the motor moves, the photodetectors between the initial and final destination (1.85 and 1.78) will be ignored because their respective AND gates have not been enabled by the decoder, although these photodetectors will respond and generate a pulse, it will be ignored.

- When the photodetector at the destination address is triggered the AND gate will pass the low to high transition to the associated D flip-flop pairs, which will clock in the destination address (11) into the magnitude comparator. As both the input and destination address are now equal, the AB lines of the comparator will both be low and the A=B line will be high. The AB lines are used in the motor controller to control FORWARD, REVERSE or STOP (high brake mode).

- Simultaneously, the same clock signal will reset the other D flip-flop pairs, ensuring that the destination address is not corrupted by the previous address.

One possible issue is that the motor will not stop quickly enough. The circuit will still work, but the mask position will be off a bit (probably not a big deal). If it is a big issue one possible remedy will be to provide a pulse width modulation signal to the motor to slow it down prior to it being triggered by the photodetector. The method would be to include two additional photodetectors, one each side of the primary detector, spaced maybe 1 away. As the 1st photodetector is triggered, the signal will enable the PWM to go from full speed to some slower speed and then when the primary detector is triggered the motor would have an easier time stopping when the brakes are put on.

It is still an involved project.it just doesnt use an encoder and a counter. It also involves more mechanical fiddling because of the alignment of each of the photodetectors. The complexity will grow if I want to add more aspect ratios than what I showed here. The encoder based design does not suffer from this expandability issue.

Well, that was a long-winded post.sorry if I bored you!

Greg

 

Non-encoder Based Motor Controller.pdf 173.3994140625k . file
Reply
Reply
post #30 of 42
1/25/10 at 11:49pm
Wow, following with great interest & way over my head. Thank you for sharing your progress with your project. This approach would avoid any programing, is that the big advantage?

John
Reply
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: DIY Screen Section

AVS Top Picks

AVS › AVS Forum › Display Devices › Screens › DIY Screen Section › DIY Vertical Masking Design......Need Design Advice