Mechanical TV Project

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Steve Anderson, Dave Moll, Andrew Davie

Mechanical TV Project

Postby FlyMario » Thu Jul 12, 2018 2:21 pm

Good evening,
I am working on a home project to build a Mechanical TV. Currently I am trying to work out the Nipkow disk. I am using a 12" (304.8mm) vinyl record as the disk itself and trying to work out the hole sizes to use. I read somewhere to use a 1/16" (1.57mm) drill bit. I was pretty pleased with that notion because I felt I would probably have no issues drilling through records at that size. In solidworks I drew the disk and placed these holes and was not pleased. The pattern did not look like it would be correct for a Baird style tv. The hole patterns would cover too much are. Thinking there would be less light let through away from the center I allowed the holes to overlap (1.2mm center to center) but still it looked much more wide than tall.

So guess I am looking for someone to maybe give me advice. Should the holes be overlapping more or am I just going to get a much smaller bit. I am going for 32 holes.

Thanks for any suggestions you can offer.
Attachments
DiskWide.jpg
holesClose.jpg
FlyMario
 

Re: Mechanical TV Project

Postby Harry Dalek » Thu Jul 12, 2018 7:57 pm

FlyMario wrote:Good evening,
I am working on a home project to build a Mechanical TV. Currently I am trying to work out the Nipkow disk. I am using a 12" (304.8mm) vinyl record as the disk itself and trying to work out the hole sizes to use. I read somewhere to use a 1/16" (1.57mm) drill bit. I was pretty pleased with that notion because I felt I would probably have no issues drilling through records at that size. In solidworks I drew the disk and placed these holes and was not pleased. The pattern did not look like it would be correct for a Baird style tv. The hole patterns would cover too much are. Thinking there would be less light let through away from the center I allowed the holes to overlap (1.2mm center to center) but still it looked much more wide than tall.

So guess I am looking for someone to maybe give me advice. Should the holes be overlapping more or am I just going to get a much smaller bit. I am going for 32 holes.

Thanks for any suggestions you can offer.


Have not made one in a while but i always use a pin instead of a drill bit which works well for plastic records or CD dvds ...if you can do it try both ideas on 2 records and see the results the gap one will of cause give the line look...i have never been able to do the over lap idea well not on purpose ! hand made disks are hit or miss i tended to make them with dvds or cds and made a few picked the best for the projects .
There are past posts about what you are up to on the forum...read read read and read some more :wink:
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 3852
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Re: Mechanical TV Project

Postby gary » Thu Jul 12, 2018 8:58 pm

FlyMario wrote:Good evening,
I am working on a home project to build a Mechanical TV. Currently I am trying to work out the Nipkow disk. I am using a 12" (304.8mm) vinyl record as the disk itself and trying to work out the hole sizes to use. I read somewhere to use a 1/16" (1.57mm) drill bit. I was pretty pleased with that notion because I felt I would probably have no issues drilling through records at that size. In solidworks I drew the disk and placed these holes and was not pleased. The pattern did not look like it would be correct for a Baird style tv. The hole patterns would cover too much are. Thinking there would be less light let through away from the center I allowed the holes to overlap (1.2mm center to center) but still it looked much more wide than tall.

So guess I am looking for someone to maybe give me advice. Should the holes be overlapping more or am I just going to get a much smaller bit. I am going for 32 holes.

Thanks for any suggestions you can offer.


If you download NipkowDXF from my web site you can enter the dimensions of the disk and (using the calc button) calculate the step size required for that sized disk. This step size is the distance between each aperture and is optimal for a rectangular aperture (or square if aperture ratio is 1).

The amount of overlap when using a circular aperture is a little controversial as it is a compromise between visible lines and definition. I would suggest you start with no overlap and if you find that objectionable then increase the overlap slightly until you are happy with it. I would suggest that the amount of overlap should not result in an area greater than the area lost when superimposing a circle of the same width as a rectangle over that rectangle.

Personally I prefer visible lines to reduced definition.

If you prefer to calculate the step size yourself then this is the formula: s = R(max) / (((n^2 * a) / 2 * PI) + ((n - 1) / 2))

Where s = step size
R(max) = centre of the outermost aperture
n = number of lines
a = aspect ratio
PI = delicious stuff encased in pastry (just kidding)
* = multiplication
/ = division
Perfecting an NBTV system is like trying to slam a revolving door...
User avatar
gary
"Fester! Don't do that to 'Thing'"
 
Posts: 2288
Joined: Sat Jan 27, 2007 11:29 am
Location: Bundanoon, Australia

Re: Mechanical TV Project

Postby FlyMario » Fri Jul 13, 2018 12:34 am

Thanks a lot for your comments guys! I have purchased some smaller bits (.4 mm and bigger) with a pin vise. Hopefully, they won't break that easily if I go slow. So my plan is to use .4mm holes. Looking at that in SolidWorks makes me feel the results would be more accurate.

As long as at the end of the day I feel I have something in the spirit of Baird Televisor I will be happy.

Even though Baird used a Counter Clockwise running motor having the "pixels" going from the outside bottom to the inside top, I for some reason can't bring myself to do that. Mine will travel Clockwise and the "Pixels" will go upper left to lower right. It satisfies my sense of normalcy.

I have built a structure out of Makebeam with a DC motor which is geared down to 1000 rpm. I will use PWM to get the motor to run at 750 rpm to accomplish 12.5fps. I know the motor will accomplish this because I did make a "Timing Light" out of an Arduino which did allow me to "stop" the record. The biggest issue I see is that the motor does not run terribly consistent. Sure I can monitor the RPM and increase/decrease the PWM accordingly however only experiments will tell me if I can get it to stabilize.

Of course, the easy way out would be just to speed up / slow down the pixels based upon the actual rpm at the time but it feels a bit out of the "spirit" of the original TV.

I plan on having the view window to the right of the spinning disk and possibly a view window at the top

I also included a picture of the build as it is (with the record removed). I mounted a stepper motor temporarily so that I can use a program to rotate the disk accurately around so that I can mark where my holes should go with my 5w laser. I have another linear slide (not shown) to move the laser left and right on the disk.

Thanks!
Attachments
20180710_160947.jpg
20180711_212234.jpg
FlyMario
 

Re: Mechanical TV Project

Postby smeezekitty » Fri Jul 13, 2018 2:32 am

1.57mm holes are definitely too big as pointed out. I ended up using a 1mm drill bit.

You wont want to use constant speed on the disk. It needs to be phase locked to the signal. Otherwise you will have rolling or an out of position picture.
smeezekitty
Evil Genius
 
Posts: 124
Joined: Fri Jan 29, 2010 11:42 am
Location: USA

Re: Mechanical TV Project

Postby FlyMario » Fri Jul 13, 2018 3:43 am

smeezekitty wrote:You wont want to use constant speed on the disk. It needs to be phase locked to the signal. Otherwise you will have rolling or an out of position picture.


So, do you guys have videos recorded for this kind of system already or do you convert video (downscale) for these Televisors? I was looking for some to download to try and started to believe I will have to create them myself.

Thanks,
FlyMario
FlyMario
 

Re: Mechanical TV Project

Postby smeezekitty » Fri Jul 13, 2018 6:01 pm

FlyMario wrote:
smeezekitty wrote:You wont want to use constant speed on the disk. It needs to be phase locked to the signal. Otherwise you will have rolling or an out of position picture.


So, do you guys have videos recorded for this kind of system already or do you convert video (downscale) for these Televisors? I was looking for some to download to try and started to believe I will have to create them myself.

Thanks,
FlyMario

There is software that will convert pictures and videos to NBTV audio and create a .wav or output directly to the sound card.
The club standard is 32 lines, 12.5 FPS with a (~175 uS) negative sync pulse on 31 of the 32 lines. The sync pulse allows a sync circuit to lock the disc to the correct line and frame rate. There are a number of different ways to do that.
Some people sync only once per frame and others sync every line. Some experimenters use dedicated PLL chips although I have been using an Atmega328 programmed in the arduino environment implementing a soft PLL and PID control.

The most well known program is Gary's Video2NBTV

I have been developing a new program called FreeNBTV since I wanted to be able to tune all the parameters. It is less stable and has a bigger learning curve than Gary's but it lets you adjust framerates, sync lengths and levels and a lot more. It also uses FFMPEG to read the video files so it will handle just about any format.
smeezekitty
Evil Genius
 
Posts: 124
Joined: Fri Jan 29, 2010 11:42 am
Location: USA

Re: Mechanical TV Project

Postby Klaas Robers » Fri Jul 13, 2018 6:47 pm

To begin with: start with prerecorded NBTV video. There is quite some on CD in the NBTVA shop, but as the shop owner is seriously ill at the moment, it is difficult to obtain that on a short time base.
However on the web pages of the NBTVA (http://WWW.NBTV.ORG) you may find .wave files of NBTV video.

1. on the NBTV home page you find at the very bottom a link: "the characteristic NBTV sound". You are guided to a page where you can listen at 12 seconds of NBTV video-sound and where you can download a .wave file of that 12 seconds.

2. on the NBTV home page follow the link "Hot news" in the top line of the page. There is a second item (not so hot), called "The MUTR NBTV compact disc". In this item is at the bottom of this short message a link "downoaded" which leads to a page with explanation about the CD and 26 minutes of NBTV pictures and sound, as well in upright polarity as in negative polarity. You may play these .wav files directly from your computer, or make a CD with both polarities on the same disc.

All this video is shot in a scanning of vertical lines, from bottom right up to top left, the Baird-way of scanning. If you do your scanning the reverse way, you still have pictures, but positioned upside down and mirrored left-right. I think Baird has selected this weird way of scanning because the visibility of the rather slow scanning is then least annoying / visible. See that you can rather simply reverse your scanning direction. If there is readable text in the picture, your reading eye movement is opposite to the scanning direction. I don't know if this applies too to just pictures.

As far as I remember the aperures on a 30 cm disc should be about 0.7 mm. The visibility of round holes is least if they have an overlap of 20%.
User avatar
Klaas Robers
Frankenstein was my uncle.
 
Posts: 1403
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Mechanical TV Project

Postby FlyMario » Fri Jul 13, 2018 8:37 pm

Thanks a ton guys!

I have downloaded the Mutr files.

Its going to take me awhile to digest this :)
FlyMario
 

Re: Mechanical TV Project

Postby FlyMario » Fri Jul 13, 2018 11:05 pm

After seeing this information I will definitely alter the idea and have the disk go Counter-Clockwise having dots go from bottom to top, right to left. At least it can be compatible with what others would normally generate with software or from available .wav files.

Glad I have not drilled holes yet.

I am curious about the LED's brightness. I had thought that I would alter a PWM pulse to lower the brightness for each pixel. The circuitry I have seen shows that people are modifying the current based upon the wave. I know that LED's (mine is 10w) can be modified incredibly fast, but if I were to alter PWM that fast the results might not be that impressive. Altering the current would be much easier. Time to dig out a MOSFET.
FlyMario
 

Re: Mechanical TV Project

Postby FlyMario » Sat Jul 14, 2018 12:56 am

smeezekitty wrote:
FlyMario wrote:
smeezekitty wrote:You wont want to use constant speed on the disk. It needs to be phase locked to the signal. Otherwise you will have rolling or an out of position picture.


So, do you guys have videos recorded for this kind of system already or do you convert video (downscale) for these Televisors? I was looking for some to download to try and started to believe I will have to create them myself.

Thanks,
FlyMario

There is software that will convert pictures and videos to NBTV audio and create a .wav or output directly to the sound card.
The club standard is 32 lines, 12.5 FPS with a (~175 uS) negative sync pulse on 31 of the 32 lines. The sync pulse allows a sync circuit to lock the disc to the correct line and frame rate. There are a number of different ways to do that.
Some people sync only once per frame and others sync every line. Some experimenters use dedicated PLL chips although I have been using an Atmega328 programmed in the arduino environment implementing a soft PLL and PID control.

The most well known program is Gary's Video2NBTV

I have been developing a new program called FreeNBTV since I wanted to be able to tune all the parameters. It is less stable and has a bigger learning curve than Gary's but it lets you adjust framerates, sync lengths and levels and a lot more. It also uses FFMPEG to read the video files so it will handle just about any format.


Ahh that is why I kept seeing a negative level. What I kept reading from the standards led me to believe from 0v to .3v to .5v (30% to 50% of 1v) was the sync pulse. However looking at the waveforms in an audio program had me confused because it was going negative. Now I get it. It is interesting to think people were adjusting the motor speed per line. I would think the disk (Record) I am using would make the changes to be too slow for that to be effective. But you know... it sounds like a fun experiment. I am a fan of PID and would probably implement it.
FlyMario
 

Re: Mechanical TV Project

Postby smeezekitty » Sat Jul 14, 2018 2:59 am

Yeah PWM isn't likely to give you good results in this case. It will probably leave artifacts in the picture. Analog control is the way to go. A 10W LED is also *way* overkill IMO. I have about 1/2W worth of 5mm LEDs and its plenty bright. With smaller apertures, you might want a little brighter.

Counterclockwise (bottom to top, right to left) disc rotation is the way to go to be compatible with existing content.

It is interesting to think people were adjusting the motor speed per line.

I find adjusting per line is not at all necessary. The disc has plenty of inertia that adjusting every frame gives a good stable picture.
The one tricky thing about adjusting per frame is you need to detect the missing pulse
smeezekitty
Evil Genius
 
Posts: 124
Joined: Fri Jan 29, 2010 11:42 am
Location: USA

Re: Mechanical TV Project

Postby Klaas Robers » Sun Jul 15, 2018 1:50 am

Indeed, it is wise to stick to a kind of standard that provides you with attractive NBTV-video program material, at least in the beginning.

In principle drilling the disc is independant from the rotation direction. If you assure that you can flip over the disc itself, so look at the A-side, or the B-side, and you can reverse the rotation direction, you can accomodate each scanning mode. The only point is to have the pitch of the lines (holes) correct.

There are computer programs that accept a certain disc diameter, and then print out a plan for drilling. Paste that print-out on your disc and drill the holes at the indicated spots. The program also calculates the optimal size of the drill. Do the drilling VERY CAREFULLY, this defines the precision of your line pattern. In fact this cannot be done by hand, but ignore this for your first discs. It is more important THAT you make a disc, than how precise it is.

LEDs can be modulated in an analogue way, the light output is linearly proportional to the CURRENT through the diode. There is a simple one-FET circuit that works better than you might expect. It is in the NBTV-Handbook, look on the NBTVA website (http://www.nbtv.org) and follow the links to Handbook, Contents. I just placed it there this afternoon for your convenience. It is a .pdf, ready for printing on A5.
User avatar
Klaas Robers
Frankenstein was my uncle.
 
Posts: 1403
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Mechanical TV Project

Postby FlyMario » Sun Jul 15, 2018 6:28 am

Thanks a ton Klass and SmeezeKitty!

I don't see the schematic on the site. I am in no hurry to deal with that circuit at this time.

Today had a disk drawing printed out and drilled the holes in my disk. After spinning it up and showing a light through the holes I feel it is at least decent enough for performing tests.
FlyMario
 

Re: Mechanical TV Project

Postby Dave Moll » Sun Jul 15, 2018 5:58 pm

FlyMario wrote:I don't see the schematic on the site.


There is a single-transistor LED driver at no. 5 in the contents. Is that it?
User avatar
Dave Moll
Just nod and pretend you understand me
 
Posts: 358
Joined: Tue Feb 27, 2007 9:11 am

Next

Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 2 guests