DIY Scanning Disk TV - Page 4 - Videokarma.org TV - Video

old_tv_nut · #46 09-07-2005, 04:09 PM

Sorry I don't have a stepper control circuit to offer. Seems to me you would still need a PLL to generate the proper pulse rate for the stepper motor, although that PLL could be built from an application note cookbook procedure because it wouldn't need to take motor/disc inertia into account.

mbates14 · #47 09-07-2005, 04:40 PM

ive got a stepper control circuit already, it is 7.5deg per step, 48 steps per revolution, and 600 steps for a 12.5rps.

I just need a circuit that generates the 600hz pulse, and varies that pulse based on the sync. probably consists of another PLL circuit, but you wouldnt need the optical fork, and it would have perfect frame lock.

i fried my 40406

swung a 30 volt rail by accident, and cooked it.

I do have an lm565 which is a pll.

mbates14 · #48 09-08-2005, 06:28 PM

welp, thats not gonna work. I cant get the motor fast enough for 12.5fps.

itll try, but as soon as I start to increase the oscillator from 485hz to any faster, the motor will jitter, and stop. Oh well.

mbates14 · #49 09-08-2005, 08:17 PM

ive been studying, and studying, and trying to figure out why it failed the first time, I ran across yantzers pages, and some other pages. I think i found my problem. I need a schmidt trigger!!! because i noticed, that his sync pulse looks like mine from the video amp, but my return loop pulse has lots of variations, and amplitiudes. But his looks just like the sync pulse, except the duty cycle is different.

I noticed he used a 3 lead optical pickup that was actually an IC that is a schmidt trigger. I looked up what schmidt triggers are, they take that same noisy variating signal, based on thresholds set by resistors, puts out a clean square wave signal!!! THAT was my problem .

So, i need to get that 4046, and try this approch. I got nothin to loose, except time and maybe anohter 4046. LOLz

old_tv_nut · #50 09-08-2005, 08:57 PM

Mike,

By "return loop pulse" you mean the pulse from the disc and your optical fork, I believe? Yes, it has to be full amplitude, should always drive the full zero to 12 volts. The schmidt trigger is a good fix for this, but it's also possible to do it with careful positioning of the fork and a good disk.

Have you been trying to work with small sync holes? They do not have to be small at all, because the 4046 is edge-triggered - it won't care if the disc pulses are wider than the sync pulses. The scanning holes need to be small in order to get good resolution, but the sync holes can be big to make sure you get plenty of signal out of the fork. Also, you could elongate the sync holes radially- they could look more like slots so that the fork position is less critical.

mbates14 · #51 09-08-2005, 09:08 PM

my whole saying is, Doesnt the edges of the waves have to be the same in order to compare the phases to be synced?? well, if thats the case, both waves are ENTIRELY different. And yes, i have been working with small holes, i just widened them. But I cant do anything till I get a pll chip.

old_tv_nut · #52 09-08-2005, 09:38 PM

Are all the pulses full amplitude and clipping at the max voltage now? That is the first important thing.

The 4046 contains more than one phase detector, connected to different output pins. The one used in all the club schematics is edge-triggered (I think positive-going edges, but I forget), so it doesn't care about the pulse width. However, it may not like pulses that rise too slowly, so plenty of signal amplitude and clipping or else a Schmidt trigger to make sure it rises quickly is the second important thing. The other nice thing about this phase detector is that it is also a frequency detector, so if the disk speed is too high or low, it will be forced back toward the correct value.

The other more common types of phase detector (including the unused ones in the 4046) are sensitive to pulse width and might like to have a 50% duty cycle on both input pulses, depending on the design. They also have only a weak frequency-offset response, so the speed has to be within a certain "pull-in" range to start with.

Can you post a picture of the oscilloscope trace of your optical pulses?

yagosaga · #53 10-10-2005, 03:14 PM

The development of color on the Nipkow disk made some progress, see:

http://www.nbtv.wyenet.co.uk/hot.htm

and

http://www.copycoder.com/phpbb2/viewtopic.php?p=126#126

(Mickey Mouse on the Nipkow disk in color.)

My own mechanical TV set is disassembled now. With a little help from my friends it will get a new and professional looking wooden house.

mbates14 · #54 10-10-2005, 04:28 PM

im still havin the problem where the whites are washed out, and darks are way too dark, you cant see them.

Im wondering if its because I used lm741 with single supply voltage. Dunno.

Anyway, that picture of the kids face you have on your site, looks like a ghost head. thats how bad it is.

old_tv_nut · #55 10-15-2005, 09:36 PM

Can you post the diagram of your video output circuit including the LEDs? If the whites are bright enough but clipped, you probably have too little resistance in the FET source circuit and/or too many LEDs in series in the drain circuit. (???) If you could measure the voltage swing at source and drain with a scopoe, that would help diagnose it.

mbates14 · #56 10-16-2005, 09:26 AM

im only using 3 in parallel, but one bleedoff resistor.

the signal does not look clipped on the scope. But at the point where its not clipped, I can barely see it. Im using 15,000mcd leds.

But when I turn it up to start to really see it, it begins to clip, and can see it clearly on teh scope. But, when I lower it out of clipping, the scope cant trigger it.

old_tv_nut · #57 10-16-2005, 10:15 PM

Hmmm...
It will be hard to see a signal on the LEDs because the LEDs are very non-linear. So I would expect the blacks (with the FET cut off) give a voltage equal to the power supply, and then if you draw even a little current, you see the LED voltage drop, and this doesn't change linearly with current, so it appears to be clipped. The resistor you have will add a linear voltage drop at the drain. This in itself should not be a big problem for getting a good picture if the resistor is not too big, as whites should not be clipping in the FET, that is, the drain voltage should not drop so low as to equal the source voltage. You should verify that this is true. What are the voltage ranges of the signals at gate, source, and drain? (Be sure to measure with the scope DC-coupled.)

What else can be wrong? Maybe you have a problem with current sharing between the LEDs - you could try putting some resistance in series with each one (maybe 100 ohms?) instead of one common resistor and see if that helps.

By the way, when you get a visible but bad image, does it look like the circuit could be oscillating? It is sensitive to putting your finger or a scope probe on different parts of the circuit?

mbates14 · #58 10-16-2005, 10:58 PM

my scopes dont have any division scales, so I cant take readings, i can only look at signals and know they are there. LOL. oh well.

yagosaga · #59 11-26-2005, 01:10 PM

My scanning disk monitor has got a new housing.
The woodworks were done by a friend, Ulrich Barkow, who is a retired physicist and the master of the city of Broitzem here. When he saw the first draft of my monitor working, he was so impressed that he asked me to make a new housing for it. He has a special workbench and tools for woodworking and will introduce me to do the same. I am fascinated and I wish I can do it in the same way. See the photos:

http://bs.cyty.com/menschen/e-etzold...sk_housing.JPG
http://bs.cyty.com/menschen/e-etzold...k_optofork.JPG
http://bs.cyty.com/menschen/e-etzold...isk_inside.JPG

Door open:

http://bs.cyty.com/menschen/e-etzold...disk_front.JPG
http://bs.cyty.com/menschen/e-etzold...isk_front1.JPG
http://bs.cyty.com/menschen/e-etzold...isk_front2.JPG

He has put the motor and the disk bearing into a wooden block, which is fixed with screws at the ground plate. The wooden ground plate now is removable. The surface of the wood is oiled.
I have constructed this set in this way that everybody can see it working from the outside and understand how it works. This is the "philosophy" behind this set. I want to use this set for teaching childs and young people to understand the fundamental principles of television. When they carefully observe this set they found the fundamental principles by themselfes and can explain it. This is not possible with modern tv sets.

mbates14 · #60 11-26-2005, 02:35 PM

You know, I just thought of some crazy idea yesterday.

Use an array of UV LEDs for the light source, but on the disk, instead of holes for scanning, use a UV pen, and and mark where the holes supposed to be, so in the dark, its got a cool looking effect.