DIY Scanning Disk TV
 

Didnt know where to post this, so I thought here was good. I thought i would try something new, so I made a scanning disk TV, I burned a disk with some test audio for the disk, and I build the disk based off the design and schematic at this site:
http://bs.cyty.com/menschen/e-etzold...anningdisc.htm

I got everything working, but I have one problem. No Sync.

I have everything built according to the schematic.

But for some reason, my motor is running full tilt boogie. and its not slowing down to the incoming sync. I verified I have a sync pulse from the sync seperator, and I have somewhat of a pulse (even though its way higher in freq) from the optical fork on the disk. So thats working, Both signals are making it to the PLL circuitry, but for some reason, the gate is fully charged, driving it into saturation, which turns the motor on full way.

hmmm. Its no working. Can anyone give me any pointers? note: If i slow the disk down with a rheostat, I can see NBTV. so, I know my circuits are working, but If i hook the motor back up to the MOSFET, it runs full tilt boogie. If I unhook the gate, and ground it. the motor quits, so I know the FET is good.


my pulse from the optical fork, does NOT no where near look like the one on the site. My sync pulse from the seperator looks the same, but the disk fork doesnt. its more rounded and lots of noise, since I have to crank the oscop all the way down on the V/div to see it.

Figure this one out: If I disconnect pin16, which is the supply of the PLL chip, it starts working.

the motor speeds up when you first power it up, when you push play, it slows down to a constant speed.

nevermind. That dont work either, it just makes the motor vary with signal, but still dont sync, as I can block the fork, and it still isnt making any difference. Oh well. :worried:

You say your sync signal is very small? The 4046 is a high-voltage CMOS which requires its inputs to swing almost the full range from zero to supply voltage. If you are not getting a large sync pulse from your sync fork you need to fix that first. If you areusing the schematic at
http://bs.cyty.com/menschen/e-etzold...BTVmonitor.GIF
Note that the variable resistor in the phototransistor emitter should be 47k (47000) ohms. The "k" may be hard to read.

I have to crank up the drive to .05V/Div to see it good on the screen. but I have to mess with manual trigger, because of the noise.

I made the motor servo and the video/seperator circuit on two seperate power supply batteries, two seperate grounds. That doesnt make since either, how does the sync pulse return, if there is no return path between both stages (grounds are seperate). Both are 12V batteries.

I can see NBTV if I brake the disk. otherwise it spins too fast. motor is getting full 12v.

the biggest pot I have is a 10k. I even tried a 47K resistor, no noticable difference on the scope, its still weak, but its a little bit stronger.

100K resistor provides the strongest from the fork, but not enough I suppose.

I got a 100 K pot, and varied from top to bottom, I seen the signal peak to about 2v, so thats OK I guess.

But still, disk spins at full speed.

If I unhook the fork from the PLL chip, it slows down to whatever I set my speed pot at. as soon as I hook the phototransistor back in the circuit, it speeds up again.


What phase should the fork be, positive or negative? im thinking there is a phase problem or something I am totally stuck. :no:

I got my phototransistor and LED from an old remote control and sensor, I took the sensor out of the IR module, and took the LED out of the remote.

im thinking there is a fault in the loop somewhere, but I read your site, and read the PLL section, and it did nothing but confuse the hell out of me :banana:

That's exactly the same result which I got when I have built the motor drive circuit according to Peter Smith's schematics:
http://www.nbtv.wyenet.co.uk/speedctrlckt.gif
First check the pulse from the opto fork. I had to vary the distance of the light emitting diode to find the point were I get the best pulse. When I have found it, the adjustment of the 47K poti was not critical.
If you compare Peter Smith's schematics with mine you will see that I had done some changes. These were necessary for getting a line sync. The frame sync is not working well now. The 4046 needs a clear-cut digital input signal. In the next step I will insert a Schmitt-trigger between the opto fork and the 4046.
Another problem is the IRF520. The original IRF510 (which I couldn't get here) needs a lower gain for regulating the motor speed. The IRF520 will be overdrived by the 4046, so I lowered the input voltage.

I tried the IRF510, I also tried a 2sk somethinanother. And it didnt work.


I notice on your scope pic, you got a clean signal from the fork. Mine is all noisey and very jumpy.


Any ideas?

(My disc is made out of cardboard, Very thin cardboard based off your excel sheets, and it scans fine. Clean picture, formatted to fit a standard sheet of paper).

I built mine based of your schematics.

I only have a TL072 and a LM741 op amps. Not the CAblah blah blah.

So I used those. The video works fine. But the motor servo is not workin period. If I disconnect the reference signal, it slows down to whatever I have the pot set at. As soon as I connect the reference signal, zrrrrroooommm. full speed. :scratch2:


My disc isnt 12 inches, its more like 6 to 8 inches (ran off a computer).

But If I get the disc to speed and scan normally by manual braking, the pulses on the scope look exact same compared to the reference signal, so I know thats not the issue. The servo isnt doing the job. The comparator isnt working, its like its adding the phases, not subtracting them.

Both the reference and fork pulses are positive. Isnt one of them supposed to be inverted? or not.


Isnt it supposed to find the difference between the two phases, and generate an offset voltage? or not. I dont know how PLL works, so bare with me. If it has to find the difference, then one of the signals MUST be inverted, or oscillation will occur? both signals are positive.

also forgot to mention, ive got everything breadboarded. I can move around, add, subtract, or whatever as necessary.

The whole entire thing is made out of cardboard. (its amazing what you can do with an old detergent box).

the disk had thick cardboard, That proved to be off center, and couldnt see the light, and wobbled.

So I used a very thin cardboard, and precisely punched the holes, including the center one. I left the pattern on the disc, and put it on the motor shaft and hotglued it. it spins up, and the circle around the hole is dead center, It doesnt bobb as if it were off-center. Its perfectly still, even though the disk is turning. So I know my disk is good.

I give up. Ive tried everything. I just shortened the fork as close as I can get it without scraping the disk, and still full speed.


CRRRAAAPPPP. Nothing works. Stupid ass TV.

mbates14,

Your PLL circuit or your opto-sensor are not working. Remove your scanning disc so the light beam reach the detector side without problem. Then measure the DC voltage reading at pin 3 of the 4046. In this condition the photo-transistor in the opto-sensor should be in full conduction and you should have close to 12V on pin 3. While you are at it, also take a voltage reading on both sides of the LED side of the opto-sensor.

Place a piece of cardbord in the slot to hide the ligh then recheck the voltage at pin 3. You should now have close to 0 volt DC.

Then I'll try to see what's wrong with your circuit.

Come to think about it, you said that you are powering your circuit from two separate batteries. It does not show on the schematic, but I think the two GNDs (-)should be connected together or your sync pulse won't have any effect on the PLL circuit !

Good luck !

Eric

Hey, What I did was built a PNP amplifier off the photodetector. I can see a good strong signal now.

And It still didnt solve my problem, but for kicks, I took the reference and fork signals and switched them on the PLL. Now, its trying to lock!!!

Itll lock, and as soon as it locks, it will quickly drop off, and lock again. and again and again. I can hear the motor slowing and speeding. So, the PLL is working.

I took the signals on pin 3 and pin 14 and switched them. Now its trying to work.

Basically itll lock the frame, and then roll, and lock as suddenly as it rolls, and keeps doing that. maybe the PLL is working now, and trying to lock onto noise, who knows.

Hey, its got more life now than what it did. :smoke:

It is morning here in Europe and I have had no morning coffee, so forgive me my bad English.
It is the same trouble I had. In the first attemps I measured the opto fork output with a microampere meter to see whether it works or not. And when it works the next problem was to adjust the motor drive in such a way that one can control the motor speed. You can do it with changing the resistor values.
Resistor at pin 13 replace with 100K. The next resistor, connected with the speed control poti determines the basic speed of the motor. Take a lower value and you have an improved speed control range. I have experimented with values between 20K and 100K. You have to find out what works best. It is stupid and boring work, but I know no better way.
A paper Nipkow disk is not the best solution. For a more stable Nipkow disk, you can use a blank cover CD. Print the attached picture on a piece of paper with your printer (paper should not be diffuse). Sting the holes with a needle, cut the disk out of the paper and put it on a CD with some glu. Take a motor from an old CD-ROM drive and you can drive the disk with the CD-ROM motor. But note that the CD motor needs a very low voltage, if the motor runs too fast, add a resistor (20 - 100 Ohms) parallel to the motor.

Already done that. I used the speed pot to slow down the motor right where it almost holds. And then I hooked the sync back up.

Like I said before, it tries to lock, and as soon as it does, the motor gets another kick from the pll and it looses lock. and itll try to lock again, and when it does, it gets another kick from the pll. hmmm.


is this because I had to switch the syncs for it to work? because remember, if I switch it back like in the schematic, it will fail to operate, and drives the mosfet into saturation, thus running the motor full speed.

that one is already at 100k. I didnt have a 147, so i used a 100.

p.s. its cardboard, not paper. And all my cdroms I have are stepper motors, not two wire motors.

besides, wouldnt I have to have a microscope to see it?

Perhaps this could be the reason why the motor looses lock. I had experimented with CD-ROM drive motors and the 4046, and this doesn't work, so I use a D.C. control with manual sync. Perhaps you get a better result if you insert a high wattage low Ohm resistor parallel to your motor.
I think, a magnifying glass is enough. But a better way might be to order a 12 inch Nipkow disk from Peter Yanczer. The picture is appr. 1" x 0.6" which is sufficient for direct view.

im using a carboard disc, and a cassette motor.

With a belt drive?

no, dont have any hubs or anything.

I worked with what I had. Carboard secured right at the shaft.

I have been trying to follow what you have been doing, and I don't see where you connected the two grounds. Did you do that? If so, you should next see if the circuit has enough range to get to the right speed.

When you block the fork, it should appear to the PLL that the disk is too slow, hence the circuit should run up to maximum speed. On the other hand, if the gate of the MOSFET is grounded temporarily, the motor should slow way down or stop.

I also think the connections of pins 3 and 14 are correct in the schematic - if the disk is slow (test by blocking the fork), the output on pin 13 should go high, turn on the MOSFET, and make the motor run fast.

It does not matter if the input pulses are wide or narrow, since this phase detector detects pulse edges. So, you should be able to use either polarity of fork signal, as long as it swings from zero to twelve volts. However, the edges should be fairly clean and quick rise/fall time.

if I block the fork, nothing happens, but the motor no longer gets its "kick"

if i unblock the fork, itll try to sync again.

its only got a 2vP from the fork

ok, I ran some more tests.

With the sync backwards, as we know, it tries to work.

Now, If i bring the optical fork closer, it speeds off. So, I switched the sync inputs back to normal. The disk went full speed.

So, I made the fork closer together, like i did before, and still nothing. no signs of speeding down. I got them so close, it was scraping the disk. I got a good strong signal from the fork too.

So, i moved the syncs back the other way, and it started to work again, and like I said if i move the fork, it speeds up. I move it back it works fine. But no accurate sync. hmmmmm. sooooo confuzzled.

its workin now.

It only works if its connected backwards, I also put a filter on the fork output.

So, its trying its damndest to stay locked. Itll lock the frame, when the frame just starts to fall, the PLL kicks it, but the motor overshoots, and itll loose lock for a split second, and slowly fall into place.

If i blow on the disc to slow it down, almost immidiatly after I blow on it, it kicks the frame back into place.

so, its working I think.

If you block the fork, what do you mean by "nothing happens" - does it speed up, slow down, or keep constant speed? It *should* speed up, since it is getting information that indicates the disc has stopped. If it doesn't speed up, there is some problem with not having enough control range and/or not having speed trim set correctly (or possibly other circuit errors, of course). Then, when the fork pulses are applied to the chip, it should slow down and eventually lock. If you have a spare chip, you might consider substituting it. It is possible to blow out some chips if you disconnect the power supply from the chip but still have full voltages on the input pins (which I think you said you tried once).

but its working now.

but it only works right when I have the speed pot set right. If i speed it up or down with the pot, it wont lock.

oddly, but surely. I got this PLL chip from radioshack when they used to carry them. They dont anymore.

itll work, if I intensionally slow the disk down, it locks the frame right back in place. But has a tendency to get a spik from the sync pulse, and overshoot it. (frame starts rolling upwards.

More or less, i have line lock, but not so much frame lock. I covered one of the holes btw.

if i cover the fork, nothing happens, but it wont sync any more, when the frame starts to roll backwards, itll just keep on doing it.

but thats when the signals are switched than from the schematic.

If i switch them back, the motor speeds all the way up. never slows down. Even if I bring the fork closes together. Still dont matter. It only works when backwards.

One thing I do notice, if i bring the fork closer together, the disk speeds all the way up. if i bring it back, it works fine. (opposite as it should, but like i said before, i got the inputs switched).

if i switch them back to normal, it wont work at all.

The only way I know to solve a loop problem is to break the loop and measure each part.

So, I think the first thing is to measure the fork pulse going into the chip (with a scope) and see what the heck it's doing when you move it. IF it is good, moving it closer shouldn't hurt, should just make the pulse sharper (steeper sides), as it clips at zero and 12 volts.

Another part is the motor control - disconnect it from the chip output and ground that lead, and the motor should stop or slow way down; connect it to 12 volts and the motor should race. If you can connect that lead to the wiper of a pot (say 10 k), with the ends of the pot at ground and +12, you should be able to manually control the speed - and the right speed should be with the wiper approximately centered. If not, your motor control circuit is not centered and it may be impossible for the circuit to reliably control it. If it's running too fast, you need to either put in a pulley system or perhaps put a high power resistor in series with the motor. I can't tell you what the value and wattage should be without knowing how much current your motor draws.

The third thing to check is the chip operation. You can look at the output with a scope or a voltmeter. Onc eyuou know the fork pulses and sync pusles are OK, reconnect the whole circuit. When the disc is too slow (like the fork is blocked), you should see a high-duty cycle waveform on the chip output that has a DC average (voltmeter reading) close to 12 volts. When the disc is running away at high speed, you should see a low-duty cycle waveform that has a DC average close to zero volts.

When everything is centered on its nominal range and the whole circuit is connected and working, the chip output should average about 1/2 of the power suypply, or 6 volts.

I'm suspicious that your motor is just too uncontrollably fast and needs to be slowed down - If you are running the speed setting close to the "slow" end to achieve this, you may need some other fix to make the adjustment more centered. But you need to check the operation of all the parts, especially first why are you getting such strange results when you move the fork.

Congratulations!
With some "fine tuning" you can improve the line lock. I don't have a save frame lock too. This depends on bad input signals, I still have to do some work here. For frame sync I have installed a switch which interupts the sync pulse for the 4046. When pressing the button of the switch, the motor slows down and afterwards it is locked. In half of starting the disk the motor will be in correct frame sync (plus minus one line). In the other half I push the button.
What I still have is some kind of jitter. The picture in sync is moving slowly up and down. This depends on the reference pulse from the opto fork. The pulse is more a sine wave than a square-wave type.

mine are just little noisy peaks comming from the fork. They get bigger when I draw it closer together, but it never goes into clipping. It rubs the disk first.

I know the LED is drawing some current, because the 680 ohm resistor gets hotter than a fire cracker.

ok. I took some measurements.

I flipped the sync inputs back to normal.


I put my multimeter on DC, and started up the CD player, and pushed pause. and powered the circuit up. The output on the IC was around 5.6v

as soon as I pushed play, it jumped to about 12.6V i pushed pause, it dropped to 11.9, but no noticable speed change on the motor. I push play again, it goes up to 12.6 I have a b+ of about 13.5vdc

if i put the syncs backwards again, it stays at 0 volts. and then itll occasionally spike to 2.4 to try to lock the picture, but quickly falls to 0v, most likely because im using the disk as the reference, and sync as the loop signal (backwards, only way it attempts to work).


If only I had a function generator, i could pump a "fake" signal in place of where the optical fork is, and see if its the circuit or the fork.

nevermind, EDIT: I didnt have my probe grounded properly.

Anyway, even though I remove sync, or anything, the voltage still holds at 12.6vdc.

its definatly the optical fork.


now that i put a ton of capacitors to bypass the power supplys really well, the noise is gone. i get good peaks from the fork, but its not working.

If i unhook the fork from the chip, as soon as I power down the CD player, or pause it, the disk stops. when i power it on, it goes.

if i hook the fork up, it stays on, never stops.

my meter reads .9vac from the fork output, and 1vac from the sync seperator output. Can I run a fork in like an LM386 audio amplifier? or not.

I figured out the problem I think.

Ok, i got the syncs back where they are supposed to be. I built an LM386 amp, and tied it to the fork. now, the output is going into clipping, but no more speed off. now it stops as you stop the disk, and speeds up to near normal when you play this disk, itll try to lock, but it wont lock the phase. itll slow down, or speed up, but never maintain.

im assuming because its clipped now, and i have to amp up the reference sync pulse. I guess?? :scratch2:

its working now, if I cover the fork, its speeds off.

if i uncover it, it slows down, but wont lock the frame, now there isnt any stable line lock. ill be lucky if it locks for 2 microseconds.