NBTV camera project

[aussie_bloke]

G'day all.

Alongside the NBTV-scope project I have recently been working on constructing a NBTV camera myself and have made one using parts from around the house. I used the following materials:
CD player DC motor with CD mount (from a scrapped car CD player)
Black base disc from bottom of DVD spindle (for Nipkow disc)
10x CD case for enclosure
C-mount lens
LDR Light dependant resistor (for image pickup)
Motor speed control circuit from my first mechanical televisor
LDR circuit

Anyhow using the NBTV template I drilled 1mm holes into the black CD sized base disc and I glued the disc to the motor's CD mount. The CD case I sanded down and spray painted it matte black. I drilled 3 holes in the case, one for the C-mount lens at front, smaller hole at back for image pickup device to collect the light and a big hole back centre to mount the motor on. I soldered the motor to a bracket and mounted it inside the case and secured it with screws and superglue. I also mounted the C-mount lens to the front with superglue as well. The LDR, considering it was smaller than the scanned area I decided (not knowing if it would work or not) to surround it with chrome cellophane to hopefully reflect light from the outer areas to the LDR and have mounted it inside an old TV tuner rotor and glued it to the back where the light passes through the holes of the disc. The LDR circuit I used from the one Harry presented in his NBTV camera project http://www.taswegian.com/NBTV/forum/fil ... 00_656.jpg and for the motor circuit I used the one from my first mechanical televisor as I was eager to get it done ready for testing.

Anyhow tested the camera out with Gary's NBTV The Big Picture software and as expected with my slaptogether setup I ain't getting a picture, but the good news is there is at least a signal response when shining the torch down the camera lens to the LDR image pickup. Gary commented on one of my videos saying I must calibrate the NBTV program before testing which I did using a NBTV converted video and then run the camera and still haven't been able to get the motor speed right for a smooth flowy video signal, am still getting stuck lines and such (too many words to explain it all).

I understand there are a number of issues with my camera setup:
A: I have no idea what speed I'm running the Nipkow disc at as I'm not using a sync circuit that can set the disc to a constant NBTV speed
B: The lens is non-rackable as I don't have the materials to make it rackable for correct back focus so hence am unsure whether I am even getting a properly focused image on disc or not, plus I am not even sure if the lens is perfectly aligned with the scan area and LDR at back nor am I even sure if the LDR at the back is perfectly aligned either
C: I wonder if the LDR circuit is even effective and sensitive enough to pick up light variations like dome sensors, phototransistors and other image pickup devices
D: The LDR being smaller than the scanned area and using chrome cellophane around it to reflect light from the outer regions of the scanned are, I am not sure if it has worked at all or worse scrambled the light pickup

I am going to soon try using a phototransistor circuit setup for image pickup and see how that goes and a sync motor speed control to make the Nipkow disc spin at constant NBTV standard speed, hopefully that will yield results though I wonder about the lens being the correct distance or not away from the disc being in its fixed position.

Should that method fail, it won't be a total loss because I can easily convert the camera into a mini CD case NBTV Nipkow disc televisor by replacing the image pickup with a LED bank light box and lens with a magnifying glass and there I'll have a mini NBTV set!;)

I have video documented my construction process in a 7 part video series on my YouTube channel, the videos can be viewed in the following links:
Part 1 http://www.youtube.com/watch?v=j0Gn3VSr_wQ
Part 2 http://www.youtube.com/watch?v=baHgZXjUuv4
Part 3 http://www.youtube.com/watch?v=BUYdSLGTljk
Part 4 http://www.youtube.com/watch?v=pxVphfOLTng
Part 5 http://www.youtube.com/watch?v=TTGGOSZVl1U
Part 6 http://www.youtube.com/watch?v=yguYWDtOrio
Part 7 http://www.youtube.com/watch?v=3xn8jQOEaX4

And here below are images of my camera from construction stage to completion:

[gary]

Gary commented on one of my videos saying I must calibrate the NBTV program before testing which I did using a NBTV converted video and then run the camera and still haven't been able to get the motor speed right for a smooth flowy video signal, am still getting stuck lines and such (too many words to explain it all).

But Troy, did you also use the alpha version of TBP AND turn off sync processing? (I noticed in your videos that you were originally using the release version).

I am not sure if you can easily remove the sensor now (if it's glued on) but one way of focusing the lens is to temporarily remove the cd disk and replace with some thin white paper of anything that is semi-transparent and then view your subject from the rear adjusting the lens until a sharp picture is obtained, i.e. a kind of rear projection.

One concern I have is the lack of a condenser lens behind the disk to focus the scanned image onto the sensor - I realise that you have a kind of reflective lens (the mylar or whatever the shiny stuff is) surrounding the sensor but I am not to sure about it's effectiveness as there is very little light available to begin with. You normally have a very short focus lens (or two) and you temporarily remove the disk and replace the sensor with some white paper (as above) to ascertain the position where the entire image falls on the sensor. Certainly with your intention of trying a phototransistor I fear you would need a compound lens system to achieve this.

The optics of this type of camera is (well at least for me) quite fiddly and difficult - the main reason I suggested Harry go FSS - much simpler but still with it's own set of problems as Harry is finding - but he has obtained a picture!

Again I would be remiss to not mention that the overly large 1mm aperture size will severely limit the resolution of your picture should you eventually obtain one (and I am sure you will!). Cheers.

[Klaas Robers]

Troy, read the part about optics and lenses in the NBTV-handbook. You may also read that short chapter on the website of the NBTVA: www.nbtv.org/optics.htm The important part for you is about the condenser lens.

Then read also the chapter about spherical aberation (next chapter) and you will understand why a condenser lens is made out of two flat-convex lenses. In between the lenses the light should be in parallel.

I agree with Gary that you will get a very unsharp camera. For a CD the apertures should be a small as 0.22 mm.....

On the other hand you are lucky to have more light than if you had apertures this small. It will be 20 times as much!!

[gary]

And of course I meant "condenser" lens not "objective" lens in my post above. I'll correct that. Thanks Klaas.

[Harry Dalek]

Hi troy

Good luck with yours you can copy the speed control stuff from mine all that would be the same .

I have an ldr 1 transistor circuit next to my dome sensor circuit but never tried it as yet been happy with the solar cells on the second sensor i made .

Must say your pretty quick knocking these things together ,i have found the flying spot camera very challenging more so than any monitor i have ever tried..its not as easy as i was thinking when i started but with Garys advice help we have got it working ...now just trying to improve what i have .

I would make a good light sensor circuit first ..and you have to make your optics adjustable may be use some white paper behind the lens need to focus what ever to the area of your nipkow scan size or you either scanning part of some or out focus...

Good to have another Aussie doing a camera ..keep trying TILL IT WORKS !!!!

[aussie_bloke]

Thanks all for the advice. Will definitely read up on the optical setup of a NBTV camera on the site, am looking at it now, am gonna have to round up some condenser lenses from somewhere, Ebay obviously. Anyhow this camera was a quick whip-up project but now I am going to give it more careful thought and refine it, I think the lens being non rackable is the biggest disadvantage of this camera as well as the disc having 1mm holes but am gonna see what I can do to build another camera with rackable lens and the right optical setup as given on the NBTV site. It's now a matter of getting the materials.

I pretty much am a big enthusiast of vintage style TV cameras and to build a working mechanical camera will be a treat so am sticking to my guns on this project. Going off topic this isn't the first TV camera I've constructed, I have previously constructed a 1970s style B&W vidicon TV camera which is a big frankenstein mass of electronics mounted on a piece of wood and believe it or not it actually works I have featured my frankenstein masterpiece on my site on the following pages:
http://www.troysvintagevideo.0catch.com/mycamera.html
http://www.troysvintagevideo.0catch.com/mycameraa.html
I also am anticipating on one day constructing an image orthicon camera and have a few IO tubes I bought off Ebay. Those interested in seeing my other projects and camera/TV/VTR collection most of it is on my site http://www.troysvintagevideo.0catch.com/index.html .

Anyhow will take all advice into account and see what I can do with this camera, otherwise it will become a mini televisor and I will work on construct a completely new NBTV camera.

[aussie_bloke]

Just a quick little update on the NBTV camera, I have read about the lens theory on the NBTV site tutorials and have been communicating with Harry on his NBTV camera project and have promptly decided to opt to use a solar cell as an image pickup device opposed to phototransistors or other small photosensitive devices due to the fact condenser lenses are not that easy to get a hold of but hopefully I will pick one up in the not too distant future. I admit using chrome cellophane around the LDR was a bit of a dumb idea as the light would bounce all over the place but I went ahead with it anyways for the hell of it with the remote possibility it might work haha. Anyhow Harry explained to me with solar cells having plenty of surface area that covers the scan area of the Nipkow disc eliminates the need of the condenser lens to focus the light onto a small area. Anyhow till I can find me a solar cell or buy one the camera project is currently on hiatus, but have a couple of other exciting projects I am working on in the meantime which I'll soon post.

[Harry Dalek]

Just a quick little update on the NBTV camera, I have read about the lens theory on the NBTV site tutorials and have been communicating with Harry on his NBTV camera project and have promptly decided to opt to use a solar cell as an image pickup device opposed to phototransistors or other small photosensitive devices due to the fact condenser lenses are not that easy to get a hold of but hopefully I will pick one up in the not too distant future.

Hi Troy
Reading the newsletters most used solar cells..but no reason not to use the others..we can make one together the solar cell i am using came from a jaycar toy solar car ..i have others but they are to big for what i want it for.

I admit using chrome cellophane around the LDR was a bit of a dumb idea as the light would bounce all over the place but I went ahead with it anyways for the hell of it with the remote possibility it might work haha.

Never know till you try i did read that one person thinks the ldr works better than the solar cells for a camera .

I suppose using a LDR using a diffuser size of the scan area so the ldr sees all of the scans as light flashes over all the area...so same idea in reverse for LEDS sort of ,but with the low light levels i do think your better off with a lens of some sort ...

Anyhow Harry explained to me with solar cells having plenty of surface area that covers the scan area of the Nipkow disc eliminates the need of the condenser lens to focus the light onto a small area.

Well it works for the flying spot idea so i see no reason it seeing the same thing behind a nipkow disk as it works in front .


Anyhow till I can find me a solar cell or buy one the camera project is currently on hiatus, but have a couple of other exciting projects I am working on in the meantime which I'll soon post.

Do you have an old calculator with solar power ? if so time to make it into a camera

[Klaas Robers]

The problem of a solar cell is its capacity (in nF). Due to the high capacity the solar cell is rather slow. That will attenuate the higher frequencies.

Any way apply a DC voltage to the cell in the polarity that it is not conducting (in the dark). That will make the capacity lower (it is a varicap as well) and will make the efficiency also higher. I don't know what voltage is possible for a normal solar cell, because it is a zener diode as well, but I think about 10 volts. the dome sensor can withstand several tens of volts.

I don't know the Pedro circuit diagram, I have to find that back in the Newsletter. It is important to know the DC voltage and the input impedance of the amplifier. Heavy negative feedback to the input can help.

On the other hand I don't like the metal lines on the surface of the cell. If you place the cell close to the disc (as close as possible) then you will see the lines in the picture you make. Remind that the picture is focussed on the surface of the disc. If that is where you place the detector close to, the picture is also sharp at the detector surface.....

[Klaas Robers]

I found the Pedro circuit diagram in Newsletter 26/3. I don't like it because
- there is no DC voltage over the sensor,
- too many variable resistors, that is not necessary in a well designed circuit,
- I don't trust the feed back cicuit.

After you have made a sensor and amplifier, test it in this way:
- Make a light tight box (shoe box).
- place the sensor with or without amplifier in the box,
- place a (red) LED in the box such that it can illuminate the interior of the box,
- modulate the LED with a square wave of 10 Hz to 20 kHz,
- and watch the output of the amplifier on an oscilloscope.

For a 32 line camera the output signal should be constant in amplitude (100%) until 10 kHz. For more lines and/or higher frame rates it should go higher.

Then decrease the LED current and see how sensitive your pick-up system is. You should see the noise of the amplifier. Steve has done this for a photo multiplier tube. Very educational.

[Steve Anderson]

With all due respect to Pedro I too find that pre-amp in 26/3 'messy' and agree with Klaas's points.

Biasing the sensor is a simple fix, wire the 100k pot across the sensor to the positive supply rather than ground. Make sure the supply is very clan and well decoupled/bypassed. or make provision for such. Or use a 9V battery, no current to speak of will be drawn, but make sure the sensor is the correct way around.

The first stage only has one pot for gain control, but I agree the biasing is hit-and-miss. It will vary with the individual specimens of the trannies as well as with temperature.

The second stage has some 'lift' starting at mid-frequencies (C2 and the 500R pot), I wouldn't have called it a 'contrast enhancer'. The biasing for the third transistor is a real no-no. The Hfe of a transistor is not a well defined parameter (unlike tube/valve gm) and will vary considerably. The specification for a BC109C is 420-800 (Philips) hence his suggestion in the last line of text of fiddling the value of R1.

I'm not sure of the function of the LDR, some form of AGC?

In fact, staring at it longer I don't think the second stage will work at all (I'm often wrong though). Imagine removing the LDR and its 4k7 pot then assume the transistor is fully off and the 100k pot is at maximum resistance. Voltage at point 'B' will be 2.75V with a +5V supply. Turn transistor on, replace the LDR/pt and that voltage will be somewhat less, perhaps less than half of before. I don't see how it could work unless there's a typo in the schematic as happened to Troy in a different thread.

I don't know if anyone else has built this but if someone has your opinion would be useful.

Steve A.

[Harry Dalek]

The problem of a solar cell is its capacity (in nF). Due to the high capacity the solar cell is rather slow. That will attenuate the higher frequencies.



Any way apply a DC voltage to the cell in the polarity that it is not conducting (in the dark). That will make the capacity lower (it is a varicap as well) and will make the efficiency also higher. I don't know what voltage is possible for a normal solar cell, because it is a zener diode as well, but I think about 10 volts. the dome sensor can withstand several tens of volts.

I don't know the Pedro circuit diagram, I have to find that back in the Newsletter. It is important to know the DC voltage and the input impedance of the amplifier. Heavy negative feedback to the input can help.

On the other hand I don't like the metal lines on the surface of the cell. If you place the cell close to the disc (as close as possible) then you will see the lines in the picture you make. Remind that the picture is focussed on the surface of the disc. If that is where you place the detector close to, the picture is also sharp at the detector surface.....

Hi Klass I have the positive to ground as in this version of a head amp..

I have made 3 working ones now ..pedros seems to me to have a little less amplification..but i put that down to it working off 5volts my others run off a 9volt battery .

Yes had been wondering about the solar cell lines on the one i am using well i will have a think about that perhaps i should just defocus it to the cell a bit so its not missing light pick up in those line areas ...i will play around with it . there are cells without these btw.

[Harry Dalek]

Hi Klass

I found the Pedro circuit diagram in Newsletter 26/3. I don't like it because
- there is no DC voltage over the sensor,
- too many variable resistors, that is not necessary in a well designed circuit,
- I don't trust the feed back cicuit.

The feed back does not seem to work changing the resistance does nothing so far testing ...i will try collector to base feed back and see.

For R1 put in a switch to switch between 2 and 4 meg resistors theres sure is a gain change happening ...this might be one area a pot would be better to see its limits .

I don't think i have seen a head amp with this yet klass do you have an example talking about the dc over the sensor.

After you have made a sensor and amplifier, test it in this way:
- Make a light tight box (shoe box).
- place the sensor with or without amplifier in the box,
- place a (red) LED in the box such that it can illuminate the interior of the box,
- modulate the LED with a square wave of 10 Hz to 20 kHz,
- and watch the output of the amplifier on an oscilloscope.

Oh i have done this sweep test on the other 2 head amps .. Gary sent me a sweep wav file ...the solar cell was not as good as the dome sensor in band width in the high end but they are still workable ...
I will try it on this one .

For a 32 line camera the output signal should be constant in amplitude (100%) until 10 kHz. For more lines and/or higher frame rates it should go higher.

Ok thats something i didn't know i will keep that in mind.

Then decrease the LED current and see how sensitive your pick-up system is. You should see the noise of the amplifier. Steve has done this for a photo multiplier tube. Very educational.

Ok that sounds like a good test i sort of tried similar using distance .

[Harry Dalek]

With all due respect to Pedro I too find that pre-amp in 26/3 'messy' and agree with Klaas's points.

Poor Pedro

It works may have to many pots ...i am not sure the circuit is he's i think its a rework but i had all the parts gave it a shot .

Going more about what he says it does than how much i liked it i have made it ..i didn't put in the ldr to try as didn't see the point of that needed yet .

Biasing the sensor is a simple fix, wire the 100k pot across the sensor to the positive supply rather than ground. Make sure the supply is very clan and well decoupled/bypassed. or make provision for such. Or use a 9V battery, no current to speak of will be drawn, but make sure the sensor is the correct way around.

Ok Steve thats some thing to try ..found listening to garys advice better run these things off battery for the start any way...
why is it the other dome sensor circuit club circuit is to ground as well..

The first stage only has one pot for gain control, but I agree the biasing is hit-and-miss. It will vary with the individual specimens of the trannies as well as with temperature.

I m using BC108's had some of these ones handy scrapped from something in the past .

The second stage has some 'lift' starting at mid-frequencies (C2 and the 500R pot), I wouldn't have called it a 'contrast enhancer'.

I was thinking he had had some control over this reason i liked it .


The biasing for the third transistor is a real no-no. The Hfe of a transistor is not a well defined parameter (unlike tube/valve gm) and will vary considerably. The specification for a BC109C is 420-800 (Philips) hence his suggestion in the last line of text of fiddling the value of R1.

I put in a 2 way switch to switch between the 2 and 4 meg resistors i put it in to see for my self ......the gain is better on one setting for sure but i haven't looked yet which resistance it is ..

I'm not sure of the function of the LDR, some form of AGC?

I was thinking its placed or directed to the subject as what you say a AGC ...i didn't bother with it even with my limited know how one gain pot it enough .....i am not fussed .

In fact, staring at it longer I don't think the second stage will work at all (I'm often wrong though).

I does work how well if it were designed correctly i am not sure but i copied it and it works .

Imagine removing the LDR and its 4k7 pot then assume the transistor is fully off and the 100k pot is at maximum resistance.

Seems a bit strange looking at it now you have i think a pot where the 82k on the collector might be a better idea .

Voltage at point 'B' will be 2.75V with a +5V supply. Turn transistor on, replace the LDR/pt and that voltage will be somewhat less, perhaps less than half of before. I don't see how it could work unless there's a typo in the schematic as happened to Troy in a different thread.

MMMM well i didn't like the ldr so didn't bother but i can test it how it is .

I don't know if anyone else has built this but if someone has your opinion would be useful.

Steve A.

[/quote]

Perhaps its only me ? i made it only because i believed hes results perhaps the circuit needs improving ..i was thinking it was a past version of the clubs dome sensor circuit .