Page 5 of 11

PostPosted: Fri Nov 02, 2012 9:31 am
by Harry Dalek
M3DVQ wrote:
harry dalek wrote:also not sure how long it would last without a cover


Not long at all...

That's what's so good about using the junctions of an EPROM. The quartz window keeps all that nasty oxygen out :D


I do have a few of those they are just UV sensitive ? i suppose good for a out door camera the sun baths us with that as well.. uv lights are easy enough to get as well.

Perhaps also the guy doing the power transistor may be if you filled it up with resin it might fix the oxygen problem one guy was bathing it in a chemical and then stuck a clear lid on it .

PostPosted: Fri Nov 02, 2012 10:18 am
by gary
Of course if one is going to go to the trouble of taking the top off a device one might as well make that device a dynamic ram chip - that way you can do away with the scanning device (Nipkow disk) altogether - see attachment.

Their were a number of variations on this theme used in the years before CCD cameras became very cheap.

PostPosted: Fri Nov 02, 2012 10:22 am
by gary
Viewmaster wrote:As I am building a NBTV televisor without any moving parts how about a NBTV camera without any moving parts too?

First thoughts are a 48x32matrix of light sensitive cells addresed in a similar way to my Niptrix.
BUT the cost would be enormous as 1532 cells would be required.
Anyway to reduce the number?


See my previous post...

PostPosted: Fri Nov 02, 2012 11:21 am
by M3DVQ
harry dalek wrote:I do have a few of those they are just UV sensitive ?


They require high energy UV to erase the memory, but using them as a crude photo-sensor they seem to work at red and green at least.

Note that this isn't using the device like a CCD as with gary's DRAM suggestion. Just the voltage developed across all the junctions connected to the power rails.

PostPosted: Fri Nov 02, 2012 12:33 pm
by Harry Dalek
M3DVQ wrote:
harry dalek wrote:I do have a few of those they are just UV sensitive ?


They require high energy UV to erase the memory, but using them as a crude photo-sensor they seem to work at red and green at least.

Note that this isn't using the device like a CCD as with gary's DRAM suggestion. Just the voltage developed across all the junctions connected to the power rails.


I have to have a good look a garys post.
But yes thats good they work on the 2 colours i was thinking just uv...they have a nice large surface area ,when i get back to a mechanical camera i will play with them i have a few doing nothing /

PostPosted: Fri Nov 02, 2012 4:33 pm
by gary
Just as a camera benchmark here is a video of Graham's marvellous results - Wave file version available on request.

PostPosted: Fri Nov 02, 2012 4:55 pm
by Harry Dalek
gary wrote:Just as a camera benchmark here is a video of Graham's marvellous results - Wave file version available on request.



Thats pretty good benchmark gives us an idea an idea to work towards so its not limited to closeups.. :!:

PostPosted: Fri Nov 02, 2012 9:49 pm
by Panrock
That's impressive, Graham!

Steve O

PostPosted: Fri Nov 02, 2012 11:33 pm
by M3DVQ
harry dalek wrote:But yes thats good they work on the 2 colours i was thinking just uv...they have a nice large surface area ,when i get back to a mechanical camera i will play with them i have a few doing nothing


Yup, that was the idea. The call went up a couple of years back at the convention to find a replacement for the dome sensor so various people have been trying to find other devices with a large sensitive area.

How much of the area of the chip is contributing to the voltage I don't know, it clearly needs more research but I'm sure someone reported a good frequency response. The only issue I can see is that because of the huge size of the package and all those unconnected i/o lines it will be a nightmare keeping interference out of the signal.
This was what ended my experiments with it because I didn't have the time or the components to isolate everything properly.

Also, you have to do the tests in a light proof box otherwise you get a massive 50Hz signal off incandescent lights, or you turn that off to work in the dark and get a feedback loop from the detector picking up its own oscilloscope trace :shock:

It was all just a bit too much effort for me since I'm not building a camera :oops:

PostPosted: Sat Nov 03, 2012 2:40 pm
by Harry Dalek
M3DVQ wrote:Yup, that was the idea. The call went up a couple of years back at the convention to find a replacement for the dome sensor so various people have been trying to find other devices with a large sensitive area.


Its a good idea i have looked at them before but didn't click to use them for NBTV camera i just knew the uv cleared them so didn;t think much more about it now i will keep them safe for a future try .

How much of the area of the chip is contributing to the voltage I don't know, it clearly needs more research but I'm sure someone reported a good frequency response. The only issue I can see is that because of the huge size of the package and all those unconnected i/o lines it will be a nightmare keeping interference out of the signal.
This was what ended my experiments with it because I didn't have the time or the components to isolate everything properly.



You can use software such as Winscope which is a pc sound card scope and a spectrum analyzer ......modulating a sweep frequency to an Led from PC sound card point it at the device connected to your head amp and use another pc to record or play live to the spectrum analyzer...you can see how good the bandwidth is ..i did it on testing the dome sensor and solar cell .

yes must be in a light proof box Or easier way in a dark room as i did at night for my tests ...tv must be off too or your flying spot scanning your plasma.

It was all just a bit too much effort for me since I'm not building a camera :oops:


I will for sure look into it when i get back to a mechanical camera but there are easy ways to do the testing.

PostPosted: Sun Nov 04, 2012 2:25 pm
by aussie_bloke
G'day all. It's been a little while since I've last updated you all on my camera/monitor project.

Got some awesome news. Gary kindly sent me a dome sensor and I built the head amp for it going by the NBTV handbook circuit http://www.nbtv.wyenet.co.uk/kit31.gif but using other transistors close to equivalent due to not having those transistors. I rigged it all up and SUCCESS, it WORKS!!!:D

I initially had the dome wired reverse polarity so got a negative picture but then switched it around and got me proper picture positive NBTV images. I have since been filming what I could think of that would look good on camera and have done nearly 2 hours worth of telerecordings of my camera in operation using my trusty 1990s Panasonic RX3 camcorder to film off the monitor! :D

The sensitivity of the dome is pretty good, which I don't need to shine the 150W lamp in close proximity of the subject for it to pick up light which is great!:D

Anyhow I have cut some small 5 second DVD quality MPEG snippets of some of my telerecordings to upload to the forum, the bigger clips I am uploading in a zip file to DepositFiles, will post the link when the zip is uploaded, there's about 475mb worth as they are all DVD quality MPEG. Here below are pics of the modification to my camera and test subjects I placed in front of the lens and the snippet telerecordings:

PostPosted: Sun Nov 04, 2012 5:20 pm
by gary
Well that's a really good start Troy - well done!

It's notable however, that you are as yet only getting silhouettes (as opposed to grey scale as you should be getting). Now this could be overloading in your head amp, or something in your monitor LED driver.

If you could post some of your raw signal recording I should be able to tell you one way or the other (of course it could be both ;-))

In any case it's only a matter of "tweaking" I'm sure.

PostPosted: Sun Nov 04, 2012 6:00 pm
by Harry Dalek
Really good results Troy !
I found the head amp was fussy with the voltage to it try it with a variable regulator ..i used transistors i had handy as well so may be that as well...i am pretty sure i made any feed back resistor adjustable in all the head amps i made having a look at your picture looks like you have done that .
Aussie aussie aussie

:wink:

PostPosted: Sun Nov 04, 2012 7:11 pm
by aussie_bloke
I made the head amp as close as possible to the original NBTV handbook schematic, problem was the transistors are not ones easily available from your local Jaycar store so I looked for what I think would be suitable substitutes, so for the BC109C I substituted a BC549, the BC212 I substituted a 2N3645 and the 2N3704s I substituted BC238Bs, going by the data given in www.alltransistors.com most of the specs were within close proximity to the original transistors of the schematic but had lower forward current transfer ratio, would lower hFE inhibit the performance of the head amp???

One thing I noticed with the dome sensor is it is very sensitive to the point my 150W lamp was shining too much light on the subjects I filmed so the subjects did look to be overbloomed and silhouettish but have to say too much of the test subjects were of silhouette pictures. I found there was some variable shading when I stuck my hand in front of the lens at a further distance away from the 150W lamp. Also I found that when I had the lamp not directly pointing to the subject to be filmed I get better shading from it. Also another factor is my Panasonic RX3 camera being very sensitive to low light does intensify the lighting from the monitor whilst the monitor produces a dim amber picture to the naked eye. And in post production I did upped the brightness and contrast to make the pictures more viewable, the first couple of clips I did put too much contrast but the rest I adjusted contrast to a suitable level below clipping. And finally the holes are drilled to 1mm which does lessen the finer detail in the picture as well opposed to drilling with a needle of half the diameter or less.

Anyhow now that the camera is working and making usable pictures I will have to start making refinements, lock the lenses in place and drill another record disc as I did a bit of a bodgy job drilling the one I am currently using.

I have now uploaded a bunch of 20 second video clips of the camera telerecordings in a zip file to DepositFiles, here is the link to download http://depositfiles.com/files/k1f2d2hqc it's about 475mb in size.

Also here are a couple more 5 second snippets:

PostPosted: Sun Nov 04, 2012 7:26 pm
by aussie_bloke
Here are a couple of wav files I made from the camera, one of the NBTV logo and one of my hand, they are recorded at 32kHz and video is on the right channel. I don't know what speed I am running my camera at so the RPM speed could be anything, all I know is I have the motor at max speed running off 12V, I am yet to refine my speed control circuit as I am using the original speed control circuit from my very first televisor which I used for all my other televisors that use DC motors, gee I'm lazy haha :lol: . I tried playing them on NBTV the Big Picture but can't get it to synchronize and produce anything recognizable. Anyhow Gary, see what you can extract from these files ;) .