Narrow-Bandwidth Television Association

A little off my topic but close enough.
I found this some time back and our forum should embrace the era of picture transmission before television ,if it can't do that here it has no where else really !
This is as good a place as any
All issues posted below have facsimile experiments in them hope you find this lost Gold interesting .

http://www.qslnet.de/member/vr2vrt/sstvhistory.htm

Back to 8 second Slow scan ,been busy last few days so had no time to try the follow up experiments to the cylinder tests last week .

My initial thoughts about a follow up test were to place 4 of the same image on the large lego block and replace the line mirror with that.

But having time to think about it unless i get the alignment right it might end up blurry more than likely .
Rethinking it since i have no lens system i think i need to for a test to use an image test card that is about the size of the mirrors to make up for that fact, or really a little larger as i have to take into account the distance from the mirror and use the distance of that as a zoom in or out to fill the frame .

I will then repeat the hand scan test and if successful its another step closer and i hope its 120 lines this time

Steve been looking into other SSTV software as well theres still lots in Dos a little frustrating not being able to use them ,now as i recall they worked in XP ? i don't recall using a lot of the dos software just in a dos only computer perhaps a dual boot to XP or xp on its own might be all thats needed .
I am looking into it i will grab an old lap top with xp and see what works with it .
I need to remake a Hamcom circuit for the com port then we will be cooking .
Theres lots of versions of the circuit i am pretty sure i made the JVfax version of it .

Results of the reflection test were disappointing to say the least not what i was hoping for ..
So trying to look into why....... i repeated the drawing on the cylinder test again that worked , i then replaced the cylinder with another square lego block with a drawing on one face of it only difference to the cylinder test really but the size of the drawing is half the size so i can still look into this to be sure but it didn't work at all from a test run today and a lot of head scratching as to why .
Only thing i can think of is the distance of the sensor to the cylinder is always the same same focus where as the square its an arc so this changes as the block rotates.

In a way at this point here where the electronics work but the optics via a reflection off a square mirror or a picture on a square block doesn't seem to work are giving sort of the same results at the same point where i was at in the NBTV mirror camera project .

I started this one to see if i could work out why ,WHY IS IT SO .

Again explaining the difference between cylinders and a square scanning and the problems of the latter i have drew some diagrams and why the cylinder has no problems of focal distance and angle where as the mirror or block has both these problems.

With a mirror or block the problem you have is center of the block you have the greatest distance then either end of the block you have the closest so focus has to deal with sort of an arc or really a triangle of distance as it rotates ,there is also at the same time an angle view of the block picture happening which i am not sure about but since same distance scanning via a cylinder works and this does not its pointing to me this is the problem.

Playing with a half lens it seems to keep the image lines of a drawing centered so i will try this next .

Maybe I have this wrong...but here goes...

You're having problems focussing the image from the rotating 'block', correct?

But as I think I understand it the block in time will actually be covered in mirrors, correct again?

If so you'l be focussing on the reflected image which surely be far more distant, I would have thought that this slight difference in sensor-mirror-subject distance as the mirror block rotates would have little impact on the focus.

But again I could have this entirely wrong.

Steve A.

Steve Anderson wrote:Maybe I have this wrong...but here goes...

I am wrong all the time so don't worry run the theory by me

You're having problems focussing the image from the rotating 'block', correct?

This is my speculation on what i am seeing it could also be a sensitivity thing or something else but i am thinking focus ,the distance of the image to the sensor is oscillating as the Block rotates with this experiment .
Perhaps the reason there are no square facsimile machine drums i know of .

But as I think I understand it the block in time will actually be covered in mirrors, correct again?

It looks like a circle in rotation

If you think of just one face of that square mirror rotating the distance to the sensor can be no closer than just out side the corners of the mirror other wise the thing will hit it no rotation ! so you have a set distance here only problem as its mirror turns it gets closer and further to it so its either a focus or you lose sensitivity some how you can be some distance pointing the head amp sensor to the mirrors i can see and hear its picking up the mirror just is that theres nothing discernible ,i know with the cylinder its best close but i have not really tested how far it will work before i loose the image some thing i have to test .
As i tried it just as square block with a picture on it taking the reflection problem away ,best case and it still didn't work so i just trying to work out what are the factors that make the results so different .

If so you'l be focussing on the reflected image which surely be far more distant, I would have thought that this slight difference in sensor-mirror-subject distance as the mirror block rotates would have little impact on the focus.

Thats what i would have thought to .
But distant or close i think filling the mirror area is more important image size ,i will see what my lens idea does /

But again I could have this entirely wrong.

Steve A.

Being wrong and finding out why is better than being wrong and thinking you were right,if its possible i would like to know why .
If the line mirror idea does not work out theres always some acrylic lenses i brought off ebay some time back i could come in handy here as a replacement idea.

OK today i again went back to the cylinder test but at higher speed just about the top speed this stepper will do before it stops, i couldn't do a frequency test on the light chopper circuit as i broke one of the light brakers off as you do ! have to hot glue it back on.

I also put in a main speed control pot and fine tune pot which makes life easier in this manual syncing i am using so far .

Well results are higher number of lines on the cylinder test ,so that was ok my cylinder was a bit wobberly and syncs a bit off but could make out the letters of SSTV now .

I then went back to the block test with the drawn picture ,tried the lens diddly squat ! worse than with out ,i then tried masking off 3 sides with dark paper and it works ,the drawing could be made out theres a light line in the picture due to the lego block being shiny .

This is good i can use some thing square so my theory on of focus is wrong and i am very glad !, it must be more the other square surface areas are perhaps causing the problem they are on the block, the mirror all i can do now is repeat this experiment and track it down ,may be they are not aligned or theres a wobble due to vibration mmmm we will see .
But a good day for a sstver

Some thing i forgot to mention was the print outs i posted a while back and things on my mind about the next steps .

On the sensor ,I had no problems on a large gray scale test card but no luck on the small pictures so far might just be the lines are to fine but need to look into this
don't want another encoder problem where the sensor does not pick up the ink well.

i have both a canon ink jet and brother laser printer so they are different enough to check against .

Here by the way are two other tries of the Block scanning just make out the SS of SSTV i made that picture to large hard to work it all out ,the other is a maltese cross drawing ...both not photographic quality but i will see if i can in time claw my way up the definition ladder.

I will also have to look into the other sensors to see what results they give ,i was before looking into sensitivity side but i suppose i should now look into a correct or good visible light band width and fine line focusing .

The maltese cross the lines do look thick to what i drew could be as well i am hand scanning the frame ,,that will spread or shrink the image to your scanning speed

I also notice the few accidents i have had knocking the sensor i might of scratched the lens seeing the same faults in the images .

I have a small amount of time to myself over the long weekend here (Kings birthday). So I thought I would at least have a look at a good old-fashioned CRT display for SSTV, as per 1950-something.

So I dug out one of the 3" CRTs I have had for the best part of a decade and mused if it would fit into the well and truly beat-up DG7-32 chassis. It will with minor 'adjustments', the tube is about 10mm shorter.

It's a DP7-5 which has a very similar two-stage long-persistance phosphor to the better known US P7 phosphor.

Having bought these (and other) tubes in 2005-6, it's about time I did something with them! But I don't have that much free time at the moment, we'll see...

Steve A.

Steve Anderson wrote:I have a small amount of time to myself over the long weekend here (Kings birthday). So I thought I would at least have a look at a good old-fashioned CRT display for SSTV, as per 1950-something.

So I dug out one of the 3" CRTs I have had for the best part of a decade and mused if it would fit into the well and truly beat-up DG7-32 chassis. It will with minor 'adjustments', the tube is about 10mm shorter.

It's a DP7-5 which has a very similar two-stage long-persistance phosphor to the better known US P7 phosphor.

Having bought these (and other) tubes in 2005-6, it's about time I did something with them! But I don't have that much free time at the moment, we'll see...

Steve A.

Hi Steve

Thats a good little crt i like the fact its short ,if you have any old crt you could use the another in a Flying spot camera which would not be that hard sharing the timing circuits camera monitor idea ,with the many types of light sensors these days for the light pick up that part would be easy too .
I suppose really any old Crt could do the Flying spot part better if that was a fast phosphor for that part .
An easy knock up would be a old scope for an experiment .

For a flying spot scanner a short persistance tube is better, for SSTV most used a standard P1 CRT which for SSTV is just short enough, but could be improved. If you were to do the same for NBTV a CRT with a P5, P15 or in European speak a DB-something phosphor would be required. e.g. DB7-5, DB13-14.

Most of these short persistance phosphors have a their major output in the blue part of the spectrum which is an ideal match to a PMT, say a 931A which is primarily blue sensitive.

But I don't have any short-persistance tubes as I haven't planned to make a flying spot scanner, though quite simple to do and no modulation of the grid/cathode circuit required.

At a push you could use a standard oscilloscope for SSTV, preferably one without an internal graticule.

Steve A.

Steve Anderson wrote:For a flying spot scanner a short persistance tube is better, for SSTV most used a standard P1 CRT which for SSTV is just short enough, but could be improved. If you were to do the same for NBTV a CRT with a P5, P15 or in European speak a DB-something phosphor would be required. e.g. DB7-5, DB13-14.

Yes thats good but i didn't know about NBTV needed a different type slower ? i suppose you really know your phosphors .
Off topic a bit but i recall a tv show on the inventor of Black and white tv phosphor by hes grandson and why the English went down this path and the Germans stayed with Green and white displays in the 30s.

Most of these short persistance phosphors have a their major output in the blue part of the spectrum which is an ideal match to a PMT, say a 931A which is primarily blue sensitive.

MY sensor at the moment would be no good for that to much in the IR range ,but LEDS can be used as light sensors and the colour ones are colour sensitive to the light its gives off a blue led will be sensitive to blue light ,i have only played around a little with colour with them as light sensors , solar cells would be a good but they too are less sensitive to Blue light .

But I don't have any short-persistance tubes as I haven't planned to make a flying spot scanner, though quite simple to do and no modulation of the grid/cathode circuit required.

Just an idea and electronic FFS would be much easier, but for me i do like the mechanical side of things.

At a push you could use a standard oscilloscope for SSTV, preferably one without an internal graticule.

Steve A.

On my experiments i have still had no luck on reflection tests ,i copied the block scanning with the 3 mirrors masked off but still no results i can say yes thats what i'm pointing at .
I seem to be having better luck rotating Block still not great but a bit better ,i might stick with this to get my image quality greater i might be trying to jump to far ahead with out fixing this first .

Spent the afternoon knocking together a second Head amp to test other light sensors,i didn't want to play around with the first one as i am happy with that.
A little different i put the VCO on the same board this time ,well my first test was the old MEL12 which i used a few times in the NBTV project head amps.
After many gos this out of wack framing was the best picture i managed to get from a cylinder test and thats the easiest test, it was near on touching the cylinder ,its hard to focus might need a lens with it .
No wonder i was getting no worthy results last time around .
I will try some other different sensors with it and see what they do later in the week.

Had some more time this afternoon and gave a few ideas a go .

On the light sensor ,tried a dome sensor an LDR and tried a tiny lens with them but no luck the little lens just lets a small amount of light to one small area of the sensor perhaps some acrylic lenses i have might be a better idea to try on those .

So really back to using the original light sensor or its type i should say as i have a few few of those it gives the best results so far .

I then have looked into using more than one of the surface areas for block scanning, i know now doing this test i am off the correct line speed, i can read more than one face and 2 faces look to be superimposed over the other 2 faces of the block ,you may notice this in some of the photos and gif animations .

My last test today was reflecting the block to a mirror and trying to read the scan off the mirror ,best i got was just seeing the frame with no image and that was having the mirror as close as possible so till this is worked out no mirror part of it will happen ,its much harder than a monitor laser idea or a FSS .

I will keep on trying and i am seeing now what is and isn't possible ,well worth swapping to SSTV to test this .

Looked into trying to use a print out to day and did a block scan on one of these ,old felix has alway been good to me in sstv so gave him a go again.

Well as you see in the slide show i got some thing out of that picture ,i tried different light levels and distances the dark one is as close as i could get .

So i think i am on the right track in looking into the focusing or detail part ,that needs to be improved to get any further .

In one of the news letters Alan Short got rid of hes sensors lens to fix the same problem,and i also have not tried a mask .