Televisor kit

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Postby PatchY » Sun Jun 15, 2008 10:33 am

Hi with the modifications here, this is my result using Gary's fine converter software

http://www.youtube.com/watch?v=UcWW4jRLtPo
PatchY
Laboratory Assistant
 
Posts: 36
Joined: Mon Aug 20, 2007 10:45 am

update

Postby LuTELLO » Mon Jun 16, 2008 10:24 am

update: Cheap pcb + poor soldering skills = potential disaster, but I did it. The new pots are clearly better, but I 'm still having sync problems. Some videos are pretty stable, and some are pretty bad, all made with video2nbtv. Anyone have any tips?

I replaced the LED but I can't seem to find something to use as a diffuser anywhere. It's almost not worth using with the included one.
User avatar
LuTELLO
Research Scientist
 
Posts: 42
Joined: Sun Dec 23, 2007 4:06 pm
Location: Town

Postby gary » Mon Jun 16, 2008 12:33 pm

It's a tough one, I have an MUTR Televisor and it has proved to be very reliable in terms of sync, but others have reported similar problems to you. The only thing that I have found to be critical for sync is the position of the disc in relation to the sensor, but I suppose that's already been gone through. On mine, I found that the optimal position occurs when the motor spindle *just* pokes through the hub (gear).

The only other thing I can think of is to look at the signal (does the inbuilt test page sync reliably?). What is your source? It is feasible for this to be too high or too low for effective sync slice. Certainly the fact that you are having variations from source item to item. One possible reason for this is that when a source is AC coupled the DC content is lost. This in turn causes the relationship between sync tips and programme material to vary depending on the programme content. This can make it impossible for the sync circuitry to distinguish between sync and programme material resulting in sync loss. An example of this is attached. You may find you can improve this situation by adjusting the level on your source device.

The choice of diffuser material is one I would like to see a complete thread on. After over thirty years trying I have never come up with a material I have been happy with, whereas others seem to have done so. It is an eternal compromise between diffusion and brightness, you get it right one way and the other suffers. Materials that seem to be easy to acquire in one part of the world never seem to be so in your part of the world, for instance, some say ground glass is the best, well I have never even SEEN ground glass never mind obtained any. Once I tried to make my own by using valve grinding compound, worked for others, never for me, sheesh. Some members recommend using draughting paper (vellum), I used this all the time years ago when I worked as a draughtsman, do you think I can find any now? not on your life. I currently use a piece of white perspex which diffuses well but drops the brightness to an extent that I have to view in a darkened room (even though I have modified the LED). Previous to that I used "grease proof paper" (you might know it as "baking paper") which, while not being perfect, is considerably better than the diffuser supplied with the unit.
Another possibility is "flexible chopping board" material, you can sometimes find these in $2 shops, but it varies a lot in colour, patterns, etc., obviously you try to find a white, non-patterned variety.

I'm sure others may have other ideas, I for one would be glad to hear them.
Attachments
sync slice.jpg
sync slice.jpg (13.22 KiB) Viewed 19241 times
gary
 

Postby AncientBrit » Wed Jun 18, 2008 10:37 pm

re diffusers.

One material I've spotted but never used is a plastic milk bottle. Not sure whether bottles outside the UK use the same material, but ours is semi-transulent.

Another tip I saw on this forum, from Klaas I think, is to use two thicknesses of diffusers but with a small gap in between to improve the spread.

Worth a try, but of course at least twice the loss,

Regards,

Graham
AncientBrit
Green padded cells are quite homely.
 
Posts: 858
Joined: Mon Mar 26, 2007 10:15 pm
Location: Billericay, UK

Postby bigscreen » Thu Jun 19, 2008 2:14 am

A very good alternative that I use, is this tape that You can buy in most
stores.
It is very thin and it difussed great when sticked directly over the holes,
so the lightsource dont need to be covered.
One layer is normal good enough, but if not, stick two layers on eachother
and there you got your diffusser....
Attachments
tape.JPG
tape.JPG (17.64 KiB) Viewed 19195 times
User avatar
bigscreen
Just nod and pretend you understand me
 
Posts: 228
Joined: Sun Jul 01, 2007 7:52 pm

Postby AncientBrit » Thu Jun 19, 2008 5:12 pm

Nice idea, Bigscreen. Easy to apply as well.


Graham
AncientBrit
Green padded cells are quite homely.
 
Posts: 858
Joined: Mon Mar 26, 2007 10:15 pm
Location: Billericay, UK

Postby LuTELLO » Fri Jun 20, 2008 4:13 pm

I just found a good diffuser. It's, well, a diffuser from the light source of a broken scanner. Thanks for the tips though. I may need them someday.

One of the biggest problems I have with the sync is that it seems to be the most stable when I adjust it so that the vertical framing is at it's worst!
So how do I fix that? (It works when adjusted properly, but not as reliable.)
Attachments
P6190001.JPG
P6190001.JPG (23.35 KiB) Viewed 19143 times
User avatar
LuTELLO
Research Scientist
 
Posts: 42
Joined: Sun Dec 23, 2007 4:06 pm
Location: Town

Postby LuTELLO » Fri Jun 20, 2008 5:02 pm

Btw, I just tried running 25fps on the MUTR by doubling the sampling rates of my video2nbtv files. Not only of course does the picture have better motion and less flicker, but the sync is usually better! Too bad this doesn't work on cassette recordings.
User avatar
LuTELLO
Research Scientist
 
Posts: 42
Joined: Sun Dec 23, 2007 4:06 pm
Location: Town

Postby lemke » Mon Sep 29, 2008 12:47 am

DrZarkov wrote:With the original LED it is not very bright, with the white LED it is quite o.k., for my televisor. The picture is very small, I use two lenses, the original one, and a "Gameboy" lense from a first generation Gameboy. That makes it quite watchable.


I received my televisor yesterday, and put in a white LED, but I am disappointed in the LED's brightness. I looks almost yellow - I used Radio Shack's 276-320 LED rated at 1100 mcd, 3.6 volt, 20 mA. Given the cost of the LED, ($5 plus US), I was expecting much better brightness.

EDIT: I should have done my homework. that rating (1100 mcd) is not really that bright, compared to what I have found available on the web.

What ratings are you all using for your televisor?
User avatar
lemke
Laboratory Assistant
 
Posts: 26
Joined: Sun Aug 10, 2008 10:39 am
Location: Seattle WA USA

Postby Panrock » Mon Sep 29, 2008 5:01 am

Think I used a 3000 mcd. I was very careful to solder it in only at the far ends of the leads, with heat shunting. The performance of these things can easily be reduced by using too much heat.

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby lemke » Mon Sep 29, 2008 10:45 am

Panrock wrote:Think I used a 3000 mcd. I was very careful to solder it in only at the far ends of the leads, with heat shunting. The performance of these things can easily be reduced by using too much heat.

Steve O


Thanks, Steve.

It just occurred to me this morning: "why not just go to the hardware store, and buy one of those disposable keychain LED flashlights?"

I don't know the rating, but I put in in the LED from the flashlight in the Televisor, and am very happy with the brightness now. The added bonus is that is was only a few bucks.


John
User avatar
lemke
Laboratory Assistant
 
Posts: 26
Joined: Sun Aug 10, 2008 10:39 am
Location: Seattle WA USA

Postby chris_vk3aml » Mon Sep 29, 2008 11:32 am

Hello from Australia ---

Be careful when assuming that the number of milliCandela gives you
an absolute measure of the amount of light emitted by the LED chip.
Often, it does not. The fact is that different LEDs have different radiation
angles, an effect varied by changing the distance of the chip behind the
surface of the LED's encapsulent lens. In this way, the same given source
chip might have only 1000 mCd measured axial intensity in the far-field
with Lambertian (cosine-law) radiation; that is with no effective lens in
front of the chip, or with the chip mounted at the centre of a
hemispherical dome.

If, however, the LED chip were placed further back from its encapsulent
lens - and some LEDs do this - the radiation half-flux angle could be
reduced to, say, 5 degrees, and the far-field milliCandela measurement
might go to 20,000 or even 40,000 mCd. A few suckers might grab
these LEDs, assuming that they're "brighter", however THE AMOUNT OF
LIGHT EMITTED BY THE CHIP IS THE SAME - the output is just differently
disposed in space! (I could go into the measurement of emitted cone
angles by explaining the concept of steradian measures here, but for
those who know, it's hardly necessary).

A better rough measure of the LED's light output is given by two factors:

(1) LED coversion efficiency.
(2) Maximum permissable continuous LED current.

The product of these two roughly give the chip's light output, but the
"optical gain" of the encapsulent lens is quite an unrelated matter!
The chip's maximum electrical dissipation figures in milliwatts may
be a better measure.

A lot of the limitations are now set by which LED producing firm has
access to recent patents on chip and design improvements, but there
are two outstanding players in the field who have attracted imitator
manufacturers, especially in China. These imitations are invariably
inferior to the branded products. The outstanding branded products
are made by:

CREE has produced white LEDs which apparently have just about
the highest conversion efficiency of any LEDs, they are frequently
used in multi-led torches - however they do not have access to
LUMILED's (the "Luxeon's") patents on high-current designs.

LUMILEDS (see their website) have produced the LEDs with absolutely
the world's highest light output - operating in the order of 5 watts,
rather than the typical 100 milliwatts of standard LED designs.

If you wanted to plug an optimum LED into that "Televisor Kit" without
any modification of the driving circuit, I would reccomend the CREE
LED product; however if you can build a heftier modulator and power
supply, a single white Luxeon of the 3-watt type would be far better.

Beware of "milliCandela ratings". They seem to be designed for those
with little knowledge - which can, as always, be a dangerous thing!

All the best to all,

Chris Long VK3AML.
chris_vk3aml
 

Postby DrZarkov » Mon Sep 29, 2008 4:15 pm

I've got some LEDs from Andrew, he bought them in Hongkong. They are titled as 40000 mcd, and they are really very bright! They are a lot better than the high-quality LEDs from Nashua with 11000 mcd. In that case they were worth the money.
User avatar
DrZarkov
I think I've had a cranial implosion.
 
Posts: 1041
Joined: Mon Feb 19, 2007 11:28 pm
Location: Kamp-Lintfort, Germany

Postby Steve Anderson » Mon Sep 29, 2008 5:50 pm

I bought a couple of white LED torches/flashlights. For two quid/four Euros a pre-built assembly comprising nine white LEDs and a laser diode...made in China of course! Included (cheap) batteries as well. Certainly bright enough.

Remove the battery holder, apply modulation, off you go.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5360
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Postby chris_vk3aml » Mon Sep 29, 2008 7:17 pm

OK, they're labelled as "40,000 mCd". That may be for
narrow-angled emission, and with the modulated LED
close behind the Nipkow disc, a narrow angle of emission
is exactly what you DO NOT want. What you need for that
kit is a LED with a fairly wide angle of emission, to
evenly illuminate the scanned area. A Lambertian radiation
polar pattern would bear nearly ideal.

OK - it's better to ignore the "mCd" rating and look
at these LED's conversion efficiency and maximum
dissipation. If the led can take 60mA instead of 20mA,
the higher current model is to be preferred regardless
of the axial far-field flux measurement. What you want
is maximum radiated flux emitted over a wide angle,
ie. the chip with the maximum flux output. Ignore the
"mCd" rating - it probably relates more to the optical
arrangement of the chip within its encapsulent lens.
Look at the device dissipation. If the device's light
conversion efficiency (or "wall plug efficiency") is
nearly the same as the other LED, you should be
finding the device that can generate the greatest
flux via increased permissable driving current, NOT
the device with high-far-field intensity obtained
through narrow-angle "optical gain".

Hope this clarifies the situation - if you'd like me to do
so, I can give a swag of Internet references to scientific
papers on this subject. VK7MJ, KA7OEI and I have gone
into this in considerable depth, while searching for diode
chips suitable for the atmospheric modulated light comms
systems that we've had operating over ranges of 150km plus.

Best wishes,

Chris Long VK3AML.[/u]
chris_vk3aml
 

PreviousNext

Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 22 guests

cron