Page 14 of 18

PostPosted: Wed Jan 02, 2013 12:03 am
by Steve Anderson
I know this will bring groans from many, but Albert, you've been framed.

Steve A.

Happy new year to all. Let's hope the '13' bit of 2013 has no significance...which it won't be of course...

PostPosted: Thu Jan 03, 2013 6:27 pm
by Viewmaster
gary wrote:Hmmmm, y'know Albert, what you are ending up with here is a NBTV version of the digital photo frame! ;-)


Here's the rear view of the NBTV digital photo frame. :)
Power supply, top right. Next to it the audio amplifiers and then sync. ccts and Steve's (A) missing 32 line pulse inserter and variable frame generator........
The kitchen sink is under the other board.........
I've decided to put that in so water is quickly available for the fire on switch on. :shock: :lol:
This is only a first unfinished assembly in order to position the holes for the 6 pot controls and switches.

PostPosted: Thu Jan 03, 2013 6:30 pm
by gary
woo hoo! - that's really coming along - and looking great!

PostPosted: Sun Jan 06, 2013 9:55 pm
by Viewmaster
If any brave soul is building a Niptrix here is a little tip.......
......Don't! :shock:
But seriously, the 48 cross wires on the matrix are very close to each other and are easily bent in towards each other.
If a short should occur it would put 2 LEDs on at the same time leading to increased current.

So, I have installed 24 pieces of stout black card bent over to
form a long V.
These are inserted at every alternative cross wire and so insulates every one from the other.

"But is it working yet?", I hear the cry go up.

I am waiting for a 7556 to arrive, all 10 IDC connectors to be made up and then ready for lift off/take cover/call the fire brigade/
/accept the wife's divorce proceedings :lol:

PostPosted: Mon Jan 07, 2013 6:40 pm
by AncientBrit
Ah Albert, great to have the encouragement and support from one's spouse!!

Also good to see 'Health & Safety' red tape over the mains input on the 'trany.

Cheers,

Graham

PostPosted: Sun Jan 13, 2013 12:22 am
by Viewmaster
Testing, testing, 123, 312, 213 :shock:

Connected PP9 battery for testing and running seperate test oscillators for clocking.
Off we went except the Niptrix's LED running sequence is wrong in about 40 places. Instead of LEDs lighting 1,2,3,4,5,6,7 they kinda light up 1,2,4,5,3,7,6, on rows and columns too

Penny dropped this morning.........
Because the 4017's used in cascade have their o/p pins on two sides of the IC and are not in any sequence I have wired some out of correct running order.

So more bleedin' work to swop about 40 connections over, then on goes the PP9 again :)

Note my little minor scope. Cute eh?

PostPosted: Sun Jan 13, 2013 12:22 pm
by Harry Dalek
No worries Albert its a little mistake but i can see but its a lot of rewiring ..shes a complicated beast !
Dr Z has a video of one of these things going from a past convention on he's you tube channel...yours will be Black and white while this uses Red leds.
https://www.youtube.com/watch?v=tF_EbWKso1Y

PostPosted: Sun Jan 13, 2013 6:44 pm
by Viewmaster
harry dalek wrote:No worries Albert its a little mistake but i can see but its a lot of rewiring ..shes a complicated beast !


It's not too bad, just 2 connections per 4017 =20 total. It was the pins on the 9-16 pin side of the ICs that were wrong.
Fortunately no components need to be changed, just wires reconnected. Thank goodness the error wasn't such as to blow up all my LEDs. I have big R in series with them at present but I still worry. :)


harry dalek wrote:Dr Z has a video of one of these things going from a past convention on he's you tube channel...yours will be Black and white while this uses Red leds.
https://www.youtube.com/watch?v=tF_EbWKso1Y


Yes, it was made by Rommert, who writes about it in an early part of this thread. His construction is very professional looking.

I bet you are still busy marking/drilling/cutting/soldering/swearing(?) away. :wink:

PostPosted: Mon Jan 14, 2013 7:30 pm
by Viewmaster
I have the Niptix running now with NBTV signals.
Results so far........

1. With less than 15ma through the LEDs the picture can be watched in a room with the normal lights on.

2. One 4017 on the line clocking is not working so picture is only 25 pixels wide. ( it was working yesterday) so that has to be attended to.

4. The third column is on all the time......discovered that when I changed leads over I have shorted a transistor, so that will have to be sorted.

5. This is the most interesting. There is a 2 pixel constant vertical shift up and down on the picture at about 3 cycles per sec.

Thinking about this I believe that the 19.2kHz pixel clock osc is very very slightly off from the pixel rate from the CD.
So, when a new line starts sometimes the 19.2 first pixel clocking has not quite started and at other times it is one cycle too late, resulting in this slight cyclic vertical shift. I may change the CD to see if the cycle alters slightly to prove this.

I shall put a trimmer in the Xtal osc circuit to try and 'pull' its freq.
Or reset the osc at every line using the 32 line pulses.
Or more interestingly, put in a free running 19.2kHz osc pot control.....this will still give a good stable picture which can be either 'stretched' or 'compressed' vertically by slight pot twiddle.
I reckon that if, for example, the osc ran at half speed the picture would only be 24 pixels high and if twice the speed only half the picture would be shown over the 48 pixels.

One can do interesting things with a matrix that you can't do with a Nipkow disk. :lol:

Other than all the above I think that the picture is very impressive when viewed at about 8 ft away on the club's test CD.

I shall put my initial results up on my site soon, together, hopefully with the correctly running pictures.

PostPosted: Mon Jan 14, 2013 8:01 pm
by gary
Yes Albert, the clock freqs will have ppm variations which will result in that drift - trimming should improve things but whether it will be enough remains to be seen. Really the clocks need to be locked some way - and the only way would appear to be using the video signal syncs - but being able to do fine adjustments should be fun to try!

Of course none of what you wrote about happened until we see video! ;-) Can't wait!

PostPosted: Mon Jan 14, 2013 8:53 pm
by Steve Anderson
Don't forget not all CD players are created equal...somewhere else on this forum a member pointed out that his CD player used a ceramic resonator as its major timing element. These are generally ±0.5% accurate, as opposed to the cheapest crystal which is ±100ppm ( 0.01%) or better. Therein maybe tho source of this...but even so with better than 0.01% crystal accuracy it will still happen, just much more slowly.

Also the original video may not be exactly spot-on too...

I would suggest as you already have done, in resetting the counter/oscillator system once every frame or line somehow.

Otherwise, congrats....

Steve A.

PostPosted: Tue Jan 15, 2013 12:00 am
by Viewmaster
Steve Anderson wrote:I would suggest as you already have done, in resetting the counter/oscillator system once every frame or line somehow.
Steve A.


Thanks Gary and Steve for opinions on the slight bouncing.

The 4060 xtal osc IC does have a reset pin so I will connect that to the 32 line pulse line and try that. There maybe a pixel error on start up of each line but it should be a constant on all lines and so cure the vertical 2 pixel jumping.

Even if this works I like the idea of manual pixel rate adjustment as the picture would not roll/slide or break up as the line and frame pulses would still control and lock the display on the Niptrix.
It would be interesting to watch the vertical expansion/contraction of the display methinks.
Baird's commutator on his lamp display would not have allowed this to be done, so maybe it has never been seen on NBTV before. :shock:
(but maybe Steve has seen it on his CRT display?)

"Roll up, roll up, see NBTV expanded and contracted before your very eyes." :wink:

Ain't NBTV just so blooming exciting :lol:

PostPosted: Tue Jan 15, 2013 3:52 am
by M3DVQ
Viewmaster wrote:Baird's commutator on his lamp display would not have allowed this to be done, so maybe it has never been seen on NBTV before. :shock:
(but maybe Steve has seen it on his CRT display?)


I have experimented with changing the speed of the signal going into my scope driver by using a PC as the source and altering the playback speed of the file.

You get all sorts of weird and wonderful patterns, especially at exact multiples of the usual speed :D

PostPosted: Tue Jan 15, 2013 12:19 pm
by Steve Anderson
Viewmaster wrote:...but maybe Steve has seen it on his CRT display?...


Yes, a positive pulse on the MR pin will stop and clear the 4060 counter, it also stops the oscillator. I have no idea how quickly these oscillators will re-start but probably quick enough for this display. What frequency is the crystal? And which output are you using (pin No.)?

Yes, the stretch/squash is simply achieved by altering the size of the ramp going into the deflection amplifiers. These 'size' controls (H & V) are either front-panel pots or internal presets. It allows you to set the correct aspect ratio so that a circle is really a circle.

Steve A.

PostPosted: Tue Jan 15, 2013 6:30 pm
by Viewmaster
Steve Anderson wrote:
Viewmaster wrote:...but maybe Steve has seen it on his CRT display?...


Yes, a positive pulse on the MR pin will stop and clear the 4060 counter, it also stops the oscillator. I have no idea how quickly these oscillators will re-start but probably quick enough for this display. What frequency is the crystal? And which output are you using (pin No.)?
Steve A.


The xtal is 4.9152 mHz and the division is 256 on pin14 to give the 19.2kHz.
I did suggest that I might loose a pixel at the start of each line and I said this because there might be a biggish delay in the reset time being anywhere near close to the pixel rate.
Probably the reset time is in a few usecs so we may well be OK.
Have sorted out two of the other problems I mentioned so we pogress.
Thank gawd all that tedious soldering is over.......I never wish to see another soldering iron in my life. :lol: