NBTV/SSTV-625 Up Converter MkIII

Where it all started as far as most are concerned and saw heavy use from the 60s through to the 80s. Colour and Hi-res modes have unfortunately pushed this system into the backwaters of SSTV. Time to resurrect interest in this simple analogue system.

NBTV/SSTV-625 Up Converter MkIII

Postby Steve Anderson » Thu May 04, 2017 5:19 pm

So, here's the new thread. Photo of the completed assembly of the MkIII, so much less wiring than the MkI. I've left the two 14-pin sockets in place...just in case! Note the 'pile-up' in the centre...I put that electrolytic cap in the wrong place!

Next I need to conduct some tests to ensure my idea of deleting the eight AND gates is valid. All being well, after that it's porting and modifying the software from the MkI version which I have already made a small start on.

Steve A.

I've just realised that four short links are missing and two resistors...oops. Now fixed.
Attachments
UP-625 MkIII 1.jpg
UP-625 MkIII 1.jpg (132.41 KiB) Viewed 18544 times
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5360
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: NBTV/SSTV-625 Up Converter MkIII

Postby acl » Thu May 04, 2017 6:23 pm

Looking good Steve,

I will hold off construction until you are happy with both hardware and software design.

Regards Chris Lewis
acl
Anyone have a spare straightjacket?
 
Posts: 439
Joined: Thu Mar 08, 2007 7:38 am

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Steve Anderson » Thu May 04, 2017 6:54 pm

acl wrote:I will hold off construction until you are happy with both hardware and software design.

That's a good idea. Things are very fluid at the moment, I wouldn't want anyone copying this just yet. This could be months in gestation, though I hope not, whatever, don't hold your breath!

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

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Steve Anderson » Fri May 05, 2017 3:08 pm

I have just realised my plan to do away with the eight AND gates is flawed, and I haven't even applied power! While tidying up the drawings the obvious dawned on me.

So unless I can come up with some other scheme it's going to be back to the MkII with the AND gates and this one laid to rest. I had a feeling it was too easy!

Added later...I may have cracked it. No additional components at all. But I need to do some speed tests to see if it's viable. The datasheets imply it is but it's not a certainty.

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

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Steve Anderson » Fri May 12, 2017 2:45 pm

Progress on this has come to a halt while I wait for my local stockist of the processor chips to receive a delivery. It's been, "Maybe tomorrow." for a week now. I could use a slightly different version but they only stock this particular flavour. A PIC18F26K20 would do fine. It's functionally the same, just more memory than the 25K20. By the time I place an order with another S.E.Asia vendor the local guys should already have them. All the other parts I have already.

I do have some 45K20s they're the same as the 25K20 but in a 40-pin package which provides an additional 8-bit port and three additional I/O lines. But they're useless here.

I did think of continuing on with the code, but until I'm sure my fix for removing the AND gates mentioned above is OK it's not worth devoting the time to it.

Updates as and when.

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

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Harry Dalek » Fri May 12, 2017 3:08 pm

Steve Anderson wrote:Progress on this has come to a halt while I wait for my local stockist of the processor chips to receive a delivery. It's been, "Maybe tomorrow." for a week now. I could use a slightly different version but they only stock this particular flavour. A PIC18F26K20 would do fine. It's functionally the same, just more memory than the 25K20. By the time I place an order with another S.E.Asia vendor the local guys should already have them. All the other parts I have already.

Updates as and when.

Steve A.


I like ic sockets !

I know how useful Ic sockets are nothing worse than having to desolder a chip but there is a much better electrical connection soldering the ic right in place and over time the socket will tarnish and perhaps cause electrical connection error .

Having said all this would over time it be a better idea once the circuit is profected to wait and solder them in place....perhaps just keeping the main memory as a socket since i have been reading the problems of these going bad .

More than likely to save money but i notice its very rarely used in devices over the last few years suppose you have to weigh up the likely hood of the ic going bad ........in a well designed circuit ? would they be needed would you ever bother fixing it in years to come ?
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Re: NBTV/SSTV-625 Up Converter MkIII

Postby acl » Sat May 13, 2017 1:34 am

Hi Steve,

I await your progress with interest.

Regards Chris Lewis
acl
Anyone have a spare straightjacket?
 
Posts: 439
Joined: Thu Mar 08, 2007 7:38 am

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Klaas Robers » Sat May 13, 2017 6:59 am

Steve, I also found in the Robot documentation that, when the converter is in the 525 line 60 Hz mode, that the first and the last lines of the 128 line memory content are in practice not visible on the monitor or TV screen. This is also theoretically not possible.

Each line is shown twice in a not interlaced field. So there are 256 lines from the memory, while the NTSC scanning is 263 lines per field. Of these 263 lines there should be a vertical blanking interval of 20 lines, in which the V-sync is situated. So in theory there is an active area of 243 lines.......

Robot starts the V-sync pulse (of 4 lines) already on line 254 of the active video area, so the last 2 lines are any way blanked by the V-sync. Then the video starts again on line 9 after the beginning of the V-sync. It is doubtfull if a video monitor will display these lines, as a blanking until line 18 is expected. Robot writes something in their documentation that the vertical size and vertical shift should be adjusted until an acceptable situation is obtained.

In the situation of 625 lines 50 Hz the situation is somewhat better. Of the 312 lines per field there is a standard V-blanking of 25 lines, 2.5 before the V-sync, 2.5 lines V-sync and 22.5 lines after the beginning of the V-sync. here the active video starts 28 lines after the beginning of the V-sync pulses. You may have seen on the photos that my monitor has already problems with the first lines. May be I should shrink the V-size somewhat.

Any way the active area is vertically shifted to the top of the screen. I looked into the diagrams, if I could shift the V-sync, but it looks that this is not simply possible, whithout adding extra TTL circuits. In principle the active video should start at line 38 from the start of the V-sync.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Steve Anderson » Sat May 13, 2017 12:00 pm

Yes, I have often wondered how they managed to get this sort of thing working in 525. In this case it's a bit of a bodge. But 625 has enough lines to fit it in with a few spare which I leave at black level. Even if it were possible to fit them in the overscan in CRT TVs/monitors mean that a few at the top and bottom were not visible. With the MkI up-converter the first line with active video is line 39 (even field) and the last is line 294, line 352 (odd field) to line 607. The entire picture is just visible on the CRT TV I'm using.

Two pdfs attached are what I use for 525 and 625 VBI reference. I also have the UK IBA's Technical Reference books which have the same information though only for 625.

If anyone is interested I have the same sort of thing for 405 and 819 lines as well as the line timings in the same sort of format for all four systems.

Steve A.
Attachments
525 VBI.pdf
(43.51 KiB) Downloaded 757 times
625 PAL VBI.pdf
(40.19 KiB) Downloaded 664 times
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5360
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Steve Anderson » Sat May 13, 2017 12:12 pm

Here's all the others, I thought I may as well upload them...

The FSTV Horiz Blanking pdf has all four systems in one file, 405, 525, 625 and 819. Includes syncs, blanking and burst (where applicable)...and 625 SECAM!!

Ignore any references to data on the web-page, the site has gone....WRONG, it has moved here now...

http://www.radios-tv.co.uk/Pembers/Worl ... index.html

Steve A,
Attachments
405 VBI.pdf
(13.81 KiB) Downloaded 782 times
819 VBI.pdf
(51.66 KiB) Downloaded 642 times
FSTV Horiz Blanking etc.pdf
(24.82 KiB) Downloaded 642 times
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5360
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Klaas Robers » Sat May 13, 2017 8:21 pm

Yes Steve, that is it. Spot on. Start your video at line 39 after the beginning of the V-sync pulse. I would like that I could make that too in the Robot, without adding extra gate ICs.

The Robot has no interlace, so it is in fact a 312 line 50 Hz progressive scan system. The vertical line counter starts at 0000 0000 at the beginning of the memory read out, this runs until "line" 256, where the video signal is blanked by bit 7. Then now at line 284 a 4 line vertical sync is gated out (284 - 287), and after line 311 the address counter is synchronously reset.

The gates for the 284 and for the 311 are both 8-input NAND gates (7430) with all bits of the line counter connected to an input, and by bending out pins before insertion in the sockets, the line numbers are selected. This is a simple mechanism, but has certain limitations. This is where software beats the simple hardware. I cannot select line 268 for the V-sync, as the 7430 only selects on the ones, not the zeros.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Steve Anderson » Sat May 13, 2017 9:26 pm

Just a thought, 268d is 0x010C (hex) or 0001 0000 1100b. A three-input AND or NAND gate would detect that correctly depending on the polarity of the output you needed. This would require an add-on sub-board though. Even if you mean 267d (counting the first line as zero) that's 0001 0000 1011b which requires a four-input AND/NAND gate.

I assume the counting for 311 is correct, the first line being line zero.

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

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Klaas Robers » Sun May 14, 2017 2:39 am

Yes Steve, you are right......., if the counter wasn't counting further. Let me explain:
The counter proceeds once in two lines. This is the counter for addressing the memory vertically. Each memory line is displayed twice.

268 / 2 = 138d = 86h = 1000 0110b
The bold ones can be gated out. Then on line 268 (and 269) there is the sync pulse, and
the not bold bits are don't cares, so
270 / 2 = 139d = 87h = 1000 0111b
is also gated out, so lines 270 and 271 are also giving the sync pulse. 4 lines V-sync, fine.

However...... the V-counter continues to line 312, so:
284 / 2 = 142d = 8Eh = 1000 1110b and
286 / 2 = 143d = 8Fh = 1000 1111b together,
will generate a second V-sync pulse of 4 lines, and

300 / 2 = 150d = 96h = 1001 0110b and
302 / 2 = 151d = 97h = 1001 0111b together,
will generate a third V-sync pulse of 4 lines.

The current situation is that also bit 3 is incuded in the AND , so the V-sync is only on lines 284 - 287 (second example).
So for a V-sync on lines 268 - 271 I need to include the zeros of b3 and b4 inverted in the AND. May be that, when everything is working, that I try to glue a 7402 on top of the 7430, input b3 and b4 in one of the 4 gates and connect the output to one of the non used 7430 inputs. It looks silly, but after the box is closed, nobody sees it.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Steve Anderson » Sun May 14, 2017 12:07 pm

I must admit I haven't studied the 400 in-depth as to how the sync waveforms are generated, you obviously have...you need to!

Many years ago there was a design for Sync Pulse Generator (SPG) in the BATC's CQTV magazine. It was a mathematically correct PAL colour SPG and was perfect in terms of Brunch blanking and the PAL eight-field sequence...important in editing. All TTL of the 74xx series and used some devices which are now made of unobtainium, e.g. 7470. I do have the article but digging through all the issues of CQTV would take some time.

As I recall around 40 chips and required approximately 1.5A at 5V. Today you wouldn't even dream of building such a thing.

Then again there are those who build 405 SPGs using tubes/valves just as would have been done in the 40s. To restore a commercially built version, OK, but to build from scratch - no thanks!

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

Re: NBTV/SSTV-625 Up Converter MkIII

Postby Klaas Robers » Sun May 14, 2017 6:53 pm

Few years ago I found at Philips several experimental ICs 610, designed by a colegue of mine in 1975 (?), a full standard PAL sync generator. This process was called I2L, integrated Injection Logic. The circuits used very low supply current, needed an input clock of 1.25 MHz and gave H-drive, V-drive, Blanking, Sync and as far as I remember the burst gate, all according to the standard. I transferred these circuits to Vic Brown, so the NBTVA members can use them.

I was sure that at the place where I found them nobody ever will use them, nobody even knew what it was. So hystorically seen this is better. There was made a second circuit, 625, compatible to the 610, that provided sync coupled PAL subcarrier, That circuit wasn't saved at the same place.

I2L was a current mode logic needing a supply CURRENT at a voltage of about 0.7 volt. In fact the supply is a diode in forward bias. Strange, but it worked at very low power. Later this was superseeded by the CMOS process. You may search on I2L and will find some on it on the web.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Next

Return to Cop McDonald's FM SSTV System

Who is online

Users browsing this forum: No registered users and 2 guests