NBTV/SSTV-625 Up Converter MkIV

[Harry Dalek]

OK guys, glad I didn't offend anyone in pointing out the previous...

Harry, resistors all 5% except perhaps R207/208 at 1% to reduce refections/ringing in mismatched coax lines, though it's unlikely, why chance it?

Caps are no issue at all as regards tolerance, only those marked 'D.C.' do need to be disc-ceramic with the shortest possible lead length to the IC package, best to put them on the underside of the board right on the device pins.

Other than that, a nice chunk of 0V copper and a compact layout, things should be ticky-boo.

Steve A.

That's ok i have done no wiring around the ic areas apart from the oscillator ,i am thinking of the 0v ground copper area round the bottom part of ic101 before moving on later in the week when i can work on it again .

Don't worry i will decouple ,i recall on the few go's on your DA circuit this is was very important to the point of not working or working with them...so rather with them !
Trying to do it all neat and testing as i go .
As mentioned a few posts back i need some advice on some of the wiring i am not sure about but not up to any of that yet .

[acl]

51795B.pdf

Handbook for PICkit 3

(1.3 MiB) Downloaded 259 times

Hi there Steve,

I managed to power up the stand-alone interface card connected to the PICkit 3 successfully as you described. It is slightly different on V8.89 but I now know how to do it so I am now the UK supplier (nearly said EU then) of these programmed devices. I don't know how i did it but as a test I loaded write 10 .asm into MPLAB and compiled and managed to program into a 26K22 device successfully . I was expecting warnings to come up as the headers in the ASM file were not the same as the programmer settings. I assume these two devices are simular. I take on board your comments about the board and will await your progress, There were a lot of interconnections. I will try my build out when you have some software avalible but then revert to a PCB or have another go later taking on board your comments.

Harry, I used single strand 30AWG kevlar coated wire wrapping cable for interconnections. It is availible in several colours so you can colour code cabling

Rgards Chris Lewis

[Steve Anderson]

Ah yes! I'm fairly sure that's one of the information files on the CD/DVD I mentioned.

Yes, I expected one or two generally minor variations between versions of MPLAB IDE, but generally they're much of a muchness.

If the .pdf files on the CD/DVD don't add up to too much I'll attach them later to this post.

Later...nearly 60MB of .pdf files, and that's in English only! Though once I got to the stage you're at I don't think I've ever needed to refer to them since.

Steve A.

Though I am lusting after the 16-bit PIC24FJxxGA002 series of devices (still DIP/DIL 28-pin and a PicKit 3 compatible), I'll continue on with these 8-bit versions...but expect 16-bit devices to displace 8-bit versions in the future. Remember the 16-bit PCs, 286 etc...now in a 28 PDIP package at around US$2.00 each!...and they don't need math co-processors to run some programs, Autocad (of the day) for example.

[Harry Dalek]

IHave a look at my layout posted on the 12/01/21. Note the large amount of green 0v tracks. Every chip should have a dedicated decoupling capacitor, the D-A two plus its compensation cap on pin 3-16. The regulators should be on-board and close by and well bypassed/decoupled too. I also rotated IC203 for similar reasons. I was hoping a largish chunk of copper might have appeared at some point.

But for the time being continue on but try to include as best as possible my suggestions here.

Steve A.

Steve i don't know what you have planned for the ground plane as in doing it with copper wire or copper foil ,strip board tracks .looking at the idea on circuit boards the copper foil tends to almost surround the circuits.

I see on your 12.1.21 post its an area pretty much between the bottom and sides of IC101 102 so between those and 23lc1024 and the DAC chip...thinking of taking one of the 23lc1024 out and placing it further down so i can put this copper foil across the board as ground touch wider .

[Steve Anderson]

I'm being delayed by my major suppliers server being uncooperative, it's just not accepting on-line orders. If it drags on a few more days I may have to take a trip across Bangkok to get what I need...which I would rather not do. The expense and the time...basically a whole day, though in some masochistic way I do like going there....but normally If I can order on-line on Monday, it's here on Wednesday and usually costs less than a quid in p&p. While I'm waiting I can get on with other things...

However, 'other things' may include DSP techniques for demodulation, watch this space...

If you recall your calculus, try this, I don't claim originality...

Demodn={Qn*In-1-In*Qn-1}/{In2+Qn2}

Here "n" is used to indicate the current sample and "n-1" is used to indicate the sample just before the current sample. "I" is the real or in-phase part and "Q" is the imaginary or quadrature part. This is a "first order" detector that works very well for our slow-scan design.

Credit KR7A in the US..or maybe Denmark...not sure...

FIR Fiter 01.gif (20.62 KiB) Viewed 7554 times

However, the divide function is a bug-bear of many microcontrollers. hence my predilection for the PIC24s..but these PIC18s should just (maybe) do the job...as well as the FIR filters required...as above...this is really a development in progress exercise...I'm trying to look beyond the 1970s methodology.

Steve A.

[Steve Anderson]

Hmm, this FIR filtering within the write processor (the demodulation part) is proving to be an entire sub-project in itself!

If you recall the original MkI had a noticeable overshoot on the left-hand side of the image as per attached. That irked me...a lot! And it still does! And I'm determined to eradicate it. There's also a small amount of contouring down the temples of her face. If I recall it was a slightly defective D-A. I was able to get rid of it with a replacement DA-08.

In that original arrangement the are three filters involved, all linear, all with potential overshoot artifacts. The first is in the modulator which band-width limits the incoming baseband video. The second is on the modulator output which cleans up the DDS generated sine-wave output, this is an unlikely suspect. The third is in the modified Robot 70 demodulator I used on the input to the up-convertor, this is my prime suspect, though the first one may not escape scot-free.

Hence my interest in using FIR filters which are phase-linear whereas the analogue versions used before are not (phase-linear analogue filters are real nightmare!). As it's done in software there's no hardware to add or build.

Stick with this, it's not going to happen overnight, such is the manner of design...

Steve A.

As a matter of interest, how is anyone going to generate their own SSTV files/audio? Or do I sense my next project on the horizon?

First Photo 2..jpg (141.14 KiB) Viewed 7528 times

[Harry Dalek]

Hi Steve i think you posted a few SSTV wav files pubs in England as i recall there might be a few vintage recordings on the net ,i was thinking since you have not programmed this to see the VIS code your converter would ignore it ?
http://www.g0hwc.com/sstv_modes.html
Some old off air recordings 1980...more than likely the VIS code used ?
https://www.qsl.net/kd2bd/sstv.html

[Steve Anderson]

Yes, a few pubs...and a few other 'landmarks' as well. Vis-codes are totally ignored as per Cop McDonald's system. It was in in its day the only standard around. With the exception of the line-rate being 60Hz/4 (15Hz) or 50Hz/3 (16.7Hz) depending on what country you were in. The MkI was 50Hz (16.7Hz) only. But it wouldn't have taken much to make it dual-standard. I should think on that for this version. It is 'only' code after all...as well as those that adhere to the original 120-line version or the more binary-friendly 128-line specification...so four versions in total...

I'm tempted to have a go (at some much later date) at the PD120 SSTV mode. But that's pie-in-the-sky at the moment...the mode that's used from the ISS to earth. But with so many PC software implementations already around I wonder if it's worth it? Probably not.

So maybe in a while I'll be looking for a something to do...maybe I'll revisit "The Brute" from years ago, a good old-fashioned CRT NBTV/SSTV monitor...yes, with tubes/valves, but probably a bit of a hybrid. Also, something I never got beyond just fag-packet sketches (and not published nor mentioned I think), a mechanical SSTV camera..a perfect companion to this thread...I've had a few ideas scuttling around in the dregs of my notepads (and my gray matter) for years...but this would be real simple, old-school stuff...though using modern, inexpensive and easy to build hardware.

Got a working 33 RPM turntable? Hang on to it, no audio pickup required, just the motor/belts/pullies required...just to get the thing to rotate smoothly at 33RPM....while carrying a balanced (approx) 500g load...mostly rechargeable batteries...and most of the mass placed in the centre next to the record spindle...Probably 4x 8.4V/9.0V PP3 (as they're called in the UK, or 6F22 in the US I think) batteries.

Steve A.

[Harry Dalek]

Yes, a few pubs...and a few other 'landmarks' as well. Vis-codes are totally ignored as per Cop McDonald's system. It was in in its day the only standard around. With the exception of the line-rate being 60Hz/4 (15Hz) or 50Hz/3 (16.7Hz) depending on what country you were in. The MkI was 50Hz (16.7Hz) only. But it wouldn't have taken much to make it dual-standard. I should think on that for this version. It is 'only' code after all...as well as those that adhere to the original 120-line version or the more binary-friendly 128-line specification...so four versions in total...

Yes it would be good to be a 7 and 8 sec converter
I found one of your pub wav's

The White Horse Shere 1.WAV

(468.79 KiB) Downloaded 268 times

I'm tempted to have a go (at some much later date) at the PD120 SSTV mode. But that's pie-in-the-sky at the moment...the mode that's used from the ISS to earth. But with so many PC software implementations already around I wonder if it's worth it? Probably not.

I think most of us like the earlier modes up to early colour then it just went crazy to many modes and i am sure those at the time didn't want a different system every few weeks if i spent my time making a monitor or early scan converter back then i would not be happy its obsolete or having to guess what systems transmitted .

So maybe in a while I'll be looking for a something to do...maybe I'll revisit "The Brute" from years ago, a good old-fashioned CRT NBTV/SSTV monitor...yes, with tubes/valves, but probably a bit of a hybrid. Also, something I never got beyond just fag-packet sketches (and not published nor mentioned I think), a mechanical SSTV camera..a perfect companion to this thread...I've had a few ideas scuttling around in the dregs of my notepads (and my gray matter) for years...but this would be real simple, old-school stuff...though using modern, inexpensive and easy to build hardware.

Arrr the Brute do you still have it in storage it was sort of my inspirational idea for the Binocular and of cause most of the monitor part from your mind using valves with a CRT i have to try always back of my mind to try , must be hell for storing electronic's in Thailand the summer heat here tends to make my old projects fall to bits ..and electronics it must be aging capacitors on that side of things ...But yea you have been working on these chip projects for a few years .

A Mechanical SSTV camera or a easier flying spot camera (slide) perhaps would be interesting ...you would try a mechanical camera ? rotating drum i have tried so its possible if i can mange it Old school now your talking Australian !

Got a working 33 RPM turntable? Hang on to it, no pickup required, just the motor/belts/pullies required...just to get the thing to rotate smoothly at 33RPM....while carrying a balanced (approx) 500g load...mostly rechargeable batteries...and most of the mass placed in the centre next to the record spindle...Probably 4x 8.4V/9.0V PP3 (as they're called in the UK, or 6F22 in the US I think) batteries.

Sounds like my PhonoVision need to put the cartridges on you got me off track here ...but using the turntable for a camera idea just need a photodiode you have the line scan with the turntable movement if you have a can drum on that and a photo wrapped around ...the frame movement for a start can even be hand moved down till you work out a geared movement review ..sstv experiments here ....a direct light SSTV camera would be even more interesting .

Steve A.[/quote]

[Steve Anderson]

Yes it would be good to be a 7 and 8 sec converter

I'll think on it...but I'm sure I have in the past and came across problem with aspect ratio I think...but not impossible.

I found one of your pub wav's

The White Horse Shere 1.WAV

All of the .wav files I uploaded all that time ago are still there, in this folder at that date...
Re: Retro SSTV anyone?
Post by Steve Anderson » Fri Feb 12, 2016 10:15 am

Arrr the Brute do you still have it in storage?

Yes, it's been accumulating a lot of dust over quite a long period!!

A Mechanical SSTV camera or a easier flying spot camera (slide) perhaps would be interesting ...you would try a mechanical camera ? rotating drum i have tried so its possible if i can mange it Old school now your talking Australian !

...a mechanical SSTV camera..a perfect companion to this thread..

I did experiment with a flying snot idea some years ago, without success, maybe one day I'll have another go...maybe.

I'll knock-up a sketch of the concept and add it here later...at 33.3RPM it will take about 3m 50s to do a 128-line scan! Sit very still for your close-up!

As you might expect from me the 'mechanics' are as simple as downright possible!

A hint, I often give my projects a silly name, this one is 'Noddy'..though no connection to 'Big Ears'...

Steve A.

...don't forget Harry, I lived not that far from you in the 70s...less the 60km south...hence the occasional Strine slips in!

[Harry Dalek]

Yes, it's been accumulating a lot of dust over quite a long period!!

Bit of a clean up worth the effort to finish that one off ,she'll be a beauty for sure .

I did experiment with a flying spot idea some years ago, without success, maybe one day I'll have another go...maybe.

Many ways to go on this ...scanner or the 2 forms for camera direct flying spot could be electronic CRT scanning a slide then moving on to a photo pretty much have a direct light camera then just have to play with the optics .

I'll knock-up a sketch of the concept and add it here later...at 33.3RPM it will take about 3m 50s to do a 128-line scan! Sit very still for your close-up!

I love the idea of using a turntable to get the line scan rest is easy ,do sketch up your idea turn table for this very interesting .

As you might expect from me the 'mechanics' are as simple as downright possible!

I think your planning a direct light camera my down fall when i tried to turn the scanner into this was i never took into account it needs to be in a light proof case only light via the the lens ..mechanically i was there had some close up results never got to a fair distance image ...live and learn
viewtopic.php?f=23&t=1890&start=135

A hint, I often give my projects a silly name, this one is 'Noddy'..though no connection to 'Big Ears'...

Oh that's like how i would envision it working with that Name ...i will wait to see !

Steve A.

...don't forget Harry, I lived not that far from you in the 70s...less the 60km south...hence the occasional Strine slips in!

[/quote]

We talk like that to keep the fly's out of our mouths ..so i blame the fly's for this form of English ! I should think your neck of the woods it would come in handy for the same reason they like the heat the fly's .

Getting back to this project i just started on the decoupling caps on the power supply few more to go didn't have much time today takes me half the day to mow the yard so only got a few in this afternoon,i will move on to decoupling caps on the crystal oscillator i had them on before but removed to place them neater this time .i can move onto the rest of it .
Any more thoughts on IC102 Steve as in to be used or not ..i am not sure if i should solder a socket in .
Edit
I found i had a heap of what i think are tantalum capacitors 100N due to room and keeping it neater i have replaced the ceramic.

[acl]

I think I have resolved the crystal issue , I was expecting a pure square wave out of this module. I purchased two more from a different supplier , different idents and they performed the same. The original supplier tested some and here are his comments. The scope did display the correct frequency of 16MHz.

'I have tested some of these in th past on a HP 1740A 100Mhz analogue scope and the output looked ok but will check again and let you know.
Please bear in mind that TTL levels are below 0.8V is a low and above 2.7V a is high.
Unfortunately due to covid restrictions in Scotland it may be the weekend before I can get to our workshop to test
regards,

'I finally got round to checking the oscillators and can confirm that they work ok. When using a scope it is better to connect the output directly to the scope input as when using a probe or cable as the shunt capacitance puts a capcitive load on the osc rounding the wveform. It is also desirable to bandwith limit the scope (my Tektronix DSO has a 20Mhz bandwidth limit function) as this reduces the harmonic output of the osc and lets you see the actual wavefom.
I am confused by your picture as it shows a 49Mhz waveform as you bought 16M oscs.
Also the voltages are are around 10x higher so your scopoe may be configured for a 10x probe.
TTL is very tolerant of a distorted waveform as the inputs are current driven and also when isnstalled in a board the short PCB tracks help eliminate the problems described above.
A 0.1uF ceramic capacitor connected across the supply and ground leads also helps the output purity.
Hope this helps'

[acl]

Hi there Harry,

Looking good.

Regards Chris Lewis

[Steve Anderson]

The output from these oscillators can often look somewhat 'grotty'. It depends on the grounding, bypassing and general layout of the circuit. All the points I so often go on about. Plus how your 'scope is hooked up to the circuit.

Also consider the waveform itself. Being a square wave it's comprised of odd harmonics only. So there's the fundamental at 16MHz, the third harmonic at 48MHz, the 5th at 80MHz and the 7th at 112MHz, which is beyond the bandwidth of a 100MHz 'scope. So it will never look as clean as you might think or hope for. And internally the micros bump that 16MHz up to 64MHz via a x4 PLL, the first odd harmonic of which (the 3rd) is 192MHz...no chance of seeing that on a hobbyists 'scope!

emct_2e_6-21.jpg (31.66 KiB) Viewed 7478 times

Steve A.

[Harry Dalek]

Hi there Harry,

Looking good.

Regards Chris Lewis

Thanks Chris i am glad you got you oscillator working bit of a catch 22 need that or nothing else works bit like us no heart beat we are goners !
I wanted to do more today but had to fix my leaf blower 4 hours later and a replaced carburetor there went the rest of the day .
I'll work more onit tomorrow unless the boss has some has other plans !