Retro SSTV anyone?

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.

Re: Retro SSTV anyone?

Postby Klaas Robers » Thu Sep 17, 2015 10:03 pm

Steve, this looks good. Yes, porches are usefull, I know. However I think it too difficult to adapt my 1973 hardware generator in DTL (Diode Transistor Logic). As earlier said, the NBTV CDs have a porch of one NBTV pixel, 1/64 of a line length, just before the sync. I haven't seen any circuit diagram that uses the end of the NBTV sync as well. For NBTV the down going edge is the only one that does it.

The .avi is nice. You are complaining about the flicker, but that is nothing compared to the flicker on the screen of a -P7 picture tube. This is the same in NBTV, where the 12.5 Hz flicker is enoying, although you get used to it. On a photo its gone. But making a film.......

In the mean time I got the monitor front end running. I said already that I expected problems with the electrolytic capacitors. It pointed out that in the limiter-FM-detector the electrolytic coupling capacitor of 10 uF between the TAA350 and the FCH191 had almost no capacity any more. Only very short needles were present on the 1k5 to ground. I unmounted the PCB and replaced the capacitor by one, slightly thicker and 22 uF. Then the circuits gave fine video signals again. I will try to make a photo.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Retro SSTV anyone?

Postby Klaas Robers » Thu Sep 17, 2015 11:29 pm

Made a photo ! This is the grey scale, this time of 6 steps.
You see the ringing of the Butterworth filter, but you also see that there is some "pre ringing" as well. This is caused by the included all pass filter. In pictures this gives an impression of sharpness. Due to the phase of the FM subcarrier the ringing is "breathing", but the photo shows the overlay of several scans.
You also see that in the bottom of the sync still some (second harmonic of) the FM subcarrier is visible. This is better visible in the much longer frame sync. In the pictures it will not be visible, as sync is always blanked. In the sixth (black) step nothing is visible.
Attachments
Scope grijssschaal.JPG
Scope grijssschaal.JPG (33.65 KiB) Viewed 10658 times
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Retro SSTV anyone?

Postby Steve Anderson » Fri Sep 18, 2015 12:28 am

Yes, more than 40 years, presumably sitting idle for long periods of time, is asking a lot of a humble electrolytic! But many from the valve/tube era will recover if given a sympathetic 'reforming'. With a small example like this it's not worth it.

The waveform looks clean and tidy, ripple in the sync-tips - who cares? It's invisible as you say and it's the edges that are the important part.

Since my last post above I haven't done much apart from doodling some of the hardware fixes I need to attend to and creating a 'TO-DO' list - so not a lot to report.

Maybe more tomorrow.

Thanks for the update Klaas.

Steve A.

Actually, about the only thing worth reporting is I checked the modulator's output spectra using MMSSTV which was 'near enough'.
Attachments
SSTV DDS Modulator Spectrum 1.jpg
SSTV DDS Modulator Spectrum 1.jpg (17.6 KiB) Viewed 10655 times
5-Step Spectra 1.jpg
5-Step Spectra 1.jpg (36.96 KiB) Viewed 10655 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: Retro SSTV anyone?

Postby Klaas Robers » Fri Sep 18, 2015 3:07 am

The word "MMSSTV" doesn't ring a bell to me, but the spectrum pictures look nice. What is it?

I did a few comparable things, I measured the output voltage of the low pass filter with an old valve tone generator connected at the input of the limiter-FM-detector. The frequency counter gave me the real frequencies, so now I know the voltages that correspond to 1200 Hz, 1500 Hz and 2300 Hz. This showed me that the FM oscillator of my SSTV hardware is not spot on frequency. I have to adjust that slighty. Also the indicated voltages in my previous document were not at all correct.

I also have taken time to draw the pass band curve in a diagram that you posted earlier in this thread. The blue line is the 7th order passive Butterworth filter. But you might also enter the values of the inductors and capacitors in the simulation program that you were using for the diagram. The "shoulder" was guessed by me and could be done better.
Attachments
Doorlaatband.GIF
Doorlaatband.GIF (39.73 KiB) Viewed 10650 times
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Retro SSTV anyone?

Postby Steve Anderson » Fri Sep 18, 2015 1:23 pm

MMSSTV is a free Windows based software package which can receive and transmit SSTV and RTTY written by JE3HHT - Makoto Mori.

But, and it's a big but, it doesn't handle the original Copthorne McDonald standard. Even the 8-second B&W mode requires VIS codes. But it is useful for the spectra display, sadly that's about it for us. The website is here...

http://hamsoft.ca/pages/mmsstv.php

Older DOS based software like JVFAX do handle the CM (Copthorne McDonald) standard but don't have the spectra display - it is a much simpler package, as you'd expect.

The plot you show is of the updated Robot 70 demodulator shown in the posting of 28/02/2015 (This thread, page 2). It's just of the filter after the differential amplifier. I modified this section of the demodulator to a) get the signal 'the right way up' by reversing the inputs to the differential amp. b) added VR101 to set the sync-tips to 0V, c) added VR102 to set the video level. The actual filter is unchanged. (Revised circuit diagram below).

The input to the filter is the junction of VR102 wiper and R120. If you want to make the filter unity gain, remove R121 (open circuit) and change R120 to 10K, the response is unchanged. You'll note - no inductors required.

R120/121 could be 22k with very little change in the response, but it would still have a gain of 0,5, -6db.

Steve A.
Attachments
SSTV Demod V2 1.gif
SSTV Demod V2 1.gif (15.16 KiB) Viewed 10642 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: Retro SSTV anyone?

Postby Steve Anderson » Fri Sep 18, 2015 6:23 pm

Klaas, yes I did notice your 'Testbeeld' file was a bit low in frequency but I was too polite to mention it. But now you have I attach the spectra...

The reason I even looked at the spectra was that on my demodulator the syncs seemed a little 'crushed' and I thought this was unlike Klaas.

The crushing is due to the sync frequencies are now in the non-linear region of the demodulator as can be seen by the V/F chart below. The voltages would now be different (see previous post) but the shape of the curve would still be the same.

I guess this should be quite a simple fix.

I have changed the values of the filter in the modulator and although there is an improvement in the ringing I now am suspicious of the filter in the demodulator (above). Changing it to a Bessel filter would probably let too much of the rectification products through, I need to muse on this...while I do I'll get on with the contouring problem.

Steve A.
Attachments
Testbeeld Spectra 1.jpg
Testbeeld Spectra 1.jpg (69.07 KiB) Viewed 11048 times
Robot 70 Linearity 2.gif
Robot 70 Linearity 2.gif (4.88 KiB) Viewed 11048 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: Retro SSTV anyone?

Postby Klaas Robers » Fri Sep 18, 2015 7:12 pm

Thank you Steve, that is something I wanted to ask you and now you did it unasked ! I got already the impression that my frequencies were not correct. Yesterday I got the sync detector working again, after > 40 years. It works with an LC tuned circuit on 1200 Hz, which is driven with the limited input signal and which amplitude is detected. Extra capacitors are added via a rotary switch, to tune the resonance to 1000, 1100, 1200, 1300 or 1400 Hz. Then, when using the now built in SSTV generator, it struck me that I had to tune the sync resonator to 1100 in stead of 1200 Hz.

I had seen as well that the output voltages of the video signals were too low, when compared to the tone generator. This is not too strange as the frequencies are dependent on the +6V supply voltage. I don't know any more which supply I used in 1973, but now I made a supply using a 7806, much better than in those days.

It is very simple to adjust the frequencies. The FM generator is a unijuction oscillator, with a capacitor that is charged via two adjustable current sources:
- one source gives a constant current and defines the frequency of the sync (1200 Hz),
- the other source is switched off during sync and is modulated with the video signal.
- This one has an offset for the 1500 Hz black level.
I will add the circuit diagram. It might be too wide for the web page, then please load it down for viewing, and a photo of the circuit board. The generator has 5 of these boards, standing upright, in a small rack.

So I can adjust the sync frequency, the black frequency and the white frequency. These potentiometers are small, era 1970 pertinax trimming potentiometers, so they may also need some adjustment. I will do that today and come back with an SSTV audio file. Then please measure the frequencies again for me.
Attachments
Schema AFM-oscillator.GIF
(10.92 KiB) Not downloaded yet
Foto AFM-oscillator.JPG
Foto AFM-oscillator.JPG (81.56 KiB) Viewed 11044 times
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Retro SSTV anyone?

Postby Steve Anderson » Fri Sep 18, 2015 7:57 pm

Klaas Robers wrote:...I will do that today and come back with an SSTV audio file. Then please measure the frequencies again for me.

Sure, no problem at all. I'll also do a screen shot, with any luck the sub-carrier frequencies being a little higher the 1200Hz grating may be of a greater amplitude - but maybe not as much as we'd like.

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: Retro SSTV anyone?

Postby Klaas Robers » Fri Sep 18, 2015 10:19 pm

I readjusted the FM oscillator and that was needed.
- The sync was on 1074 Hz, that was adjusted first to 1200 Hz.
- Then still black was too low and white was too low as well.
It was quite some change. Both should be Ok now. Steve, surprise me.

Therefor I post two SSTV files in 8 bit 8 kHz:
- The well known Test picture
- a checker board. I will add a picture of it.
This is a headace file, as the line sync is preceded by a black line and followed by a white line, or reversed. So the sync will be moving, even the "center of gravity", and it is almost impossible to get nicely positioned rows of blocks. However it is a nice file to check the frequencies, as there is only sync 1200, black 1500 and white 2300.
Attachments
Testbeeld2.wav
SSTV black and white Test frame in 8-bit and 8 kHz
(234.5 KiB) Downloaded 493 times
Dambord.wav
SSTV black and white Checker board in 8 bit and 8 kHz
(234.5 KiB) Downloaded 477 times
Dambord.GIF
Checker board as it should look like. The grey border is NOT part of the picture!
Dambord.GIF (1.96 KiB) Viewed 11036 times
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Retro SSTV anyone?

Postby Steve Anderson » Sat Sep 19, 2015 1:17 am

I got those files, thanks.

A quick check with the scope verifies they are spot-on, see below - no time to re-name the files - tomorrow.

But on the screen there's something very wrong, and MMSSTV can't seem to understand the signal. It's the same with my files too. I suspect an earth/0v has dropped off somewhere or something else is wrong with the untidy lash-up here.

Ignore the cyan trace, I just forgot to turn it off.

I'll have a go at sorting things out in the morning, it's just after 9pm here...

Steve A.
Attachments
TEK0082.gif
TEK0082.gif (43.06 KiB) Viewed 11028 times
TEK0083.gif
TEK0083.gif (43.92 KiB) Viewed 11028 times
TEK0084.gif
TEK0084.gif (44.55 KiB) Viewed 11028 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: Retro SSTV anyone?

Postby Klaas Robers » Sat Sep 19, 2015 1:46 am

Thanks Steve sofar.
I see that also on your oscilloscope the white for Testbeeld2 is slighly lower than for Dambord. I noticed that already here. Testbeeld comes from a different source than Dambord. The latter comes from my pure hardware generator, lots of DTL flipflops, no more than two in each dual in line chip. Testbeeld comes from the newly built 1 MB EPROM source. This one has a small cermet potentiometer to adjust the amplitude of white. I expect that MMSSTV will show that Testbeeld white is not 2300 Hz, but slightly lower. I have readjusted that small potentiometer already and now both are the same.

I see that you have lots of noise on the video. I don't see that on my analogue oscilloscope, but I haven't looked at my recording, only at the output signal of the generator......

Something to do.

1th edit: In the mean time I did some experiments.
- Indeed, when playing back my own files via the SSTV demodulation circuits, the demodulated video looks noisy as well.
- My first thought was: the lousy 8 kHz and the lousy 8 bits.... but
- when recording at 16 bits and at 44.1 kHz and playing back, the demodulated video looks exactly as noisy.
- I also looked at the played back FM subcarrier and it was also not as clean as the FM subcarrier directly from the generator.

I have no idea what happened here. The laptop was running free of the mains, on its built in batteries. Do you know how the recording level of a laptop is adjusted? I don't have a line input, so I have to use the microphone input. I have to experiment with that. May be THAT is too noisy.

2nd edit: It is not the 8-bits and 8 kHz. It was the attenuator to the microphone input. It was far too high ohmic towards the PC-input. Now I used 500K to 50 ohm and the 50 ohms connected to the microphone input suppresses the excessive noise. Here is a new 6 x 6 checkerboard with much less noise.

And there was no automatic volume control. I had to select the resistors by hand until the amplitude was good enough.
Attachments
Dambord2.wav
(234.5 KiB) Downloaded 444 times
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Retro SSTV anyone?

Postby Harry Dalek » Sat Sep 19, 2015 9:25 am

Hi Klass it is possible to use your laplops Mic input ....i have had lots of laptops depends what you have what you have to do .
https://www.youtube.com/watch?v=c7i6LJs3b6E

This above works if you are lucky and your laptop can record stereo...
\

IF only mono a litte circuit

http://www.epanorama.net/circuits/line_to_mic.html
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: Retro SSTV anyone?

Postby Steve Anderson » Sat Sep 19, 2015 2:02 pm

I love to hate those 3.5mm connectors used for audio I/O on PCs and other devices, about as reliable as the British summer.

The gain of the mic and line inputs (if any) is controlled within the codec so the gain is going to be controlled in software usually via an on-screen knob or slider. Even in the humble and now very dated AC97 codec there's a whole load of useful I/O that we never have access to unless you want to 'knife and fork' your mainboard.

Klaas, Dambord2 is much better but the spectra has a load of junk above 2300Hz - what that might be I have no idea. The waveform looks clean and the screen-shot is OK (ignore the sync-jitter and the junk top and bottom, my problem). I'm always suspicious of free third-pary software until I'm convinced its reliable, I would tend to ignore this.

It might be worthwhile re-doing Testbeeld again, the second version as you mentioned was very noisy in the mid-grey areas.

My problem late last night was as suspected, some earths/0v had drifted off into outer space.

Steve A.
Attachments
Dambord2 Screen Shot 1.jpg
Dambord2 Screen Shot 1.jpg (135.07 KiB) Viewed 11015 times
Dambord2 WFM 1.gif
Dambord2 WFM 1.gif (43.15 KiB) Viewed 11015 times
Dambord2 Spectra 1.jpg
Dambord2 Spectra 1.jpg (19.49 KiB) Viewed 11015 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: Retro SSTV anyone?

Postby Klaas Robers » Sat Sep 19, 2015 9:06 pm

Hi Harry,
this is more or less what I did, but I have to optimise it. My recording software (Gold Wave) has a constant recording level. In fact I like that, as it is now ME that defines the amplitude of the recorded signal. But i did it in a wrong way, that is in respect to noise.

Steve,
the screen shot looks very fine. And the squares are aligned not bad at all. Also the demodulated video wave form looks good. Are the levels DC-wise comparable with what you are used of?

Then I don't understand the "spectrum" of MMSSTV. The sync is a very narrow peak, this looks good, so why are the blacks and the whites so wide? I expect that the program is not a real spectrum analyser, but it measures the time between all successive zero crossings and makes a histogram of this. That is what I should have done. In a spectrum amplitude is also taken in aspect, but that is of no use for FM signals.

Steve, how much time of SSTV signal do you need to make a "spectrum? If it is short, one second or so, then I can send a short phragment at 48 kHz and 16 bits resolution. But no, this cannot be the reason. Have you looked at the wave form of the replayed audio? These should be sine waves or almost sine waves. That is what I see on my oscilloscope when I look to the headphone output of the PC.

Here is a new Testbeeld file, however in 32 kHz 8-bits.
Attachments
Testbeeld3.wav
(937.62 KiB) Downloaded 431 times
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Retro SSTV anyone?

Postby Steve Anderson » Sat Sep 19, 2015 10:36 pm

My fault again, overloading the input to the PC line in (again), reducing the signal to 1.0V p-p results in a much cleaner resultant spectra. The peaks at 1500 and 2300 are quite broad compared to the modulator I'm using here.

Looking at the sub-carrier sine wave of Danbord on the scope it is quite 'jittery', as if some noise is getting into the VCO control line or the unijunctions are being affected in some way.

It may be worth generating a file of a whole black frame to see if any harmonics of sync or black are creeping in too.

I'll have a go at Testbeeld3....

Around 10 seconds of 16-bit/48kHz mono should be enough, it should be a file of only around a megabyte.

The 'contouring' problem turns out not to be contouring from the multiply routine. I added the required hardware (one pot, one resistor), removed the multiply routine - and it's still there! There appears to be two factors, one is our good old friend RAM I/O timing differences, but it's the lesser of the two. The second one I'm still trying to pin down.

Steve A.
Attachments
Dambord2 Spectra 2.jpg
Dambord2 Spectra 2.jpg (19.21 KiB) Viewed 11003 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

PreviousNext

Return to Cop McDonald's FM SSTV System

Who is online

Users browsing this forum: No registered users and 1 guest

cron