Update to SSTV demodulator

Forum for discussion of SSTV topics. Slow Scan television (SSTV) is a picture transmission method used mainly by amateur radio operators, to transmit and receive static pictures via radio in monochrome or colour.

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Update to SSTV demodulator

Postby Steve Anderson » Sat Sep 14, 2019 6:54 pm

The original SSTV demodulator that I used was simply an update on an old design by Robot from the 60s or 70s. It's not perfect, but the updated version performed quite well.

My biggest gripe was the overshoot introduced by the output Butterworth filter, otherwise I was quite content with it. Sure it could be improved in various areas, but that overshoot, particularly on the extreme left of the picture really bugged me. So here's my attempt at something better...

Generally the demodulator is exactly the same as before, the departure is in that culprit output filter. Here it's replaced with a software FIR (Finite Impulse Response) filter. This should be phase-linear suppressing that darn overshoot. After that I'll try and address the other concerns raised. However here's my first iteration...yet to be built...

Note the output filter as shown is the same as the original, this will be changed, hence the red text for component values.

The processor used is simply some I had in stock, but it needs to be a PIC18 device for the internal hardware multiply device. PIC16s and the like don't have them and a software routine is probably too slow. The micro has a USB internal engine which is not used here...but maybe in the future...
Attachments
SSTV v3 Demodulator 1-Model.gif
SSTV v3 Demodulator 2-Model.gif
SSTV v3 Demodulator 2-Model.gif (46.8 KiB) Viewed 12929 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: Update to SSTV demodulator

Postby Steve Anderson » Sat Sep 14, 2019 7:31 pm

For the SOIC devices a local company here makes adapters for SOIC to 0.6" DIL, ideal for Veroboard and the like. Not only 8-pin versions but practically all other versions of SMD devices. They have a US distributor, but not one in Europe. Also most available from Futurlec in Oz. (Spelling corrected now)...and in English...sort of...

https://www.futurlec.com/SMD_Adapters.shtml

The individual pieces simply snap apart, pre-scored.

The website (as shown on the PCB) is almost entirely in Thai...sorry guys...also prices are all in Thai Baht, then add 7% VAT. Around 30THB to the US$, 38 to the UK quid.

If you know of other sources for similar products please post them here...we can't afford to be locked out of newer/better devices just because they're not 0.1" pin-spacing...

Steve A.
Attachments
SOIC8-DIL8 Adapter.jpg
SOIC8-DIL8 Adapter.jpg (54.99 KiB) Viewed 12928 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: Update to SSTV demodulator

Postby Harry Dalek » Sun Sep 15, 2019 6:00 pm

Hi Steve good to see your back to SSTV reading if i have got it right the TLO84 demodulator circuit has been replaced with the pic 18 circuit .
Is this for in time the SSTV to 625 line converter ?
Any case interesting i am watching .
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: Update to SSTV demodulator

Postby Steve Anderson » Tue Sep 17, 2019 9:09 am

Well, the original SSTV demodulator could be used anywhere you needed to turn the FM-based SSTV signal into a baseband signal - for a display, whatever. However there were shortcomings within it - for me it was/is that output filter that adds the overshoot on large signal changes, say from syncs to peak white at the start of a line. That filter has a lot to do, pass decoded video up to around 1kHz and reject everything above 2kHz. In the process the signal has artifacts introduced by the filter - i.e that overshoot.

In the my original posting/copy of the Robot demodulator the input waveform is frequency doubled by the action of the full-wave rectifiers (as per the usual arrangement with power supplies). So syncs become 2x1200Hz (2400Hz) and white becomes 2x2300Hz (4600Hz). This is where this filter earns its pay-cheque. Getting a sufficient degree of rejection of frequencies above 2kHz, yet passing those below 1kHz isn't so easy whilst all the time retaining the correct phase relationship between the signal components, which is tricky with analogue filters.

So I'm hoping, really hoping, that this will be the answer, even if not definitive, I hope an improvement.

That to me was the major issue, but as Klaas pointed out there are other things that could be improved. For me, if I can get rid of that overshoot I'll be a happy bunny! However, as Klaas mentioned the 'faulty' output filter does add an element of detail sharpening to the result - as ever a trade-off...

Not having used or designed software defined filters before - I'm straying into new areas. They certainly hold a lot of promise, it's a case of 'wait and see'...

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: Update to SSTV demodulator

Postby Steve Anderson » Tue Sep 17, 2019 2:10 pm

Having re-read your posting Harry, the micro and output filter replace all after VR202 in the original diagram, all the upstream bits are still required, IC201/202 and associated components. I'll knock-up a combined diagram and post it in a short while...

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: Update to SSTV demodulator

Postby Steve Anderson » Tue Sep 17, 2019 2:46 pm

Having some hassles with Autocad, I'll do a test posting to see if all is OK...

Steve A.

Well it posts OK, just dunno why the lines are so fine....I'll try another...
Attachments
SSTV v3 Demodulator 3-Model.gif
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: Update to SSTV demodulator

Postby Steve Anderson » Tue Sep 17, 2019 2:57 pm

Well, I'm going to have to find out what's happening here...second part is the same...

Steve A.

I may have to wait until my Autocad guru returns from Europe this coming weekend to sort this out...
Attachments
SSTV v3 Demodulator 4-Model.gif
SSTV v3 Demodulator 4-Model.gif (36.7 KiB) Viewed 12897 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: Update to SSTV demodulator

Postby Harry Dalek » Tue Sep 17, 2019 4:17 pm

OH i see Steve i was not sure what the new ic's did ,your posts give me an idea now of what you have planned .
Mainly for us that love classic SSTV an improvement on the system keeps it alive so that's a good thing .
Am i right in saying again reading your posts you are yet to construct this add on to the demodulator and also the program for the PIC ic .
I can see it could be used just in a classic p7 monitor or a converter .
BTW i never knew Klass was a Professor I came across a PDF on motor speed control and it was mentioned ,You could be one as well as far as i know ! :wink:
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: Update to SSTV demodulator

Postby Steve Anderson » Tue Sep 17, 2019 6:20 pm

My interest in SSTV started with an article in the early 70s in the magazine 'Electronics Australia', I was living in a suburb of Melbourne at the time (Frankston). I can't recall what the item was about except it was SSTV-related. The idea of being able to send pictures (albeit stills) via a voice-circuit fascinated me. It still does today even with the Internet, mobile phones, jpgs and all the rest of it. In concept it is so simple...and as was done 'back then' using valves/tubes only if you wished.

Yes, this has yet to be built, but the analogue part (the first diagram) I'll simply re-use the first version (no changes envisioned there), so all I have to build is the micro, D-A and the 'new' output filter, the old Butterworth filter is simply disconnected. Though I'll probably leave it in circuit to see the difference between it and the 'New Improved Version'.

The D-A part doesn't need to be 12-bit, it's just I have them here in stock, in fact the data sent to it is only 8-bit. So with the spare outputs on the micro it could be adapted to use something like a DAC-08 parallel D-A. However they require split +/-12V supplies which for some reason puts a lot of people off. I actually prefer them as they make designing DC-coupled designs so much easier. No -3db point at low frequencies, a big plus-point with NBTV and SSTV.

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: Update to SSTV demodulator

Postby Klaas Robers » Wed Sep 18, 2019 8:53 pm

Steve, do I see that you keep the analogue low pass filter in the circuit? If you do that you keep the non flat phase characteristic and the ringing overshoot only after the transient. However you might program the FIR coefficients such that the FIR-filter compensates for the ringing overshoots. I don't know yet how to do that, but may be "time inversion" is a way.

I expected that your FIR-filter will replace the analogue butterworth filter. It is still a pity that my experiments earlier gave no usefull results.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Update to SSTV demodulator

Postby Harry Dalek » Wed Sep 18, 2019 9:33 pm

Oh i hope it works out Steve ,i am always interested in improved SSTV for classic Slow scan ~!
On the dual supply i think most people think you have to have a 3 wire secondary transformer i use this solution with the single diodes all the time with regulator s ...used it in the binocular monitor and other projects...so i know that works well .
Not sure about the full diode bridge idea not tried that its new to me apart i know it does not work at all with out those caps on the AC ?
plus-9v5.jpg
plus-9v5.jpg (16.98 KiB) Viewed 12829 times
Attachments
dual-dc-power-using-2-terminal-transformer.jpg
dual-dc-power-using-2-terminal-transformer.jpg (37.7 KiB) Viewed 12829 times
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: Update to SSTV demodulator

Postby Steve Anderson » Thu Sep 19, 2019 2:22 pm

Klaas Robers wrote:I expected that your FIR-filter will replace the analogue butterworth filter.


That is the idea, the FIR filter is fed from the output of the adder/subtractor (IC01d, pin 14) with a trimpot to set levels. In other words it replaces the Butterworth filter (IC103 and associated components), though I'll leave the Butterworth version in circuit so I can compare the performance. The most recent circuits above show the feed to the FIR filter taken from the existing Butterworth filter input..i.e. the filters run in parallel, not one fed from the other. The link is shown as point 2A in the diagrams above. This is only to compare the performance of one against the other. If the FIR filter performs better than the Butterworth the Butterworth will 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

Re: Update to SSTV demodulator

Postby Steve Anderson » Thu Sep 19, 2019 2:44 pm

Harry, yes the first circuit is valid and works as advertised. BUT the ripple is now at 50Hz (or 60) rather than 100Hz (or 120). Which means more ripple the regulators have to get rid of, and for the same load generally larger capacitors are required and therefore higher peak currents in the diodes and transformer, this results in higher heating of the transformer windings. Generally half-wave rectification is less efficient than full-wave. If the load is very light, then OK, but if you're wanting to use the transformer to it's full capacity then full-wave is far better.

I've not had a problem with getting centre-tapped secondary transformers, e.g. 12V-0V-12V rather than just a single 12V secondary. If you're using old phone-chargers or other 'wall-warts' then you may get stuck.

As for the second circuit, I've never tried that arrangement, I've seen similar a few times in the past, but it's messy, uses more components, it simply doesn't look 'right' to me. Get the right transformer for the job is my advice. Or get a second one. In my opinion it's a 'botch'.

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: Update to SSTV demodulator

Postby Steve Anderson » Thu Sep 19, 2019 5:38 pm

Klaas, the output filter after the D-A will be changed, probably to something like a Bessel with it's much gentler phase characteristics. Which is why the component values on the circuit diagram are currently in red. There's a lot of 'cut-and-try' involved here.

If the SSTV subcarrier was (say) 10-12kHz things would be a lot simpler. But that isn't case...

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: Update to SSTV demodulator

Postby Harry Dalek » Thu Sep 19, 2019 5:55 pm

Steve Anderson wrote:Harry, yes the first circuit is valid and works as advertised. BUT the ripple is now at 50Hz (or 60) rather than 100Hz (or 120). Which means more ripple the regulators have to get rid of, and for the same load generally larger capacitors are required and therefore higher peak currents in the diodes and transformer, this results in higher heating of the transformer windings. Generally half-wave rectification is less efficient than full-wave. If the load is very light, then OK, but if you're wanting to use the transformer to it's full capacity then full-wave is far better.


What i do is use the largest capacitors i have before the regulators even used a high capacitors after the regulators and never noticed a problem ,i understand nothing is for free you have to adjust for the lack of full regulation ...if you are stuck its an easy fix than having to buy another center tapped Transformer .

I've not had a problem with getting centre-tapped secondary transformers, e.g. 12V-0V-12V rather than just a single 12V secondary. If you're using old phone-chargers or other 'wall-warts' then you may get stuck.


Some times its good to have the 12V-0V-12V AC option on your bench power supply if you don't want to buy another transformer right away ...
Have to say I did pick up a lot to tips on power supplies from your Anderson monitor build .

As for the second circuit, I've never tried that arrangement, I've seen similar a few times in the past, but it's messy, uses more components, it simply doesn't look 'right' to me. Get the right transformer for the job is my advice. Or get a second one. In my opinion it's a 'botch'.

Steve A.
[/quote]

i feel looking at it again and your advice i think i will stick to the 2 diode idea ..
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

Next

Return to SSTV

Who is online

Users browsing this forum: No registered users and 6 guests

cron