PIC 405 line pattern generator

Hi Folks,

Only just read the recent posts in this thread and two give the team concern.

The only bit that could do with being added is the "hack switch" that allows copy protected chips to be copied

Does this switch actually exist? if so I could really do with knowing about it.

I would hate to see an otherwise excellent project/thread that Frank has invested a great deal of time in, closed due to others members discussions of illegal ways to hack IC's.

This practice is not something we wish the forum to be associated with. Speaking as owner and on behalf of the moderation team who could be equally held accountable by condoning such posts, any further discussion on hacking chips will result in the thread being closed. I hope you understand our stance in this matter, lets keep posts to the matter in hand which is creation/construction of a PIC 405 line pattern generator.

Note: The UK enacted the Copyright, Designs and Patents Act, 1988, c. 48, § 213, I believe the act/ law fully protects chip topography and code

Understood, perhaps I should have clarified with a little context, I have several of my own projects which I have lost the code for in the great hard disk crash of 2012, I'd really like to get the original hex code back but obviously there isn't a 'legal' method of doing so that I know about.

Thanks, Dave.

I am after having quite an interesting battle with a new generator that I built to find out, why, what looked like a hum bar was appearing on screen.

All TVs that it was tried on displayed a perfectly steady test card C from the Aurora. The generator was tried on three different modulators the Aurora used as modulator only, a ETF modulator and another home built modulator that is similar to the Aurora's modulator. No matter what combination of modulator or TV that was tried a slight hum bar could be seen traveling up the screen.

The hum bar did not shade or brighten the area that it passed through as it moved up the screen, it just pulled the lines slightly to the left and appeared to moved the lines apart as it passed through them. This could be seen best at the top and boom of the picture where it would lengthen the castellations as it passed through them.

When scoped no mains ripple could be found on the output or the supply.
In order to eliminate any ground loops the generator and modulator was run from battery's but this made no difference either. Attention then tuned to the microcontrolers oscillator and to get rid of any suspicion the resonator was changed for a crystal. I then noticed that this microcontroler unlike the microcontroler that I was using at the start could run a crystal up to 25 MHz, I fitted a 24 MHz crystal and changed the software to suit. this removed the need for the the X4 PLL and any jitter it introduced but still no joy.

In order to confirm that it was mains hum and nothing else I decided to inject some mains frequency into the misbehaving generator. This was done by connecting the secondary of a small 9V transformer between 0V and the transistors base, a 15k resistor was put in series with with the transformers secondary to limit current. The primary of the transformer was connected to mains via a variac. The variac was wound up until a hum bar could be seen on screen and sure enough it tracked the hum bar from the generator in both direction and speed, at least now I knew for sure that it was mains related.

Next I found another Sony 9-90UB that I could run from DC. This was ran from the bench power supply and when the generator was tried on it the hum bar had disappeared. Then running the TV from mains the hum bar reappeared. Finally some progress!

Then I fitted a 'Y' splitter to the output of the modulator so I could feed two TV's, one ran from mains the other from DC. There was no hum bar on the one ran from DC while at the same time the one connected to mains had a hum bar. On the face of it this would appear to suggest that the hum was been introduced after the modulator but why then would one generator produce a hum bar while another would not?

After another bit of prodding around the cause was found, any ideas what it was?
I will post the answer in a while.

Frank

I once had a problem that turned out to be an earth/neutral transposition on the mains socket, the RF source also being grounded. In my case, I discovered it by accident when I was lucky enough to observe a marked decrease in the effect when a 3kW fan heater was turned down to 2kW as a very cold office started to warm up!

See:  https://www.radios-tv.co.uk/an-elusive-hum-problem/

This couldn't happen with an un-isolated chassis, of course, because of the aerial isolator.

So my money is on some form of earth loop.

Hi Terry
I had a look at your link, that was an unusual fault and just goes to show how something seemingly unconnected can trow a spanner in the works.
Right to the end I too thought that it had to be a hum loop of some sort and I had chased my tail trying to find it for quite a while.

After hooking the generators up to the scope again and comparing the output of a good and misbehaving generator a difference could be seen and the penny finally dropped.

I said in an earlier post that the frame rate was not exactly 50Hz. I had thought that this would not be a problem but I was wrong.
The difference between the two outputs displayed on the scope was when the scope was synced to mains, the misbehaving generator output passed across the display much faster that the good one. In other words it was more out of sync with the mains than the one that didn't display a hum bar.

The misbehaving one took 4S to cross the screen while the good one took 9S. I then checked the output of the Aurora and it took 30s to cross the screen.
I often heard that TV transmissions were synced to the mains but never realized that if it wasn't that it could cause such a problem.

All the good generators used a resonator from the same lot, while the misbehaving one was from a different lot and must be closer to 6MHz as substituting a crystal made no difference.
I set about seen what could be done to improve the situation and after doing the maths it looked like I should be getting a better frame rate with a crystal than what I was getting.

Out with the scope once again and checked the misbehaving generator, flicking through all the patterns revealed that apart from the two test card patterns all others patterns were moving across the screen at a much slower pace. Scoping a good generator showed the test card patterns moving across the screen in one direction (<50Hz) while all other patterns were moving in the opposite direction (>50Hz). This could only result from the test card patterns having more lines than the other patterns.

Sure enough after checking through the code two extra lines were found in a portion of code that was common to both test cards. The lines that forms the top and bottom castellations are produced in blocks. The last line in the bottom castellations is a modified line to allow for the interlace etc. This line is not part of the bottom block and should have been subtracted from it but wasn't hence an extra line was added to each field.

After removing the extra lines the generator is producing a much better picture, tough there is still a tiny bit of hum on it if you look at it very closely as it is still more out of sync with the mains than the first generators that I built. It would appear that the hum bar is more intense the further the frame rate is from 50 Hz and if it gets reasonably near to 50 Hz it diminishes altogether. This possibly just an optical illusion?

In the light of this, to get the frame rate closer to 50 Hz I will reduce the total line length from 99 uS to 98.83 uS. This will be done by reducing the sync pulse from 9 uS to 8.83 us. When I get it done I will make it available on the website as a download.

In the meantime I have updated the software download on the website to one which now has the correct number of lines in all patterns.

Frank

I often heard that TV transmissions were synced to the mains but never realized that if it wasn't that it could cause such a problem.

Oh yes it was!

Here's the history: in the beginning, transmissions were asynchronous, i.e: mains locked. The advantage of this was that stationary hum on a picture can be virtually invisible to the viewer, provided it is not excessive. No doubt that was very important in the early days when electrolytic capacitor technology kept values relatively low.

However, this could lead to some very interesting effects. With many sets, hum could affect one or both timebases, leading to slight S bends on verticals and linearity variations over the period on a frame - as you have now witnessed, Frank! Any insert into a live programme, whether from another studio, telecine or this new fangled video tape, would be from an external source which, although also mains locked, had a two in three chance being fed from a different phase, so the field sync might vary ±16.66mS from the studio sync. It might be better to say from +16.66mS to +33.33mS here. To avoid a nasty jump, the studio sync would be switched from mains lock to the incoming video, so the sync generator was allowed to free run until the field sync pulses agreed. What happened in the receiver was any combination of a number of effects, which would vary from receiver to receiver. This happened fairly quickly, so quite violent changes in line speed could occur, together with the other effects.

I was sitting reading with my back to our Ekco TV (a T281, I think) on one occasion when I heard the sudden change in line frequency and, on instinct, glanced round to catch the tail end of the characteristic field linearity convulsions. My mother said "How did you know that was going to happen?". (She couldn't hear a 10.125kHz line whistle, even with her head virtually inside the cabinet - I tried!)

Asynchronous working is not suitable for colour transmission - everything has to be synchronous with the colour subcarrier - so, after the report of the Pilkington report on the future of Television, which gave the go ahead for BBC2 on 625-lines and, later, colour, the BBC decided to switch to synchronous transmission with startling results because customers could now see the hum effects continually moving, which made them highly visible!

It was quite a busy time, particularly with older sets ...!

Hello Terry

Here's the history: in the beginning, transmissions were asynchronous, i.e: mains locked.

That is not asynchronous, that is synchronous! To be asynchronous, there would have to be no mains lock.

Regards,

Dave.

Oops! Obviously got it wrong way round! 

Hi Terry
Thanks for the explanation and history, it shines a lot of light on everything.
it's interesting to know how things have developed and why.

Frank

As somebody who has battled with programming PICs for many an hour, I think you should be congratulated on your work. It really is marvellous to see, I'm very impressed. Well done.

Hi mkstevo
Thanks for your comments. I enjoy playing with PIC's but as you say it can be a battle sometimes. I have spent many hours trying to find out why something which seamed totally irrational was happening, inevitably it always turns out to be a school boy error that I have made, but It's all part of the fun.

I should soon have the software done with shorter lines.
I had intended to reduce the total line length to 98.83 uS but this turned out to be a non runner.
Reducing the length from 99 uS to 98.83 uS would be done by shortening each line by one instruction cycle, which isn't a problem for the lines however in ordered to keep the timings correct each broad pulse in the field sync needs to be reduced by half that amount which cant be done.

Instead I will reduce each line by two instruction cycles to 98.66 uS. The broad pulses can then be reduced by one instruction cycle.

Frank

I have the new software finished, it can be found at that same link and I have also built a more compact version that uses phono sockets in place of BNC.

pictures below.

Frank

That is... astonishingly small, Frank! Amazing...

As soon as I've got this Dynatron chassis off the bench, I'll do the long-awaited test report on the generator. Which will also force me into getting my Sony CVM9-90 working!

I have the new software finished, it can be found at that same link and I have also built a more compact version that uses phono sockets in place of BNC.

Very smart and very elegant!

Thanks chaps.

The next thing that I will build to go with it, is a System A modulator that will cover all 13 Channels. It's a design that I fond over on UKRRRDF by David Robinson.
When the time comes I will start a new thread for that.

Frank

Thanks chaps.

The next thing that I will build to go with it, is a System A modulator that will cover all 13 Channels. It's a design that I fond over on UKRRRDF by David Robinson.
When the time comes I will start a new thread for that.

Frank

That sounds like a truly excellent addition to the arsenal, Frank. I've looked at the design and it uses the same modulator IC as the Aurora standards converter.

I'd better get the spectrum analysers sorted out (TBH the Marconi works, but is limited to 400MHz: the HP goes to 1.5GHz but is faulty).

Hi Cathovisor
That's the one, It uses two MC44BS373CA modulator chips that are getting difficult to find. I bought my lot on ebay about a year ago, there were several suppliers listed back then. I took a look today and cant find any however there are still some listed on Aliexpress.

Being able to switch between channels is a great advantage particularly when you have both London and Birmingham TV's.
I have it practically finished and it appears to preform well.

Frank

A pair of readily available NE612 / SA612 will work, though only to lowest Band III channel.
Cheap DIL on ebay.
The trick to get AM is to pull down ONE only of the balanced input pins. Try 22k I think. Or 47k variable in series with 2k2. Decouple the pin for LF video or audio. Use series capacitor on other input pin to video / audio.

The Oscillator is simple to use with ceramic, crystal or LC. A varicap can be added for FM (525, or 625 or other) or DC fine tuning. Phase modulation can be done by adding a proportion of LO out to the Mixer out.

Works fine for LW /MW audio modulator too. Without unbalancing the input (or adding in the Osc Out) it's a Double Side band Suppressed carrier mixer and Local oscillator.

Hi Michael
That's an interesting little chip that I have not come across before, it's good to know there are alternative ways.

Before I settled on the MC44BS373CA I was thinking of using a pair of MC1946 and two AD5891 DDS modules as there oscillators. It should have been able to do all the Band 1 Channels which is all I need really.

Frank