Negating LF problems of NBTV.

Forum for discussion of electronic television. Generally, stuff to do with CRTs and not using mechanical displays.

Re: Negating LF problems of NBTV.

Postby Steve Anderson » Wed Apr 01, 2020 3:46 pm

Now, there is a possibility to use a processor that is exactly the same except less program memory. The PIC18F26K22 has 64kB (32,768 single word instructions) of program memory, I can't see any way that amount would be needed. It's siblings are...

PIC18F25K22, 32kB (16,384 single word instructions)
PIC18F24K22, 16kB (8,192 single word instructions)
PIC18F23K22. 8kB (4,096 single word instructions)

The smaller the memory, the cheaper the chip. I'm thinking that I could get this into 4,096 single-word instructions...we'll see. Some instructions are double-word (32 bits) but not often used.

Also as the program memory decreases in size, so too does the internal RAM, but very little is required in this application.

It is also possible to use the internal 10-bit A-D for the video channel - I'll have a look into this...unfortunately these micros don't have an equal resolution D-A, so on the output side things will be the same.

You start out with an idea or concept - then simplify...if possible!

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: Negating LF problems of NBTV.

Postby Klaas Robers » Wed Apr 01, 2020 10:50 pm

I know this effect Steve. In the past we had EPROMs 2716. And then there came 2708 chips. They had half the storage capacity. But if you tried to program them beyond this limitation, it worked! The 2708 chips were in fact 2716 chips that had "a problem" in the higher part of the memory space. You never knew where it was, so a certain chip would work, while another chip woud cause problems. On the other hand, the faulty 2716's could still be sold as a 2708.

But if there was a problem in the first 8k.....

I expect that there was someting, a bonding pad, that when grounded, inverted the highest address line, so that the second 8k was swapped to the first 8k. During production it was grounded or not.

So why not using the cheaper ones if you don't need the program space?


Afterwards editted: No, the 2708 trick was simpler. There were two types: 2708 and 2708A. The highest address line pin of the 2716 should be kept 0 (grounded) for the 2708, and made 1 for the 2708A. So nothing was needed. Clever guys.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Negating LF problems of NBTV.

Postby Steve Anderson » Thu Apr 02, 2020 12:52 pm

I've looked into using the internal A-D for the video channel but there isn't a spare pin available! I could free up one of RC0-RC3 (IC101 pins 11-14) but it doesn't really make the device simpler and the cost savings are nothing much in terms of the overall project cost. So I'll leave it as is.

With a change to the 40-pin versions of the same chips (PIC18F46K22 etc.) there would be no problem at all but I'm reluctant to go down that route.

The reason I'm using the 64kB version micro in the prototype is simply I have them in stock and given I can get the code into a smaller memory device, then the change will be made. Nothing in terms of the circuit/hardware would need to change other than the micro itself, they're all available in 28-pin DIP packaging in this 'family'.

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: Negating LF problems of NBTV.

Postby Dave Moll » Thu Apr 02, 2020 6:26 pm

Klaas Robers wrote:In the past we had EPROMs 2716. And then there came 2708 chips. They had half the storage capacity. But if you tried to program them beyond this limitation, it worked! The 2708 chips were in fact 2716 chips that had "a problem" in the higher part of the memory space. You never knew where it was, so a certain chip would work, while another chip woud cause problems. On the other hand, the faulty 2716's could still be sold as a 2708.

But if there was a problem in the first 8k.....

I expect that there was someting, a bonding pad, that when grounded, inverted the highest address line, so that the second 8k was swapped to the first 8k. During production it was grounded or not.


I'm sure you're all aware, but just to clarify that the 8K and 16K here are bits rather than bytes, so a 2708 is 1KBytes and a 2716 is 2KBytes.
User avatar
Dave Moll
Anyone have a spare straightjacket?
 
Posts: 460
Joined: Tue Feb 27, 2007 9:11 am

Re: Negating LF problems of NBTV.

Postby Steve Anderson » Thu Apr 02, 2020 7:24 pm

Mentioning these old EPROMs reminds me of the first micro I ever programmed. A Motorola MC6809. Well the micro couldn't be programmed, an external EPROM held the instructions/code. It also required external RAM, and for this sort of application an external UART. None of which could reach the speeds needed here. Quite simply we've got it easy these days!

I can still get the MC68B09 (the 'high speed' 4MHz version) here at around US$4.20, but with the rest of the 'stuff' needed it simply doesn't make sense.

Then there's issue of an assembler that turns your code into machine executable bytes. There are free versions available on the web, but how good are they? The one I used in the 80's was from Crossware, still selling the same product - but at GBP365.00 - forget it!

I'd love to do a 'retro' project using the MC6809/MC68A09/MC68B09, but it just doesn't make sense...

Though one could say the same of using thermionic devices (tubes/valves)...I plead guilty as charged your Honour...

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: Negating LF problems of NBTV.

Postby Steve Anderson » Fri Apr 03, 2020 3:39 pm

Now the circuit may look a little complex, but here's the initial layout for the logic/processor board. There'll be a few other items to be added but it's not a huge undertaking.

The board with the filters & gain/loss stages will be larger then this...

This board incorporates the micro, the dual A-D and the dual D-A

Steve A.

Postscript..."The board with the filters & gain/loss stages will be larger then this..." Actually maybe not, though it may be 'busy'. Changing to just two TL084's may be the answer...
Attachments
Initial layout 1.jpg
Initial layout 1.jpg (62.9 KiB) Viewed 11480 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: Negating LF problems of NBTV.

Postby Steve Anderson » Thu Apr 09, 2020 3:11 pm

Work on this and the SSTV-625 up converter has come to a halt. All suppliers of components are closed here due to COVID-19, even those with on-line ordering and postal delivery, and there's not many of them.

So all I can do is make a start on the code. But with hardware chunks being missing it cannot be tested or debugged.

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: Negating LF problems of NBTV.

Postby Steve Anderson » Tue Apr 14, 2020 2:50 pm

Just a short note - I may be changing to a newer processor, e.g. PIC18F26Q10 - not definite as yet. Half the price and with a lot of extra in-built peripherals.

I haven't checked if the pin-out is the same as the '26K22, most often they are.

Steve A.

Basically the pin-out is the same though several peripheral I/Os can be re-mapped to different pins - I doubt I'll be needing to do that.
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: Negating LF problems of NBTV.

Postby Steve Anderson » Sat Apr 18, 2020 5:34 pm

I've decided to 'stick with the devil you know' and use a PIC18F26K22 micro during development, then shoe-horn the code into a smaller memory device later - if possible...though I am fairly optimistic about that.

I still need to work on the filters though.

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: Negating LF problems of NBTV.

Postby Steve Anderson » Sat Apr 25, 2020 12:30 pm

I've massaged the filters such that the video filters are Bessel and the audio filters are Butterworth. The Bessel filters give less over/undershoot on video signals (syncs for example), but because of their gentle attenuation characteristics may introduce some aliasing - wait and see if that's an issue.

The audio filters are Butterworth which do have some overshoot/undershoot and phase inaccuracies, but unless you have 'golden ears' and are using high end audio gear you'r unlikely to notice it. Keep in mind what this is being used for. The advantage is steeper cut-off with less aliasing which might/can be heard.

The input stages also now have a gain of five such that a standard 1V p-p video signal and a 1V p-p (approx. 300mV RMS) audio signal is more compatible with signal sources.

This is not 'set in stone' and there are very likely changes will be required as a result of experience.

Steve A.
Attachments
PC NBTV Interface 5A.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: Negating LF problems of NBTV.

Postby Steve Anderson » Thu May 21, 2020 2:09 pm

Very little done on this or anything else in almost a month, a major upheaval/re-organisation of my workshop/office has been underway, almost complete. It's amazing the amount of junk one accumulates over just a few years!

But I hope to be 'back on the case' once the weekend is over.

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: Negating LF problems of NBTV.

Postby Harry Dalek » Thu May 21, 2020 2:48 pm

Steve Anderson wrote:Very little done on this or anything else in almost a month, a major upheaval/re-organisation of my workshop/office has been underway, almost complete. It's amazing the amount of junk one accumulates over just a few years!

But I hope to be 'back on the case' once the weekend is over.

Steve A.


Good Luck with the clean up ~!
Know what that's like ,i have to do it every now and then as i loose track of what i have done with some parts more than anything i need to do an audit again .
As they say one mans junk is another's Treasure ! :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: Negating LF problems of NBTV.

Postby Steve Anderson » Sun Aug 02, 2020 6:23 pm

Well, it's now August and the sort-out of the office/workshop is virtually complete, it's taken far longer than I expected! However some 'real work' also interrupted the process. Now I hope to resume this project...

The first thing I need to do is make sure that the PC can accept the incoming data rate and also output it at 921,600 baud. If that's the case then it's all 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: Negating LF problems of NBTV.

Postby acl » Fri Aug 07, 2020 8:31 am

Hi Steve,

Look forward to seeing your progress on the SSTV/NBTV decoder.

Regards Chris Lewis
acl
Anyone have a spare straightjacket?
 
Posts: 440
Joined: Thu Mar 08, 2007 7:38 am

Re: Negating LF problems of NBTV.

Postby Steve Anderson » Fri Aug 07, 2020 2:18 pm

Hi Chris, I'm not sure which project you're referring to as I have four currently in development...so here's a quick status report...

A) The AD9850 DDS oscillator & PC interface/control...work in progress.
B) The DC-coupled PC-based NBTV + audio record/playback interface...work in progress.
C) The MkIII version of the SSTV-625 up-converter...on hold until the above complete.
D) A NBTV-625 up-converter...design almost complete, will follow on from 'C' above. Not mentioned so far on this forum. (Watch this space!).
E) For the future, combine 'C' and 'D' above into one unit, though it's not currently on my 'to do' list, just an idea.

There are one or two other ideas rattling around in my empty skull, but they're quite a while off yet...

Steve A.

P.S. If anyone wants to suggest a future project that I may be able to put together, ideas welcome...e.g. 'B' above, remove the requirement of the PC, record/playback on a SD card as part of the unit. If you've got some older SD cards of 2GB or less hold on to them for this, I really don't want to mess with the e-FAT format, FAT16/FAT32 is OK, though the cards will not use any of the conventional formats...they'll be used simply as a glorified EEPROM - which is really all they are.

Using the same format as the PC-based version of 'B' above, a 1GB SD card will record over three hours of NBTV audio and video.

...however there is a caveat to this... using the proprietary SD card interface requires an expensive (for us lot) licence, the SPI interface is licence-free but of course slower.
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 Electronic NBTV

Who is online

Users browsing this forum: Google [Bot] and 9 guests