NBTV without any moving parts. THE NIPTRIX

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Steve Anderson, Dave Moll, Andrew Davie

Postby AncientBrit » Wed Oct 10, 2012 5:36 pm

Gary and others have mentioned using a cap to extend the on illumination time of each LED.

Are we talking here of a simple transistor constant current source feeding each LED with the cap on the base of the transistor?

This would give isolation, allow a smaller cap and reduce drive requirements but at the expense of yet more components.

Graham
AncientBrit
Green padded cells are quite homely.
 
Posts: 858
Joined: Mon Mar 26, 2007 10:15 pm
Location: Billericay, UK

Postby Viewmaster » Wed Oct 10, 2012 6:31 pm

M3DVQ wrote:If you only want to use it with crystal controlled sources (e.g. a computer) then you can run the counters from a crystal clock divided down and re-synchronise it to the line pulses to stop it drifting enough to cause a picture tear. This is what I did with my oscilloscope driver circuit and it works very well.


I am still mulling over all the options if I build this display.

At the pixel clock rate of 19.2kHz a suitable crystal would be 1.2288 mHz
divided down 6 times.
I have Googled but cannot find a supplyer who would cut a single xtal for me if I took that route.
Could you tell me where you obtained your crystal ?
Thanks.
“One small step for a man,"......because he has Arthritis.
Albert.
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1292
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby Steve Anderson » Wed Oct 10, 2012 6:34 pm

Viewmaster wrote:
M3DVQ wrote:At the pixel clock rate of 19.2kHz a suitable crystal would be 1.2288 mHz. Could you tell me where you obtained your crystal ?


You could use a 12.288MHz Xtal (which is a standard value...and cheap) and divide that by 10, half of a HC390...

Steve A,
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 4252
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Postby Viewmaster » Wed Oct 10, 2012 6:57 pm

Steve Anderson wrote:
Viewmaster wrote:
M3DVQ wrote:At the pixel clock rate of 19.2kHz a suitable crystal would be 1.2288 mHz. Could you tell me where you obtained your crystal ?


You could use a 12.288MHz Xtal (which is a standard value...and cheap) and divide that by 10, half of a HC390...

Steve A,


Gee, many thanks for that, Steve. I didn't realise that our NBTV pixel rate
could be obtained from a standard xtal.
“One small step for a man,"......because he has Arthritis.
Albert.
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1292
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby Steve Anderson » Wed Oct 10, 2012 7:14 pm

Viewmaster wrote:Gee, many thanks for that, Steve. I didn't realise that our NBTV pixel rate could be obtained from a standard xtal.


19.2kHz is a standard baud rate for modems/RS232, as is it's divisions, 9.6kbaub, 4.8, 2.4, 1.2...and it's multiples, 38.4, 57.6 (x3)...the fastest I believe analogue modems went.

Farnell list over 40 flavours of 12.288Mhz Xtals, about 10 of which are wire-ended (not SMD) and around 50p each.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 4252
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Postby M3DVQ » Thu Oct 11, 2012 2:16 am

For my scope driver I used a 3.2768MHz crystal divided down by 64 to generate a 51.2kHz clock. This clocks a 12 bit synchronous counter of which 5 bits give the X scan and 7 give the Y scan. This means 32x128 but the voltage swings were scaled appropriately to give the correct aspect picture.

If you divide the master clock by 128 instead (to 25.1kHz) and have your counter 11 bits then you end up with 32 lines each 64 pixels, that would make a total of 2048 LEDs. I know this is then 2:1 rather than the club 3:2 but I am always drawn towards powers of two as they're so much easier to work with in digital electronics :wink:
M3DVQ
Just nod and pretend you understand me
 
Posts: 331
Joined: Sat Feb 10, 2007 10:52 am
Location: Lincolnshire

Postby Viewmaster » Thu Oct 11, 2012 6:30 pm

M3DVQ wrote:For my scope driver I used a 3.2768MHz crystal divided down by 64 to generate a 51.2kHz clock. :


Well, thanks for that, but when Steve mentioned that a standartd xtal could be used I started agoogling again and found this.......

A Cmos 4060 has a built in oscillator which can use a xtal. It can also divide down by a factor of
16/32/64/128/256/512/1024/4096/8192/16384, just by a single pin selection. Amazing to me.
Also, more agoogling, I found another standard xtal of 4.9152 mHz. This divided down by 256 gives that 19.2KHz pixel rate. Even more amazing to me. I have never used a xtal, so this may be a simple way for me to have a go.
“One small step for a man,"......because he has Arthritis.
Albert.
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1292
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby Steve Anderson » Fri Oct 12, 2012 12:49 am

I was going to suggest the 4060 but be careful if you're going to use the CD/HEF version at 5V, it's maximum frequency is limited, worst case to 3.5MHz. At 5V use a HC4060, good to 80MHz plus.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 4252
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Postby Viewmaster » Fri Oct 12, 2012 1:13 am

Steve Anderson wrote:I was going to suggest the 4060 but be careful if you're going to use the CD/HEF version at 5V, it's maximum frequency is limited, worst case to 3.5MHz. At 5V use a HC4060, good to 80MHz plus.
Steve A.


Thanks again for guidance. Things are firming up, I think ! Using.......

That 4060 xtal for pixel clocking at 19.2 kHz.

Peter Smith's missing 32 frame pulse (narrowed) to reset all 10x4017 cascades to zero for the column and rows and also to reset the pixel clock.

Grant Dixon's 80 transistor circuit (PNP and NPN) matrix drives to the LEDs. This will allow me to set up the LED's current, balancing brightness against blowing them all up.

I feel like a chicken now , picking here and there at various folks brains, Peter, Grant, you and others here, Steve.
Thanks to all.
BTW, have just finished 200 soldered joints....only 3,300 more to go to get into the asylum. :lol:
“One small step for a man,"......because he has Arthritis.
Albert.
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1292
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby M3DVQ » Fri Oct 12, 2012 3:03 am

Yep, the 4060 was what I used for my oscillator and clock divider too :D
M3DVQ
Just nod and pretend you understand me
 
Posts: 331
Joined: Sat Feb 10, 2007 10:52 am
Location: Lincolnshire

Postby M3DVQ » Fri Oct 12, 2012 3:10 am

The reason I did it 32x128 was that by keeping the number of pixels per line to a power of two you can use one long counter (in my case 12 bits) which is completely synchronous without any extra logic.
To have 48 pixels per line for example you have to have have decoding logic reset the counter when it reaches a particular value and clock the line counter, then you're adding all sorts of propagation delays and the potential for glitches etc.

That's why I decided it was simpler to stick to powers of 2 :)
M3DVQ
Just nod and pretend you understand me
 
Posts: 331
Joined: Sat Feb 10, 2007 10:52 am
Location: Lincolnshire

Postby gary » Fri Oct 12, 2012 7:34 am

Sounds logical to me ;-)
Perfecting an NBTV system is like trying to slam a revolving door...
User avatar
gary
"Fester! Don't do that to 'Thing'"
 
Posts: 2303
Joined: Sat Jan 27, 2007 11:29 am
Location: Bundanoon, Australia

Postby Steve Anderson » Fri Oct 12, 2012 12:39 pm

Viewmaster wrote:BTW, have just finished 200 soldered joints....only 3,300 more to go to get into the asylum. :lol:


Ah! So you've decided to go ahead with this, I wasn't sure if this was for real or 'armchair engineering' as I so often do.

There's nothing wrong with 'mixing and matching' from various sources, it's done in industry all the time.

Gary's suggestion of using a PC power supply is worthy of consideration unless you have a suitable PSU kicking around already...or a spare car battery! Otherwise you're into what could be a bulky and expensive supply to construct.

There are a number of sites showing you how to convert a PC PSU into a bench supply on the 'net, a quick google should turn up a number. It seems quite simple to do. Here's one example...although I wouldn't call it 'laboratory' standard it's good enough for the vast majority of what we do here...

http://www.wikihow.com/Convert-a-Comput ... wer-Supply

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 4252
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Postby Steve Anderson » Fri Oct 12, 2012 1:09 pm

Not wishing to throw a spanner in the works, has any thought been given to sync pulse time? If using 48 pixels during the active video time of a nominal 2.375ms (allowing 5% of 2.5ms for syncs) results in a pixel rate slightly higher than 19.2kHz....around 20.21...kHz...

I'm gonna make a run for it...

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 4252
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Postby M3DVQ » Fri Oct 12, 2012 1:24 pm

If you have any old AT supplies still hanging around as I do, they are much easier to power up as they generally don't have a minimum load to power up - just a big mains switch. Of course if you want to draw a vast current at 12 volts then they may not be beefy enough.
M3DVQ
Just nod and pretend you understand me
 
Posts: 331
Joined: Sat Feb 10, 2007 10:52 am
Location: Lincolnshire

PreviousNext

Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 6 guests