The PMT in control of the Devil

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Re: The PMT in control of the Devil

Postby Steve Anderson » Thu Jul 20, 2017 4:36 pm

With 12-bits the ramp should look much smoother than the one shown. In fact it should look identical to an analogue ramp generator except when you zoom right in. Even with only 10-bits it should look like a smooth ramp. At 12-bits there should be over 4000 little steps in the waveform, 10-bits, over 1000.

There is also the problem of resistor tolerance. Over 6-bits the resistors should be 1% (64 steps), when you get to 8-bits you're looking at 0.1% resistors...and so on. This is why most D-As are a single chip with some form of internal R2R or current-switching arrangement where the matching can be trimmed by a laser during production.

Also the outputs when using 10/12-bits need to be equally matched when using something like a 4040, they will be close but not close enough for 10 or 12-bit accuracy.

With 12-bits the clock needs to be 2^12 times the ramp frequency, for 25Hz that's 25 x 4096 = 102.4kHz, a quarter of that for 10-bits.

Really, once you get to 8-bits or more, using discrete resistors becomes impractical.

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: The PMT in control of the Devil

Postby Andrew Davie » Thu Jul 20, 2017 4:40 pm

I think I did OK with my resistor DAC so it's possible...

Image
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: The PMT in control of the Devil

Postby Steve Anderson » Thu Jul 20, 2017 4:45 pm

Yes, depending on the application that may be OK. But note the kink in the centre, that's more than likely the the MSb output resistor being slightly off. I recall you fixed it by replacing the resistor or something. With 6-bits you can 'get away' with using looser-tolerance components if they are 'batched-matched'', all from the same batch made at the same time and new.

Beyond that you are seriously just wasting bits and potentially processing power or logic complexity.

Wikipedia...

https://en.wikipedia.org/wiki/Resistor_ladder

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: The PMT in control of the Devil

Postby Andrew Davie » Thu Jul 20, 2017 4:49 pm

Steve Anderson wrote:Yes, depending on the application that may be OK. But note the kink in the centre, that's more than likely the the MSb output resistor being slightly off.


Yes, d5 (=32) would be slightly out here. I did this with hand-tailored "resistors" which were actually two in parallel and another one in series so that I could try to get the resistance "really accurate". Then I spent a lot of time measuring actual resistance and choosing the appropriate individual resistors that got me closest to my "goal". I realise the differences between accuracy and precision, the limitations of the multimeter, and also the issues you are pointing out. After all this DAC is only 6 bits. But still, it's not bad :P
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: The PMT in control of the Devil

Postby Steve Anderson » Thu Jul 20, 2017 4:57 pm

If you have a circuit simulator you can see the effect of resistor tolerance on the output voltage. Say using a R2R ladder comprised of 10k and 20k resistors, all exact except one. The degree of influence does depend on where in the ladder the 'off'' resistor is, but it is an error. And in practice it's not just one resistor.

For 12-bits you're hoping for an accuracy of 1/4096, about 0.025%.

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: The PMT in control of the Devil

Postby Harry Dalek » Thu Jul 20, 2017 5:59 pm

Steve Anderson wrote:If you have a circuit simulator you can see the effect of resistor tolerance on the output voltage. Say using a R2R ladder comprised of 10k and 20k resistors, all exact except one. The degree of influence does depend on where in the ladder the 'off'' resistor is, but it is an error. And in practice it's not just one resistor.

For 12-bits you're hoping for an accuracy of 1/4096, about 0.025%.

Steve A.


I missed the last few posts till now working on the circuit today just about everything discussed had dawned on me ,not being digital savvy...i have learned the hard way ...The clock needs to be faster again the resistors need to be 1% even there checked ...so i gave up today
You mentioned most Digital Analog chips now have some form of internal resistors ...that would save circuit board space and oh the wiring ! what would you say is worth buying Steve here i am going order some ic's and stock up a bit so that would come in handy if you could point me in that direction on that ic ?
I would love to do away with the resistor ladder .
BY the way seeing the numbers needed for the clock with this idea it sounds like i would have to go crystal now do you think ?
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: The PMT in control of the Devil

Postby Steve Anderson » Thu Jul 20, 2017 6:50 pm

I feel what needs to be pinned down is the modes this multi-standard monitor needs to work at. How many lines maximum at the maximum frame/field rate.

If you want to go (say) for 256 lines at 50Hz non-interlaced that's a line rate of 12.8kHz. Now if the picture was square, and ignoring syncs, blanking and the like the pixel rate is about 3.3MHz. This would also be the line timebase clock rate. In practice it would have to be 20-30% higher, even more so if using a 4:3 aspect ratio.

You can see the numbers are getting bigger all the time.

Even 405 at a field rate of 50Hz requires a pixel clock of around 6MHz to provide the required 3MHz video bandwidth.

You're starting to get into some serious design and construction methods.

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: The PMT in control of the Devil

Postby Klaas Robers » Thu Jul 20, 2017 8:10 pm

You can try the IC DAC8. It is an integrated ladder network for 8 bits. and the irregularity at 127 to 128 is almost invisible on an oscilloscope.

For the horizontal sawtooth you think that you need many bits to make the wave form smooth, but with a small capacitor you can smooth the fine staircase to a sawtooth. Then the jump back is slightly slower, but all TV standards have a kind of sync pulse at that moment, so you wouldn't see it. It is blanked!

For the vertical sawtooth this is not needed, a staircase is even better than a sawtooth. Also the jump back is done instantaneous, much faster than with whatever sawtooth. And don't be afraid for higher frequencies. Your limit is the max. clock frequency of the 4040.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: The PMT in control of the Devil

Postby Harry Dalek » Thu Jul 20, 2017 8:44 pm

Steve Anderson wrote:I feel what needs to be pinned down is the modes this multi-standard monitor needs to work at. How many lines maximum at the maximum frame/field rate.

Hi Steve
Well i started off seeing what was possible to view on the devil monitor .
The flying spot idea here was really to get me started and display this camera idea on this monitor and learn ... the flying spot camera i am hope can out do this monitor for a future monitor that would be above 100 lines i was thinking this is wise or i would have to build it all over again so it is better to at least do 32 line 60 or 64 so i can test it it works and then test the higher lines rates on new monitor .
Line rate above 100 or 200 lines might be a limit for now as with the 4060 with the doubling of the frequencies i am not to bothered if it has not been used before ... 120 125 so on its more around there whats easier to make 25 hz frame rate seems to be used around those line rates but if 12.5 is only possible thats ok too .
i always had in mind to build another monitor for the higher line rates when i learnt from you these little CRT tubes have there limits raster dot size .

If you want to go (say) for 256 lines at 50Hz non-interlaced that's a line rate of 12.8kHz. Now if the picture was square, and ignoring syncs, blanking and the like the pixel rate is about 3.3MHz. This would also be the line timebase clock rate. In practice it would have to be 20-30% higher, even more so if using a 4:3 aspect ratio.

You can see the numbers are getting bigger all the time.

Even 405 at a field rate of 50Hz requires a pixel clock of around 6MHz to provide the required 3MHz video bandwidth.

You're starting to get into some serious design and construction methods.

Steve A.[/quote]

I can see it gets very large using the digital idea needs a high crystal clock rate for the bandwidth correct ? old color tv crystals 4.43 and such 20mhz i have might be useful here /
I started off here wanting a clock with line a frame rates to sync 2 sawtooth oscillators both worked independently not together ...to go now all digital is fine but i have to ask advice here as i am hopeless at this side of it . :roll:
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: The PMT in control of the Devil

Postby Harry Dalek » Thu Jul 20, 2017 8:53 pm

Klaas Robers wrote:You can try the IC DAC8. It is an integrated ladder network for 8 bits. and the irregularity at 127 to 128 is almost invisible on an oscilloscope.

For the horizontal sawtooth you think that you need many bits to make the wave form smooth, but with a small capacitor you can smooth the fine staircase to a sawtooth. Then the jump back is slightly slower, but all TV standards have a kind of sync pulse at that moment, so you wouldn't see it. It is blanked!

For the vertical sawtooth this is not needed, a staircase is even better than a sawtooth. Also the jump back is done instantaneous, much faster than with whatever sawtooth. And don't be afraid for higher frequencies. Your limit is the max. clock frequency of the 4040.


OH thats an interesting solution Klass i will google that ic see if i can work it out ...that would be as Steve mentioned a good idea i didn't like the resistor ladder only the matching of resistance does need better resistors than i had handy .
OK i will look into that...Edit looking at a few examples 555 say a 4040 or 7493 some thing like that then the DAC ic so in my case 4060 4040 and a DAC ic yes ?
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: The PMT in control of the Devil

Postby Steve Anderson » Thu Jul 20, 2017 9:28 pm

I used a DAC08 in the NBTV/SSTV-625 up-converter for the video output. It has enough bandwidth. It's downside is it requires + and - supplies, otherwise it's very easy to use.

Attached is a partial of the circuit of the MkI version (the MkIII is virtually identical). Ignore the AND gates, a few minor changes would be required here and there for your application. The supplies don't have to be +/-12V, +/-5V would be OK too. In fact they don't have to be equal, say +12V and -5V.

Steve A.

That damned default font....but it should give the idea.
Attachments
SSTV-625 10.gif
SSTV-625 10.gif (31.16 KiB) Viewed 10603 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: The PMT in control of the Devil

Postby Harry Dalek » Thu Jul 20, 2017 9:57 pm

Steve Anderson wrote:I used a DAC08 in the NBTV/SSTV-625 up-converter for the video output. It has enough bandwidth. It's downside is it requires + and - supplies, otherwise it's very easy to use.

Attached is a partial of the circuit of the MkI version (the MkIII is virtually identical). Ignore the AND gates, a few minor changes would be required here and there for your application. The supplies don't have to be +/-12V, +/-5V would be OK too. In fact they don't have to be equal, say +12V and -5V.

Steve A.

That damned default font....but it should give the idea.


The positive and negative supplies are fine that does not bother me Steve ..i was googling them whats coming up is DAC08c or DAC08cp or DAC08e or DAC08en seems to be a lot of different version of them ,i would need 2 for the H and V sawtooth ...16 pin /
i will review those projects you mentioned
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: The PMT in control of the Devil

Postby Harry Dalek » Thu Jul 20, 2017 10:51 pm

I found this on ebay ...most i have searched for are at the price of these postage but i would have a few years supply here..... this lot any good ?
http://www.ebay.com.au/itm/TESLA-MDAC08 ... SwSypY~0Dd
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: The PMT in control of the Devil

Postby Steve Anderson » Fri Jul 21, 2017 11:21 am

Not heard of Tesla before, not at least relating to chips. But at the price it's worth taking a chance. I'm fairly confident they'll be OK. The type numbering differences are almost superficial, most of these devices are the same, just a slightly different specification.

I did a search for the company and came up with two, one in Europe the other in Bangladesh. Nether appear to have a web-site.

If you do go ahead perhaps you could send me a couple and I'll drop them into the NBTV/SSTV-625 up-converter and confirm they're OK.

I agree, the differences between the type numbers can be confusing, sending you off for hours trying to find out if there are any significant differences. But without a Tesla datasheet we'll just have to try them out.

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: The PMT in control of the Devil

Postby Harry Dalek » Fri Jul 21, 2017 1:11 pm

Not heard of Tesla before, not at least relating to chips. But at the price it's worth taking a chance. I'm fairly confident they'll be OK. The type numbering differences are almost superficial, most of these devices are the same, just a slightly different specification.


Looking around the price is good with the postage its about the same as buying the cheapest 2 chips with free postage but you get over 20 of them.

I did a search for the company and came up with two, one in Europe the other in Bangladesh. Nether appear to have a web-site.


These are via the USA but looking at the data sheets they are Russian by the looks of it .

If you do go ahead perhaps you could send me a couple and I'll drop them into the NBTV/SSTV-625 up-converter and confirm they're OK.


Will do no problems it will be interesting to find out i will send some spares.i have the data sheets but will still post the spares off to you .

I agree, the differences between the type numbers can be confusing, sending you off for hours trying to find out if there are any significant differences. But without a Tesla datasheet we'll just have to try them out.

Steve A.


I am as i said not familiar with the chips at all so its a learning curve ...
I will just send off this post and if the data looks ok i will order them .the others i have seen on the selling sites pretty much 10 to 20 dollars a chip.
Attachments
MDAC08EP-Tesla.pdf
(381.64 KiB) Downloaded 427 times
mdac08.pdf
(652.52 KiB) Downloaded 438 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

PreviousNext

Return to The Tasmanian Devil VCR139....A for Andrew

Who is online

Users browsing this forum: No registered users and 1 guest

cron