Mini version of Scophony - its do-able

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Steve Anderson, Dave Moll, Andrew Davie

Re: Mini version of Scophony - its do-able

Postby Gregory » Tue Apr 26, 2016 1:03 am

Thanks Gary for all the info, very useful. One question - with the xtal oscillator how would you get the frame or line beginning to coincide with the actual spot position or in other words the black frame band to be at the bottom of the screen and the line black band on the left - I assume this may have to be done manually?
Gregory
Laboratory Assistant
 
Posts: 37
Joined: Wed Feb 10, 2016 5:06 pm
Location: Nicosia, CYPRUS

Re: Mini version of Scophony - its do-able

Postby gary » Tue Apr 26, 2016 12:55 pm

Gregory wrote:Thanks Gary for all the info, very useful. One question - with the xtal oscillator how would you get the frame or line beginning to coincide with the actual spot position or in other words the black frame band to be at the bottom of the screen and the line black band on the left - I assume this may have to be done manually?


Because of your original circuit I had imagined this would be "free running" this system. To lock to line and frame sync you need to extract the sync out of the video - since you are doing that you may as well use those syncs to trigger your clock and so a 555 would be fine.

OTOH if it IS a free running system then you could manually lock a) by making the clock frequency variable, or b) allow the odd train pulse to be "lost" - that may be as simple as just switching the clock train off momentarily - there is room for experiment there.
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

Re: Mini version of Scophony - its do-able

Postby Steve Anderson » Tue Apr 26, 2016 8:23 pm

This thread has almost got me to do a "Care and Feeding of Stepper Motors" article or thread...as applied to NBTV of course.

Gary points out that generally the average stepper motor needs to be fed with an Sine/Cosine waveform, or even a Cosine/Sine waveform depending on which way you wish the thing to rotate. Usually this is done in a simple binary fashion with just a few gates, feed in the pulses, the motor will (should) respond. Because of the bang-bang digital waveform is why as your printer advances the paper it is so noisy...at least mine is...

This is the 'cogging' Gary mentions. The motor moves from one position to the next quite quickly then sits idle until the next pulse(s) that edges it forward...or backwards. (Sine/Cosine Again).

Now, there are a large number of issues here to deal with. First is the rated Voltage and winding Inductance. Torque is proportional (as an approximation) to winding current. If the pulse is quite short thus not allowing the current time to build up the motors torque output will be quite low, thus probably getting nowhere. i.e. there's a limit to speed and mechanical load. The product of the two basically defining the power output of the motor.

I have to say that driving a stepper motor of the common variety at 8kHz is likely to be a disappointment. At least marginal. Please prove me wrong.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 3856
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: Mini version of Scophony - its do-able

Postby Gregory » Tue Apr 26, 2016 8:48 pm

Steve Anderson wrote:I have to say that driving a stepper motor of the common variety at 8kHz is likely to be a disappointment. At least marginal. Please prove me wrong.

Hi Steve - The 8 khz refers to the steps the driver is receiving, The stepper driver has a micro-steps feature which basically divides the full step into smaller steps (ie this produces the sine wave drive required using PWM) I have set the dil switches to divide by 16 micro-steps, which is the smoothest selection available, so one full step is divided into 16 smaller steps, and so the actual stepper motor receives 8khz/16 = 500 full steps per second.

The resolution of the motor I am using is 200 full steps per rev so I am getting a speed of 2.5 revs per second and with my 8 sided frame scan mirror this gives me the 20 frames per second I need. I have found the motor runs smoothly on this speed, but to start it I still need to start form a lower frequency and ramp up.
Gregory
Laboratory Assistant
 
Posts: 37
Joined: Wed Feb 10, 2016 5:06 pm
Location: Nicosia, CYPRUS

Re: Mini version of Scophony - its do-able

Postby Steve Anderson » Tue Apr 26, 2016 8:55 pm

Ah! I was going to mention micro-stepping, but thought best not to.

All I can say is if it's meeting your demands. well done!

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 3856
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: Mini version of Scophony - its do-able

Postby gary » Tue Apr 26, 2016 8:58 pm

Gregory wrote:
Steve Anderson wrote:I have to say that driving a stepper motor of the common variety at 8kHz is likely to be a disappointment. At least marginal. Please prove me wrong.

Hi Steve - The 8 khz refers to the steps the driver is receiving, The stepper driver has a micro-steps feature which basically divides the full step into smaller steps (ie this produces the sine wave drive required using PWM) I have set the dil switches to divide by 16 micro-steps, which is the smoothest selection available, so one full step is divided into 16 smaller steps, and so the actual stepper motor receives 8khz/16 = 500 full steps per second.

The resolution of the motor I am using is 200 full steps per rev so I am getting a speed of 2.5 revs per second and with my 8 sided frame scan mirror this gives me the 20 frames per second I need. I have found the motor runs smoothly on this speed, but to start it I still need to start form a lower frequency and ramp up.


Oh and with the microstepping you probably won't need to drive it with sine waves (to eliminate cogging) as the microsteps ultimately approximate a sine wave (16 might be too few but with the inertial smoothing also going on I think it might be ok).

I got up to 750 RPM with a simple L/R driver (and high voltage) - I have had a micro stepper driver for quite some time but I haven't had the time to try it out :-(

And yes ramping will always be necessary to get through those pesky resonances it has to pass through.
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

Re: Mini version of Scophony - its do-able

Postby Harry Dalek » Wed Apr 27, 2016 12:08 am

Hi Gregory

From what i found you can only start up slow with the type of circuit idea your using ,the polygon motor circuit driver board they pretty fast ramp up the motor speed very quickly .

With what you are using on manual control to get to a set speed or speeds as you want it multi system ,it is a step a above a DC motor so your drift will be there but a lot slower than with one of those,you will also want to know the speed and control your line motor mirror using an opto switch with one of these type mirrors you with get a very good steady line pulse you can use in a feed back to control the motor ...Having made something like this i found the line speed is critical but the frame speed as in seeing a synced image is not ...very easy in manual control ! Sort of like a zoom in out effect as the frame motor speed is adjusted the image effect is picture size quality and i suppose movement is effected but you can see all this so its easy to adjust .
So i would be worrying more about your line speed control than any thing ...do it all in manual control have a play you will see it works but there's drifting a bit then you can work out how to get that line mirror to lock to the video line sync pulse if that's what you what you want in time .
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: 4105
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Re: Mini version of Scophony - its do-able

Postby Gregory » Sat Apr 30, 2016 7:44 am

Made some good progress this week - got the green laser modules at last (10 mw & 30 mw). I have now installed the 10 mw one. Before removing the module from the laser diode I measured the current which was 240 ma. The series resistor has now been adjusted to give 220 ma on the laser diode, which now gives a much brighter raster than the original red laser diode. The 30 mw laser diode I will leave for the very end, once all the experimenting is done, although the 10 mw one is reasonably bright and it may not be necessary to go to 30 mw

30 lines.jpg
30 lines.jpg (62.87 KiB) Viewed 4808 times

60 lines.jpg
60 lines.jpg (79.58 KiB) Viewed 4808 times

Also the 555 clock oscillator with the start up ramp has now been wired and tested on the breadboard. The clock frequency is 8 kHz with a ramp up time of around 2 - 3 seconds

555 clock with ramp.jpg
555 clock with ramp.jpg (65.8 KiB) Viewed 4808 times

The next phase is to build on the breadboard a new driver for the line scan using the TDA5140a IC as in the circuit posted earlier. Although the current lines scan driver works well, it has a big disadvantage that as it uses surface mount components and unmarked IC, there is no way of connecting it to be able to sync automatically. Also this driver often will not start without spinning the mirror to start it off. Hopefully if the TDA5140a IC works well it may be possible to find a way to use the line sync with this.
Gregory
Laboratory Assistant
 
Posts: 37
Joined: Wed Feb 10, 2016 5:06 pm
Location: Nicosia, CYPRUS

Re: Mini version of Scophony - its do-able

Postby Harry Dalek » Sun May 01, 2016 3:57 pm

Hi Gregory
Yes the Green laser is pretty good i always thought those were the brightest ,getting very close now
You have a horizontal display not that it matters just turn its side for Baird scanning .
The line scanning is the hardest but not impossible even in manual control .
Really good work very compact .
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: 4105
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Re: Mini version of Scophony - its do-able

Postby Gregory » Thu May 05, 2016 10:01 pm

Harry Dalek wrote:Hi Gregory
You have a horizontal display not that it matters just turn its side for Baird scanning .


Harry - Good idea for the baird system, eventually when I make the case, must make it to stand on the side too.

The line scan driver with the TDA5140a IC has been wired and tested on the breadboard and the ready made brushless motor driver has been removed. It now starts every time, but I did not realize that the TDA5140a IC requires a high power voltage control for the speed control input as this is connected direct to the output drivers. The speed can also be controlled by just varying the supply voltage to the driver, as it would be with a DC motor.

Also one question if anybody can help please, for 30 lines 12.5 frames/sec I can use a CD player or computer for the signal source, but what can one use for 60 lines at 20 frames/sec?

From what I understand 10 khz bandwidth is sufficient for 30 lines/12.5 f/sec but for the 60 lines/20 f/sec, I would need around 65 khz bandwidth.
Ideally it would be nice to have a stand alone device like a CD player or MP3 player, but laptop would be OK, if I can get that from the sound card output.

If anyone has suggestions would be most welcome.
Gregory
Laboratory Assistant
 
Posts: 37
Joined: Wed Feb 10, 2016 5:06 pm
Location: Nicosia, CYPRUS

Re: Mini version of Scophony - its do-able

Postby gary » Fri May 06, 2016 12:02 pm

60 lines at 20 fps and 4:3 aspect ratio (you don't seem to have explicitly defined the aspect ratio but you have mentioned a 6cm x 8cm screen so I assume 4:3)

SR = 20 x 60 x 60 x 4 / 3 = 96000 (BW of 48kHz)

Theoretically modern sound cards can handle 192kHz sampling (96kHz BW).

I have never been able to get around to testing that the output filters of such a sound card pass that BW though, as, of course, the human ear can't hear anything above around 20kHz, so it remains to be seen if the manufacturers do or don't take advantage of that to relax the filter response requirements.

I can provide you with the necessary software to produce the signal if you wish to experiment.
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

Re: Mini version of Scophony - its do-able

Postby Gregory » Sat May 07, 2016 5:26 am

gary wrote:60 lines at 20 fps and 4:3 aspect ratio (you don't seem to have explicitly defined the aspect ratio but you have mentioned a 6cm x 8cm screen so I assume 4:3)

SR = 20 x 60 x 60 x 4 / 3 = 96000 (BW of 48kHz)

Theoretically modern sound cards can handle 192kHz sampling (96kHz BW).

I have never been able to get around to testing that the output filters of such a sound card pass that BW though, as, of course, the human ear can't hear anything above around 20kHz, so it remains to be seen if the manufacturers do or don't take advantage of that to relax the filter response requirements.

I can provide you with the necessary software to produce the signal if you wish to experiment.

Hi Gary,

Thanks for the info and I have been looking further on the web to see if there are any cheap options to do this. Unfortunately most sound cards are filtered with the exception of a few which are expensive option ($250 -$500}, so maybe a used sound card could be a solution.

Also I have seen DVD-A players can output up to 96kz, but again they don't come cheap, although I have seen some used ones for less than $100. I noticed however that these are not compatible with wav files, but accept WMA files - do you think this would be suitable?

Gary in your experience, what solution have other club members used to overcome the problem of bandwidth, and is there a cheap option available?
Gregory
Laboratory Assistant
 
Posts: 37
Joined: Wed Feb 10, 2016 5:06 pm
Location: Nicosia, CYPRUS

Re: Mini version of Scophony - its do-able

Postby gary » Sat May 07, 2016 12:09 pm

WMA: In general any form of lossy *audio* compression will introduce unwanted artefacts into NBTV video and is to be avoided if at all possible.

Once you are over the normal audio bandwidth you are generally in uncharted territory where you are unlikely to find "off the shelf" solutions.

One notable exception is the Aurora Standards Converter (http://www.tech-retro.com/Aurora_Design ... erter.html) which has been used successfully by a number of members. However it is not a "cheap" solution.

Other solutions I know of include:
DIY interface for the S/PDIF (digital interface) to consumer digital systems (CDs DVDs etc) and soundcards.
PIC/Microprocessor/microcomputer based standard converters.
DIY ADC/DAC systems.

There could be more options, if I think of any I will let you know.
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

Re: Mini version of Scophony - its do-able

Postby Gregory » Sun May 08, 2016 7:03 pm

Thanks Gary, I will also keep looking, maybe I will be lucky.

I have at last got the first pictures from the televisor, and have done all tests on the 32 lines for the moment. There is no sync on either the frame or the line and although the image is not bad for such a crude laser modulator, there is slight slow rolling of the image both horizontally and vertically, so sync will be a must. The image is bright with a good contrast, so the green laser diodes should be sufficient in power.

Also the laser modulator, which is basically a power FET with resistor current limiting needs to be re-designed, probably for a PWM type, but I will leave this until last

Below is the schematic, so far, and I will be updating this in future posts, until the project is complete.

schematic televisor 1.0.jpg
(57.87 KiB) Not downloaded yet

The next phase will now be to do the frame and line sync. I will be using a paper printed disk with 8 black bands for the frame mirrors and 15 black bands for the line mirrors, so the frequency signals form the photodetector will correspond exactly with the frame and line sync signal. 4046 ICs as I have seen on the posts here seems to be good solution to control the syncs once the two sync signal are separated.

As soon as I successfully do the sync, should be able to upload some pics, of the screen images for 32 lines.
Gregory
Laboratory Assistant
 
Posts: 37
Joined: Wed Feb 10, 2016 5:06 pm
Location: Nicosia, CYPRUS

Re: Mini version of Scophony - its do-able

Postby Gregory » Fri May 13, 2016 4:15 am

Manage to do the line scan encoder disc and fitted a CNY70 reflective sensor/led as in the photo below. I coupled this onto a single transistor common emitter amplifier and go a decent square wave out of this. Also the sync separator has been completed and works as expected.

line scan sensor.jpg
line scan sensor.jpg (80.74 KiB) Viewed 4621 times

There is however a HUGE problem when trying to sync the line scan motor. After a lot of playing about I managed to get a sync - well sort of - but this is very shaky and can drift after a few minutes beyond the capture range of the 4046.

After a lot of checking, I believe the hard disc motor will not be suitable for the line scan motor, in its present form. Although the HDD motor is capable of very high speed and is excellently engineered, it has very low torque, meaning that any adjustment to its speed is slow and making the feedback loop time large. Perhaps with a phonic wheel it could work, but I don't know for sure. In addition the BLDC driver takes time to reach its final frequency, even if the voltage control to it is rapidly change, as it constantly monitors the rotor position before increasing its frequency.

The BLDC driver, although reasonably stable, can cause slight changes to its RPM as it is monitoring the rotor position constantly, and if this drifts, it can cause unexpected very small, but sudden changes to the speed.

I have looked into the possibility of using a stepper motor for the line scan. The 1.8 degree stepper motors unfortunately can only go to around 1000 RPM or thereabouts, but for 60 lines at even 12.5 frames/sec one needs 3000 RPM with the 15 sided mirror and 4800 RPM for 60 lines 20 frames/sec.

A second option is 7.5 degree stepper motors, and here for the same pulse rate the RPM is about 4 times as much. I have scavenged a Matsumi M49SP-2K stepper motor from an old laser copier and tried it out to see the maximum RPM unloaded. The spec shows a maximum of 2100 full steps per sec which corresponds to 2400 RPM.

After testing with a 25 volt supply with the stepper driver set to 0.8 Amps the best performance was with 1/16 micro steps, the motor reaching 5500 RPM before cutting off - if I got things right, which is quite a bit more than the spec

So major changes, I will need to re-mount the line scan mirror with the stepper motor, luckily it should fit into the existing base, provide I cut a hole in the base for the motor. Hopefully after this I should have a good sync, but will have to wait for another stepper motor driver to come.
Gregory
Laboratory Assistant
 
Posts: 37
Joined: Wed Feb 10, 2016 5:06 pm
Location: Nicosia, CYPRUS

PreviousNext

Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 5 guests

cron