Starting at the beginning with Mirror Screws

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Re: Starting at the beginning with Mirror Screws

Postby Panrock » Fri Aug 28, 2020 10:01 pm

I am now have a little time and am intending to take this forward: convert my Mirror Screw to project a 'real' 120-line colour picture onto a screen. It 'sucks' having to watch a virtual/reflected picture from across the room!

I think... a vertical line of laser light would give linear results shone at the centre/axis. This could be produced by shining a laser through an acrylic rod (cylindrical section).

I also think... a converging vertical fan of light from LEDs and rod (focused further on, on a screen) would do the same, but with possible vignetting at the line edges.

I could be wrong.

Karen's Slat Angle Timing Corrector and RGB Converger would be as useful as ever. A screen would capture just the corrected edition of the picture.

The screw is 180mm tall. So a 4:3 picture would be 240 mm wide. I therefore reckon a correctly proportioned '12-inch' diagonal picture should be available
about 7½ ft from the screw.

I am looking into using R G and B laser diodes, such as THIS. These devices are extremely eye-unsafe. I welcome all warnings. Maybe stretching the beam into a line-ended fan will mitigate? Maybe they could be under-run by the modulator circuit? I value my eyesight and the eyesight of others even more, to the extent that, unless adequate safety protections can be built in, this may put the stopper to the whole project. Are ready-made laser modulators available these days? I'm back to being new to all this, having just got back in the saddle.

I think I may start by getting a ready-made "collimating+line" laser lens and see if this can be made to work with a narrow-angle LED...

Thanks for reading. Your advice is valued.

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: Starting at the beginning with Mirror Screws

Postby Klaas Robers » Sat Aug 29, 2020 4:28 am

Steve, these laser diodes on their own don't give a parallel focussed beam. When you look into the specification you find the "divergence", which has a FWHM (Full Width Half Maximum) of 7 deg in one direction and 22 deg in the other direction. You can see that too in the graph a few pages further.

Laser diodes always give an oval shaped beam, where the points of origin are not identical, but are still very close to each other. In a laser pointer there is a focussing lens that tries to make the beam parallel in both perpendicular directions. Without that lens you have the oval spot.

You should not modulate the laser, but switch it fast on and off. By pulse dutycycle modulation you may get an analogue modulation, without any form of gamma.

Good luck
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Starting at the beginning with Mirror Screws

Postby Steve Anderson » Sat Aug 29, 2020 11:19 am

Good see you back Steve...

Klaas is correct, I did try to modulate a laser diode with an analogue current (not voltage) source, the problem comes from the 'lasing current' [1] below which it simply stops emitting light...around 30% of the nominal current. So you lose all the detail in the darker scenes. With colour/RGB each diode will 'give up' at a different point leading to severe colour shading in the lowlights. PWM or similar is the way to go. Very similar to what you did on a previous LED mirror-screw monitor, in fact that circuit could be adapted with a few simple mods. You may recall my laser-link in the newsletter which used the same concept (no idea of the issue number).

Incorporating Gamma correction would require a modified ramp generator which drives the three comparators. You could use a 'breakpoint corrector similar to the one of mine also published in the newsletter many years ago (again, no idea of issue number). Actually I've just thought of an easier/better way to do this - let me look into it...

Or, slightly more complex, use a micro with a look-up table to do the same job. That way you can adjust the curve without any hardware knifing-and-forking...and potentially do away with those three comparators...I'll think on this further...

Steve A.

Afterthoughts...

There's a problem with going the micro route and processing three signals at 120 lines - the class of micros I'm used to simply cannot process that amount of data at the speeds required, even at a 64MHz clock speed. I can't recall what frame rate you use but even at 12.5Hz it's too much data. So it's the old analogue method...

[1] In the datasheet you linked to, they call the 'lasing current' a 'threshold current'...a better term I think...

For a 120x120 frame (square) there's 14,400 pixels, at 12.5Hz frame rate that's 180,000 pixels per channel per second, i.e a bandwidth of 90kHz per channel, not including any increase to allow for syncs and/or blanking.

Steve, if you can pass on the lines (assumed to be 120), aspect ratio and frame rate that might help...what would be the source of signals? Your Aurora converter? Any data/timing info would help greatly...I guess you're aiming for next years NBTVA convention...hopefully by then we'll be clear of this dreaded virus thing so it can go ahead...
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: Starting at the beginning with Mirror Screws

Postby Panrock » Sat Aug 29, 2020 8:32 pm

Hi Klaas.
Klaas Robers wrote:Steve, these laser diodes on their own don't give a parallel focussed beam. When you look into the specification you find the "divergence", which has a FWHM (Full Width Half Maximum) of 7 deg in one direction and 22 deg in the other direction. You can see that too in the graph a few pages further.

Laser diodes always give an oval shaped beam, where the points of origin are not identical, but are still very close to each other. In a laser pointer there is a focussing lens that tries to make the beam parallel in both perpendicular directions. Without that lens you have the oval spot.

Thanks for pointing this out. I thought the specified beam widths were looking a bit wide... and I didn't know about the oval spot.

Klaas Robers wrote:You should not modulate the laser, but switch it fast on and off. By pulse dutycycle modulation you may get an analogue modulation, without any form of gamma.

Yes, good idea. When you mentioned this, I immediately thought of Steve A's PWM circuit, which he very kindly designed for use with my 'Grosvenor' (with later added aperture correction too!). Steve A originally intended this to cope with 80-line NBTV, so it could perhaps be 'stretched' to cope with 120-lines... though this is also using a relatively fast 25Hz frame rate...read on.

Steve Anderson wrote:Good see you back Steve...

Nice to be back! I've been rather occupied with building a Band I transmitter. This blows out so much waste heat, it could heat a house on its own. Could be useful now it's turned colder.

Steve Anderson wrote:Klaas is correct, I did try to modulate a laser diode with an analogue current (not voltage) source, the problem comes from the 'lasing current' [1] below which it simply stops emitting light...around 30% of the nominal current. So you lose all the detail in the darker scenes. With colour/RGB each diode will 'give up' at a different point leading to severe colour shading in the lowlights.


This sort of useful information you only get from those who 'have been there first'; ie. posters on this forum. I did have an inkling of this, having many years ago built a (poor) laser diode modulator circuit for use with a red torch on a Mihaly-Traub outfit. I also remember how the diodes would immediately destruct at the merest suggestion of a noise spike at switch on.

Steve Anderson wrote:PWM or similar is the way to go. Very similar to what you did on a previous LED mirror-screw monitor, in fact that circuit could be adapted with a few simple mods. You may recall my laser-link in the newsletter which used the same concept (no idea of the issue number).

I'll go and have a look for this.

Steve Anderson wrote:Incorporating Gamma correction would require a modified ramp generator which drives the three comparators. You could use a 'breakpoint corrector similar to the one of mine also published in the newsletter many years ago (again, no idea of issue number). Actually I've just thought of an easier/better way to do this - let me look into it...

This isn't a CRT display. I don't see why we need to bother with Gamma correction at all. Though if we do, it can be switched in and out on the Aurora.

Steve Anderson wrote:There's a problem with going the micro route and processing three signals at 120 lines - the class of micros I'm used to simply cannot process that amount of data at the speeds required, even at a 64MHz clock speed. I can't recall what frame rate you use but even at 12.5Hz it's too much data. So it's the old analogue method...

I think this was your reasoning when designing the original analogue drivers for my 120-line screw. This, incidentally, rotates at 25Hz, not 12½Hz. It uses LEDs, not lasers. As stated previously, I'm wondering whether I can now use bright LEDs instead of lasers here... with a line-generating and collimating lens intended for lasers. This would mean I could use the existing drivers.

Steve Anderson wrote:Steve, if you can pass on the lines (assumed to be 120), aspect ratio and frame rate that might help...what would be the source of signals? Your Aurora converter? Any data/timing info would help greatly...I guess you're aiming for next years NBTVA convention...hopefully by then we'll be clear of this dreaded virus thing so it can go ahead...

25Hz. 120-lines. 4:3 horizontal oblong. Aurora WC-01. RGB parallel colour.

I'll know next week whether this will take the LED or Laser Diode route. If laser, I'll be looking for a suitable PWM driver, with soft start.

Finally (one can never say this enough): thanks for the help!

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: Starting at the beginning with Mirror Screws

Postby Steve Anderson » Sun Aug 30, 2020 12:50 am

Panrock wrote:Finally (one can never say this enough): thanks for the help! Steve O

Now in my dotage and on the scrapheap I have plenty of time...though I do protest my capabilities...

Enough of that...I would go with what you know works, having done something similar to this in the past it might just need a degree of refinement...primarily based on bandwidth requirements..and mechanical constraints...I still feel LEDs are better than trying to use lasers, from an optical viewpoint and possibly modulation irregularities...no idea what, but you never know...LEDs (Luxeons) are familiar and a known quantity...I suggest sticking with them or even the chain of LEDs used before...

I think Klaas is correct. solid-state laser diodes may be more than a little disappointing...great for on-off applications (PWM) but not linear...plus there is their optical constraints...

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: Starting at the beginning with Mirror Screws

Postby Panrock » Thu Sep 03, 2020 2:07 am

I've now obtained a laser line-forming lens and two different examples of narrow-angle LEDs to try. I had been hoping one of these might have compatible optical characteristics with an uncollimated laser diode, but it appears it doesn't. No line is formed.

The next logical step would therefore seem to be to design and make a laser diode driver circuit to enable a diode to be safely fired up. Then try some examples of laser diode with the line lens.

However, I am worried about the safety aspects. A rotating mirror screw is going to reflect the laser beams all over the place, besides where they're wanted - on the favoured screen for viewing. Indeed, there could be scores of uncorrected vertical sync displaced 'pictures' projected all over the room! Anyone placing their head in the path of these beams could get an unpleasant dose of flickering laser light. Admittedly, these would not be tight stationary laser beams, but moving beams stretched into narrow vertical lines and chopped up by the screw slats. But would this be safe? Cumbersome light screening measures might be needed.

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: Starting at the beginning with Mirror Screws

Postby Steve Anderson » Thu Sep 03, 2020 3:20 pm

Panrock wrote:The next logical step would therefore seem to be to design and make a laser diode driver circuit to enable a diode to be safely fired up. Then try some examples of laser diode with the line lens....Cumbersome light screening measures might be needed. Steve O


If you have the datasheets for the laser diodes (or can point me to them) I'll have a look-see. Are you starting from scratch or modifying your previous mirror-screw monitor? Some of the electronics could be re-used, perhaps with a few mods.

Well some of your previous mechanical cameras/monitors weren't that small or lightweight. For transport the light screening could be made out of cardboard and folded up. When needed just unfold and there you have it. It only has to block light so it can be lightweight. (Is that a pun?). It's a prototype so I wouldn't be too concerned over the 'beauty' of the device. If it works as hoped it can be 'tarted up' later.

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: Starting at the beginning with Mirror Screws

Postby Panrock » Fri Sep 04, 2020 2:19 am

Hi Steve A,

Yes, I'll be modifying my existing Mirror Screw. :mrgreen: (Incidentally, what is 'MrGreen' supposed to signify? Possibly how I'd feel at the prospect of ever having to make another!)

I've had a nose about on Ebay, but it looks a good range of R G and B laser diodes is to be found at RS, out of which I've selected:

877-5241

161-1868

758-7810


Most 'blue' laser diodes seem distinctly violet rather than blue. Though these ones should be OK. The picture colorimetry is rather out of my hands.

On further reflection, I agree that light masking should be possible. Perhaps the three laser diodes with lenses could be mounted facing inward, side-by-side below the centre of the back-projection screen bearing the 12-inch 120-line 'color' picture... all done with mechanical television... Yum!!

I'm hoping that the above laser diodes are a standard fitment that will just push into the lenses I shall obtain.

So, next to make the Drivers. Grateful for the assistance offered Steve A.

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: Starting at the beginning with Mirror Screws

Postby Steve Anderson » Fri Sep 04, 2020 1:13 pm

Those laser diodes all seem good choices, my first thought is whether they may need some form of heat-sinking, there is quite a moderate amount of power being dissipated in each one...though I didn't notice any mention of it in the datasheets. I'll look back over the previous build(s) and see if they're adaptable to drive these diodes..

I assume only one diode per colour, correct? I've just looked them up on the RS site...25 quid each! Plus VAT! Wow!! So I would hope it's only one per channel!

Steve A.

I think 'Mr. Green' is supposed to mean an 'Evil grin'.

The RS site has a link to an application note which shows heat-sinking...in this application I think it's going to be essential. I assume the green and blue laser diodes have similar or the same application note.
Attachments
Application note Red laser diode.pdf
(185.51 KiB) Downloaded 325 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: Starting at the beginning with Mirror Screws

Postby Steve Anderson » Fri Sep 04, 2020 4:52 pm

Update on previous PWM modulator/driver for the laser diodes.

Laser Diode PWM 2-Model.gif
Laser Diode PWM 2-Model.gif (14.33 KiB) Viewed 8619 times


Note that the laser supply has been dropped to 12V and that there's two 12V supplies, +12VA and +12VB, they need their own regulators to keep the PWM switching currents away from the input circuitry. Two 7812 should do fine, the +12VB regulator will need a heatsink, the chassis being ideal. Don't forget the input/output capacitors for them.

Some 'fiddling around' may be required for each R11, but try it as it is at first.

Also as I suggested before, changing the TIP3055s to BF259s may be worthwhile as they're faster, transitioning the threshold point of the lasers rapidly. But try it as it is first. It's also worth considering switching MOSFETs.

There are no changes to the ramp circuit, input op-amps or comparators, only the output drivers.

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: Starting at the beginning with Mirror Screws

Postby Panrock » Fri Sep 04, 2020 11:42 pm

Thanks! :-)

This is, in the main, the old Grosvenor circuit, which continues to work - with the WC-01 - very well today, giving cracking definition at 30-lines, with the addition of your aperture corrector (no longer needed for this application).

It looks like I shall have to be very careful with powering arrangements. The application note even talks of avoiding switching fluorescent lights on and off in the vicinity! Would a 1N5818 spike limiter diode be a good idea across the 12v? Yes, one laser diode per colour raster. These are expensive and I won't want to blow them. To put this into perspective though, the budget in parts for the Mirror Screw ended up well into four figures.

I haven't yet checked, but I hope proprietary heat sink rings will be available. The profile of one of the laser diode cases is different to the other two.

Can you remember the switching frequency, which appears to be unchanged? The video bandwidth is 240 KHz.

Cheers,

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: Starting at the beginning with Mirror Screws

Postby Steve Anderson » Sat Sep 05, 2020 12:48 pm

Panrock wrote:Can you remember the switching frequency, which appears to be unchanged? The video bandwidth is 240 KHz. Steve O

Ah! That does need addressing. I think the original ramp generator ran at around 100kHz, it needs to be 500kHz or more now, I doubt that circuit is up to the task. I'll look into it.

Now the PWM frequency is 500kHz or more, those '3055s are going to have to go, even the BD139s might be marginal. So it looks like switching MOSFETs it's gonna be. Each one is switching well under 1A so they'll be quite easy to find.

I'll also check out the op-amp bandwidth, the TL081s might be OK, if not I'll try and come up with something...

Steve A.

I've gone back on the MOSFET decision, their gate capacitance is quite awkward to get around even at these low frequencies. So re-examining the BD139's I suggest we go with them, they're actually faster than I thought...an Ft of 100MHz...I doubt they'll need heatsinks, but have a few small ones handy and allow space for them.

Although the TL081s should be OK I've changed them to LF356s, same packaging and pin-out, somewhat greater bandwidth.

I'm of the mind to change the comparators to LM319s. Two comparators in one 14-pin DIL package. As they are faster than the 311s the 4th/spare one may be OK in the ramp generator...I'd like to try them myself first though before committing others to them. Though that does mean a trip across Bangkok to get them, but I was due to go soon anyway, so no problem...now we have the elevated Skytrain and the extended underground there's no road transport required - so much easier, quicker and cheaper...unlike two decades ago...
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: Starting at the beginning with Mirror Screws

Postby Panrock » Sun Sep 06, 2020 5:54 am

You're going to a lot of trouble over this Steve A!

Just make sure you only fit this in when genuinely convenient.

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: Starting at the beginning with Mirror Screws

Postby Steve Anderson » Sun Sep 06, 2020 12:37 pm

Panrock wrote:You're going to a lot of trouble over this Steve A!

Not really, assuming it functions as advertised I can file the circuit away knowing that if I need something like this in the future I can just drop it into whatever application I'm working on. A few of the NBTV things I've done here have had a commercial function sometime later. As for going across town to get some bits, I was going to go this week anyway...

Steve A.

P.S. The BD139's could be replaced by BD135's or BD137's, the same but with lower voltage rating, but all OK for this application.

I also suggest getting hold of about 1m of screened audio cable (single), it doesn't need to be video or RF coax as the extra capacitance of the audio cable will help to reduce the chances of static or other short pulses blowing the laser diodes. The stuff that's often included with domestic audio gear.

As these diodes are quite expensive it might be worth investing in a pair of 'shorting pliers' so there's no chance of zapping the diodes as you work on them. When you close them there's a latch at the handle end that holds them firmly closed, the inner surface of the jaws are serrated ensuring good contact. Or use ordinary ones with a strong rubber band at the handle end. I guess you could get away with a 'croc clip' or even a paper clip.

I think I got the ones shown below in a shop in Singapore, many moons ago...to release them you 'sort of' twist them, it's obvious when you have them in your hand...unlike scissors they're ambidextrous...I wonder if the "Left-Hand Shop" in London is still going? Everything they sold was for left-handed people...it was somewhere in the Wardour Street area... no, I'm not left-handed, but I had a friend who is and he told me about this place...thinking more, surgeons use similar tools to close vital passageways as they work on your internals, though I would think the jaws had smooth inners...got a user friendly doctor? I think they call them 'clamps'.

I've found the left-handed shop...

http://www.anythingleft-handed.co.uk/shop.html
Attachments
Shorting Pliers 1.jpg
Shorting Pliers 2.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: Starting at the beginning with Mirror Screws

Postby Panrock » Mon Sep 07, 2020 3:21 am

Hi Steve A.

The BD139's could be replaced by BD135's or BD137's, the same but with lower voltage rating, but all OK for this application.

Got those in stock!

I also suggest getting hold of about 1m of screened audio cable (single), it doesn't need to be video or RF coax as the extra capacitance of the audio cable will help to reduce the chances of static or other short pulses blowing the laser diodes. The stuff that's often included with domestic audio gear.

Should have some of that lying around too.

As these diodes are quite expensive it might be worth investing in a pair of 'shorting pliers' so there's no chance of zapping the diodes as you work on them. When you close them there's a latch at the handle end that holds them firmly closed, the inner surface of the jaws are serrated ensuring good contact. Or use ordinary ones with a strong rubber band at the handle end. I guess you could get away with a 'croc clip' or even a paper clip.

Yes, the proper thing is worth getting here. Could prove cost-effective!

I wonder if the "Left-Hand Shop" in London is still going? Everything they sold was for left-handed people...it was somewhere in the Wardour Street area... no, I'm not left-handed,

But I am! :lol: Thanks for the link.

Steve O

PS. I've just gone and run up my Grosvenor, which uses your circuit. This colour Nipkow display is "daylight bright", so much so in fact that the flicker is objectionable at 12½ Hz. So I now run it at 16 Hz. Your circuit remains sharp and loafs along...
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

PreviousNext

Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 32 guests