A new 90-line Mirror Screw

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Re: A new 90-line Mirror Screw

Postby Panrock » Wed Oct 09, 2024 6:42 pm

Hi Steve A,
Steve Anderson wrote:If you can scan said 'missing chunks', that's an improvement on nothing! If it's basically ledgible that should be OK, no real need for OCR. I realise this would be quite an effort, so don't go crazy on this.

The .pdf in our possession is the 1st May 2016 edition. The paper copy is the 21st July 2016 edition. There may be numerous differences between the two. I therefore will scan the entire later edition as an image file set and let you have it.

As the present custodian of Karen's Timing Corrector, perhaps I should make clear that the existing unit will not be modified - but preserved untouched, to be used and enjoyed as it is. Discussions about possible developments, chip updates and other improvements could be useful and interesting though, since they might lead to the development (and eventual construction) of another design: perhaps a 'club' Timing Corrector(?) Such discussions are likely to remain well above my head!

Steve Anderson wrote:One thing I need clarifying, is the frame rate 25fps or 12.5fps at 90 lines? What, if any, are the line and/or frame sync arrangements?

The frame rate will remain at 25Hz for 90-lines. In Karen's own words: "lines are identified by simple counting". Both line and frame pulses are taken from the Aurora. The frame pulse resets the count.

Steve O
User avatar
Panrock
Green padded cells are quite homely.
 
Posts: 892
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: A new 90-line Mirror Screw

Postby Steve Anderson » Wed Oct 09, 2024 7:50 pm

Hi Steve, with perhaps further scans from you I may be able to reconstruct the waveforms required, like back porch, sync duration, front potch, and the same for vertical. Are the signals separate (RGB+H+V, five in total?). Are there 'hidden lines' like 625 where only 575 are active with picture information, the rest devoted mostly to field/frame sync or vertical blanking? I'm sorry to be a pain here, but without accurate info an accurate result is unlikely to occur.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5470
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: A new 90-line Mirror Screw

Postby Panrock » Wed Oct 09, 2024 8:18 pm

Hi Steve A,

You're never a 'pain' and your involvement is much appreciated. I'm not well placed right now to start up the Aurora and take a look at the waveforms, but all the information should be available in the Aurora manual I sent you. From memory, its mechanical standards' outputs don't include syncs and things like back porches. They are simple line-after-line signals, with one following immediately on another. Again, from memory(!), the (separate) line and frame reference outputs provided are 1:1 square waves.

Hope this helps,

Steve O
User avatar
Panrock
Green padded cells are quite homely.
 
Posts: 892
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: A new 90-line Mirror Screw

Postby Panrock » Thu Oct 10, 2024 3:12 am

I've now received the first quote for the cutting and polishing of the slats. This is for a goodly sum, but I expected that.

I have communicated the requirement that the block resulting, when the slats are assembled together, should present a flat, lineless, mirror finish on one of its faces. It is this surface which will in effect be seen under the picture! Every blemish will be visible. But get this right, and the reward will be an NBTV image of matchless quality.

Steve O
User avatar
Panrock
Green padded cells are quite homely.
 
Posts: 892
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: A new 90-line Mirror Screw

Postby Steve Anderson » Thu Oct 10, 2024 11:20 am

Panrock wrote:...but all the information should be available in the Aurora manual I sent you. From memory, its mechanical standards' outputs don't include syncs and things like back porches. They are simple line-after-line signals, with one following immediately on another. Again, from memory(!), the (separate) line and frame reference outputs provided are 1:1 square waves. Steve O

OK, I'll have to sit down and study said document. I must have missed all that in my initial quick scan through...

A little bit later....I have to admit I'm getting a bit 'confuddled' here, in the original 120-line version all I did was primarily the RGB LED video driver circuitry, correct me if I'm wrong and have forgotten something. The Aurora does 'its stuff right out of the box'. Karen did the 'heavy lifting' for convergance/registration of the colour channels, plus here some alterations may/will be required for the change to 90-lines and the changes brought about as a result. A reduction of line rate from 6kHz (166.7uS/line) to 4.5kHz (222.2uS/line) assuming the same frame-rate of 50Hz. Or have I again got something totally wrong here? All very possible, my memory is rubbish and it was some years back!

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5470
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: A new 90-line Mirror Screw

Postby Panrock » Thu Oct 10, 2024 6:22 pm

Steve Anderson wrote: All very possible, my memory is rubbish and it was some years back!

Join the club. :)

I don't think there is any pressing issue here that needs attention. As you say, you designed the LED driver circuitry. Its bandwidth will be more than enough for 90-lines.

Karen developed and built the Timing Corrector from scratch. This worked 'out of the box'. I used it both on the 60-line and 120-line versions of my original giant mirror screw. (I now call this the 'lump'). Its slats were designed to be build-up'able into a screw of either standard.

No modifications of Karen's unit were required when switching standards. It wasn't line frequency-sensitive. It simply performed its magic line-by-line until told to start again by a frame pulse. Equally, I'm sure it will be happy working at 90-lines.

So, stand easy and enjoy the ride. :)

Steve O
User avatar
Panrock
Green padded cells are quite homely.
 
Posts: 892
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: A new 90-line Mirror Screw

Postby Panrock » Sat Oct 12, 2024 2:51 am

A view of a single slat, and all the slats as they will be stacked for polishing; in effect a 90-line raster. I'm hoping visibility of the borders between the lines will be minimal.

I'm not using position-indexing pegs and holes this time, unlike with the last screw. This didn't prove accurate enough to be really helpful and there are other ways to set up the critical angles.

Steve O
Attachments
general.png
User avatar
Panrock
Green padded cells are quite homely.
 
Posts: 892
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: A new 90-line Mirror Screw

Postby Dmitrij » Sun Oct 13, 2024 8:28 pm

Hello. It is convenient to use a vernier angle meter to adjust and control angles. Sorry for my English.
Attachments
45741_01_big.jpg
Dmitrij
"Fester, enough of the light-bulbs!"
 
Posts: 8
Joined: Tue Dec 03, 2013 11:29 pm
Location: Russia. Moscow.

Re: A new 90-line Mirror Screw

Postby Panrock » Mon Oct 14, 2024 12:54 am

3дравствуйте!

Thank you for this. Your English is fine. Yes, that looks a very good method.

I am proposing to make a small, accurately-sized Vee-shaped wedge plate, place this next to a slat corner and tap it into position until its edge matches the profile of the supporting slat below. As far as I can see, that should leave the angle pretty much correct. The test will be if 90 such adjustments result in a 360-degree screw. If they don't, I will adjust the 'Vee' until they do.

Steve O
User avatar
Panrock
Green padded cells are quite homely.
 
Posts: 892
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: A new 90-line Mirror Screw

Postby Steve Anderson » Wed Oct 16, 2024 4:03 pm

Steve, I can't recall if the previous version of this mirror-screw had some form of timebase corrector, possibly Karen did this? It's function would be to iron-out any small resisdual slat angular displacement by delaying/advancing the video signal to the 'line-O-lite' by a few milli/microseconds on a line-by-line and frame-by-frame basis.

I also cannot recall which set the timebase/speed of the display, was it the screw/motor combination or the incoming video waveform? The obvious one is the incoming video, but there may be a reason why not.

Steve A.

P.S. I have reversed my previous decision r.e. this forum. I just need to develop a thicker skin. Subject matter closed.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5470
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: A new 90-line Mirror Screw

Postby Panrock » Wed Oct 16, 2024 8:57 pm

Steve Anderson wrote:Steve, I can't recall if the previous version of this mirror-screw had some form of timebase corrector, possibly Karen did this? It's function would be to iron-out any small resisdual slat angular displacement by delaying/advancing the video signal to the 'line-O-lite' by a few milli/microseconds on a line-by-line and frame-by-frame basis.

Yes it did. We were reviewing this when discussing Karen's manual recently. The price for corrections of the line mis-timings become corresponding jagged edges of the picture. A future re-developed edition of the Timing Corrector could perhaps mask these edges out? But such a project - maybe a design for a 'club' version - would be a major undertaking for you.

And don't forget the present design depends on being able to closely work in tandem with the pulses from a WC-01. Not many folk (I know only of one other) have one of these boxes.

By the way, the scanning of Karen's final paper Manual would be a fair task. But I'm more than happy to do this if you would like it.

Steve Anderson wrote:I also cannot recall which set the timebase/speed of the display, was it the screw/motor combination or the incoming video waveform? The obvious one is the incoming video, but there may be a reason why not.

The sheer inertia of the 'giant' mirror screw - along with a convenient feature offered by the WC-01 converter - indicated the best way to arrange 'sync' would be to adjust the converter's timing to match the the screw, rather than the 'traditional method' the other way round (locking the screw to the signal).

So, the screw is run up to about 25 Hz, and a frame pulse is provided by the screw, once-per-rotation. The WC-01 locks to this... (Darryl said the tolerance is 7%.) Ergo... the frame (and line) are instantly correct and there's no 'hunting'.

But how does the Timing Corrector know when there's a new line starting - for it to delay? That'll be provided by the reference line frequency square wave from the WC-01. As for which line it is (and what stored delay it merits), that'll surely be according to the count since the last frame pulse.

Update:
Just received the second quote for making the slats. It's for less than half that of the first one. So, Full Steam Ahead!


Steve O
User avatar
Panrock
Green padded cells are quite homely.
 
Posts: 892
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: A new 90-line Mirror Screw

Postby Steve Anderson » Thu Oct 17, 2024 1:22 pm

This is all a bit of a conumdrum. Is the motor speed locked to the mains frequency? i.e. synchronous? But as the WC-01 receives pulses from the screw I would think it follows the screw. This is an area I've not applied the 'grey matter' to much in the past, but I may have to now!

Once the screw is built up and locked in place as a single lump on its shaft then a 'calibration delay' for each slat would be needed. Or have I missed something completely here?

Seeing the previous version worked OK it may be a case of copying the principle and applying it to 90-lines anstead of 120. With some minor changes (I hope) the same electronics could become switchable between 90 and 120 lines. Others too in theory.

I'm still ruminating on all of this, there are quite a few variables involved! Bearing in mind I have little experience in the electro-mechanical area. (Pun actually unintended).

Steve A.

Update:
Just received the second quote for making the slats. It's for less than half that of the first one. So, Full Steam Ahead!

That makes a change!!
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5470
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: A new 90-line Mirror Screw

Postby Panrock » Thu Oct 17, 2024 3:00 pm

Hi Steve A. "Vex ye not". It really is dead simple.

Steve Anderson wrote:This is all a bit of a conumdrum. Is the motor speed locked to the mains frequency? i.e. synchronous? But as the WC-01 receives pulses from the screw I would think it follows the screw. This is an area I've not applied the 'grey matter' to much in the past, but I may have to now!

The motor for the heavy stainless-steel 120-line screw was originally intended (I understand) to drive a washing machine! It had a governor... so yes, it operated like a synchronous motor, but also had considerable power behind it. Used with suitable pulleys and a belt, it spun the screw at 25 Hz.

The relatively light-weight and smaller aluminium 90-line screw will be powered by an ex-heater fan DC motor from a car. It will be directly coupled to the screw. This will run at any speed you want and the supply voltage will simply be adjusted to get it within 7% of 25 Hz. This is the tolerance band of the Aurora, within which it can adjust the 'speed' of the whole 90-line signal so that it will lock.

Steve Anderson wrote:Once the screw is built up and locked in place as a single lump on its shaft then a 'calibration delay' for each slat would be needed. Or have I missed something completely here?

No, you're spot-on.

Steve Anderson wrote:Seeing the previous version worked OK it may be a case of copying the principle and applying it to 90-lines anstead of 120. With some minor changes (I hope) the same electronics could become switchable between 90 and 120 lines. Others too in theory.

Yes it's really simple. So simple in fact that no switching should be needed between 90 and 120 lines, just as it wasn't at 60. Karen's creation should work as well on any number of lines up to 128.

Steve O
User avatar
Panrock
Green padded cells are quite homely.
 
Posts: 892
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Re: A new 90-line Mirror Screw

Postby Steve Anderson » Fri Oct 18, 2024 11:41 am

Panrock wrote:Hi Steve A. "Vex ye not". It really is dead simple.

Steve Anderson wrote:Seeing the previous version worked OK it may be a case of copying the principle and applying it to 90-lines anstead of 120. With some minor changes (I hope) the same electronics could become switchable between 90 and 120 lines. Others too in theory.

Panrock wrote:Yes it's really simple. So simple in fact that no switching should be needed between 90 and 120 lines, just as it wasn't at 60. Karen's creation should work as well on any number of lines up to 128.Steve O
Hmm, would like to see how Karen engineered this. but I don't know if her very clever creation has been published anywhere? Conceptually it could also be applied to discs and other mechanical scanning displays. It may also clear up the 'hunting' often seen.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5470
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Re: A new 90-line Mirror Screw

Postby Panrock » Fri Oct 18, 2024 1:34 pm

Steve A, my understanding of this is limited and based on how I assume it must work.

I will scan in her paper Manual, which I have not studied - it's above my head - and get a copy to you as soon as possible. This may - or may not - elucidate matters for you.

As far as I know, Karen didn't publish the Timing Corrector on the web. Karen published her other work HERE.

Hope this helps,

Steve O
User avatar
Panrock
Green padded cells are quite homely.
 
Posts: 892
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 1 guest

cron