Synchronisation

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Synchronisation

Postby Andrew Davie » Mon Feb 27, 2017 2:03 am

I was just thinking about synchronisation and the "taped up" 32nd hole, and how it all works.

As far as I now understand it, the NBTVA standard has synch pulses for 31 of the 32 scanlines.
These pulses are then synched via a phase-locked loop with the incoming signal from the IR sensor looking at 31 holes on a disc (the 32nd being taped up).

Presumably the motor is being driven at some voltage, and the PLL modifies the voltage - increasing if the IR pulse arrives late, and decreasing if it is early.
And between pulse detection, the disk has some inertia but is also slowing due to friction. And there's the characteristics of the drive belt to consider.

But simplistically, if you drive the motor at some constant voltage, the disc will rotate at a constant speed. There may be acceleration issues changing from one voltage to another.

So I got to thinking - why is this operating at 400 Hz (32 scanlines at 12.5 Hz)? Well, 31 if you want to be pedantic, so 387.5 Hz.
Why wouldn't it work with just a single synch hole? The electronics already provides a "frame synch" pulse, so the PLL could use this and mask off 31 of the 32 holes.

I find it hard to understand how if a potentiometer can be used to trim the frame lock to "reasonably OK" - that is, a slow roll - then why is it necessary to be doing a PLL at a much higher frequency than humans would use to keep that picture stable. PLL would have much better control on voltage precision, at a much higher rate (even at 12.5Hz) than a human. It should be able to do a splendid job of locking to a frame, with just a single IR hole.

So what am I not understanding? Why isn't it done this (much simpler) way?
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Synchronisation

Postby Klaas Robers » Mon Feb 27, 2017 7:51 am

In principle you are right, however with one sync pulse per revolution the inertia of the disc should be very large. That implies that the correction of the position of the disc should be very slow, and it will take say more than a minute to get the disc more or less in synchronisation.

Then secondly, you would like to have the picture quite stable. With one hole per revolution the position of that hole should be kept very precisely. This is all made easier by a factor of 32 with 32 sync holes per revolution. That is the reason for 32 sync holes per revolution.

And then fame sync....... Leaving out one sync pulse is a difficult solution. The synchronisation solution of Peter Smith gives some form of frame synchronisation, but in many attempts it synchronises at an offset of one line. In my monitor I made a different frame synchronisation, where I have an extra sync hole, one per revolution. That sync hole should coincide with the (missing) sync pulse, or better coincide with the frame pulse coming from the second one-shot. If not, it supresses the line sync and alows to run the disc free. It is easiest if you can control the disc then by hand and steer it to the correct "full frame". At that moment the line synchronisation switches on automatically and fixes the frame in the viewing position. I know that I am the only one doing it in this way.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Re: Synchronisation

Postby Andrew Davie » Mon Feb 27, 2017 11:53 am

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

Re: Synchronisation

Postby Andrew Davie » Mon Feb 27, 2017 12:03 pm

I've been reading up a bit.

As I understand it, standard NBTV systems drive a motor via pulse width modulation - PWM - effectively an ON or OFF signal to the motor delivered at the rate that the synch pulses are coming in from the video source. And Klaas that sits with your comment about the time it would take to synchronise - for if we are only pulsing the motor once per revolution yes that would be true.

However, there are other ways to control a motor, right? By varying the voltage. True, when you apply a low voltage to a motor that is not spinning, probably nothing will happen. But I wasn't thinking of doing it directly. What I was asking if it's done/possible is keeping a "motor voltage" value (MVV) and modifying THAT based on the synch pulse (frame based). When we need to go faster, we don't apply full-power to the motor briefly - we increase the stored MVV, and apply THAT to the motor. In this way, MVV starts at 0 and rapidly increases to "full power", motor will spin to overspeed and then MVV is decreased - motor slows, and we're in our PLL.

The analogy is that the electronics are controlling a potentiometer just like a human would. The potentiometer being the MVV equivalent.

In other words, going from the PLL directly controlling the motor to instead controlling a MVV and the MVV controlling the motor. And the end result of this would be the voltage to the motor would be pretty much exactly that required to run the disk at correct speed. And variations would be subtle and gradual so much easier on the motor, no? I assume it would be quieter/smoother. Also less likely to cause jerkiness in the belt drive.

That's what I have been trying to articluate - hope it's clear. Seems to me to get rid of any problem with taking ages to get up to speed, and also caters for poor-quality video source because the MVV will change slowly compared to a PWM all-on or all-off solution. Would this work?
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Re: Synchronisation

Postby Klaas Robers » Mon Feb 27, 2017 8:01 pm

This is exactly how the CMOS 4046 works.
- There is a capacitor that is charged to a certain voltage.
- This voltage is also the voltage on which the motor runs (without discharging the capacitor).
- If the video sync pulse and the "hole pulse" coincide, nothing is done,
- but if there is a small difference the capacitor is charged or discharged, with a fixed current, during that amount of time.
- Although, to come to speed, the real process of chargeing and dischargeing is somewhat more complicated, this is the principle.

But the inertia of the system should be much larger (slower) than the repetition rate of the sync pulses. This is with a sync pulse repetion rate of 400 Hz, and is not with a rate of 12.5 Hz. So the disc should not be able to change its speed noticably within the time between two measuring points, sync holes.

That is the reason that I have split up the process of line pulse synchronisation and frame synchronisation. Only if the frame synchronisation is close to its optimal position (between + and - one line) then line synchronisation is allowed. And you hear the line frequency in a mechanical way when in operation, a kind of 400 Hz hum coming from the motor and the disc.

But.... if you think that you can get it to work..... try it. The sync separator gives a V-pulse, which is at the second half of line 1.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands


Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 23 guests