Wavy horizontals

[AncientBrit]

Albert,

I think we all make our own decisions in NBTV on what 'is' and 'is not' acceptable to use.

Some purists will baulk at anything later than valves (tubes).
Others are prepared to embrace more modern solutions such as microcontrollers. And some even combine them!

But that's the fun of this hobby, there's no commercial outlet so little constraint on what path to take.

It all makes for a most absorbing hobby, just wish I'd come to it earlier,

(and sorry for getting off topic)

Kind regards,

Graham

[holtzman]

I think the wavy pattern is caused by mechanical factors. I also experienced various kind of waves, but managed to figure out the reasons - and the reasons may be quite strange.
It looks like the faze of oscillations is the same as the disc rpm. So it's pretty much like the disc goes off-center. Please note that this may be not only because it's mounted off-center on the shaft, but also because the bearings, whole system vibrations, vibration of your camera, etc.
It does NOT look like the problem caused by wrong sync or alterating rpm. If you look carefully, you should notice changes in lines density corresponding the waves. I mean, there are regions with lines closer to each other, and further. This happens if disc is off-center.
Attached is example of another wavy error, horisontal since the format was landscape. These waves appeared because of faulty pattern of holes. It was manually transfered from plotter printout, after the paper had shrinked/extended in one direction. Note the faze is different.

[Viewmaster]

Thanks for you suggestions Holtzman.
Because the camera is mounted sideways its lines run parallel to the 32 lines.
I thought it may be that, but mounting it at 90 degrees makes no difference.
Yes if off centre, lines are closer/further apart. At present, because I can adjust the centering by hand, it is within about a pixel. This will lead to 2 pixel waves but doesn't account for the up to 8 pixel waviness I have seen.
Re vibration of the camera.....Unbolting the camera and moving it about by hand makes no difference either.

Certainly any total picture bounce is caused by RPM variation as a seperate strob flash shining onto the sync wheel shows, so why not the wavyness too? Maybe.

I now believe it is a combination of 2 or more factors, each one adding to the wavyness effect. Small factors, which in themselves I have not noticed, giving the effects I can see.
The bearings are new, sealed ball bearings running free, and I don't believe they are causing any problem. But maybe?
I must try and devise some tests to try to eliminate the various possibilities.

[Viewmaster]

(and sorry for getting off topic)
Graham

No need to appologise for that, Graham, as I started this off topic bit, not you.

Anyway, on this forum, we have very flexible moderators who stretch the
32 lines to their breaking point at times. ......and the forum is the richer for it.

[gary]

The bearings are new, sealed ball bearings running free, and I don't believe they are causing any problem. But maybe?

Do you have, or have access to, a dial gauge Albert?

[Viewmaster]

The bearings are new, sealed ball bearings running free, and I don't believe they are causing any problem. But maybe?

Do you have, or have access to, a dial gauge Albert?

Yes I do Gary, (if I can find it), and I know what you are suggesting.
Will let you know if the shaft is running out of true.....bet it isn't

[gary]

Also measure the edge of the disk (if it is a true circle - it needn't be of course), and the front face of the disk.

I am sure you are right and the shaft is running true, it's just that I have had so many bad experiences regarding runout with off-the-shelf solutions that I just don't trust them, call me cynical...


Oh, and of course runout will not explain why you are getting a variation in the amount of distortion.

[Viewmaster]

Oh, and of course runout will not explain why you are getting a variation in the amount of distortion.

Run out is very very slight. You asked about a dial gauge. The dial gauge on the shaft shows zero when running slow and about .0005 Inches (approx .01 mm) when at 750 RPM.

So, bearing in mind that a pixel on this set up is .004 inches then it is certainly not a shaft problem, as the waviness at 4 pixels (which is roughly the norm) is a .016 inches (.4mm).

I disconnected the 47uF C on the 4046pll output, but except for general picture bounce the waviness doesn't change.

When the camera is on there is some small interferance to the picture, but this cannot be anything to do with the waviness as turning it off and viewing the picture with an eyeglass, the waviness is the same.

Something odd I find is that moving the sync wheel (which has the standard 32/31 blanked off segment) around, relative to the disk, makes little difference in the framing. That does seem wrong and might be a clue.

It's my old arch enemy, the 4046 to blame somehow although when outputting clean sq waves the waviness is still there.

I might take a small bet soon that it's speed variation during every frame that is the main cause.
But previous doubling the weight of the flywheel, (which I would have thought would help to smooth out variations) and also deliberatly running it out of balance doen't seem to alter the waviness.

[gary]

It's my old arch enemy, the 4046 to blame somehow

well that should be easy enough to prove - take it out and do a manual sync - if the waviness disappears there you have it.

[Viewmaster]

It's my old arch enemy, the 4046 to blame somehow

well that should be easy enough to prove - take it out and do a manual sync - if the waviness disappears there you have it.

I am going to completely isolate the motor from any internal circuit.
I shall run the motor from a battery with series pot. That should prove whether it's mechanics or that other stuff.

[gary]

Yep - good idea

[Viewmaster]

With battery driving motor, although the frame line is not as flat as yours, Gary, it is not wavy, and is always the same.

So now rebuilding the sync sep and 4046 circuits on a seperate board . I have rather long leads to the 200k and 10k pots at present, so this may not have been a good thing to do.

Also changed the metal flywheel to a set of CDs (!!).
In this way I can alter how much inertia goes into the system, from 1 up to 10 CDs.

One thing of general interest is that any slight picture bounce is much slower the more CDs in the flywheel.
That is what one would expect, of course.

[gary]

Well done Albert, well that's a new one on me - so it seems hunting is happening on a sub frame basis - so yes a heavier fly wheel should help. But of course, it shouldn't be hunting at all ideally.

[Viewmaster]

Well done Albert, well that's a new one on me - so it seems hunting is happening on a sub frame basis - so yes a heavier fly wheel should help. But of course, it shouldn't be hunting at all ideally.

Maybe a 1 ton flywheel would straighten things out.

I noticed yesterday that if the camera is showing both the top and bottom frame lines of a picture, sometimes the two are of a completely different shape.
The top frame line, say, might be quite flattish and the bottom frame line very curved ....now I don't think this can be camera distortion as it doen't always happen !
Gee, ain't I 'aving fun.

Maybe if I use Des O'Connor CDs as flywheels it will cure the problem.

[Viewmaster]

I took this video of track 45 this morning. The first part, as you can see, shows a very wavy frame line.
I then just briefly touched the disk's flywheel and immediately the frame line greatly improved as shown, giving quite a good picture for the Nipper.
So this quick change is not the disk, bearings etc but some electronic gizmo pulsing the system every frame I would guess.
The slight cross black bands are in the video compression I did to reduce the video download time and nothing to do with the Nipper.

The new sync/4046 board is nealy finished so we shall see.