NBTV Camera Question

[Harry Dalek]

Just a small amount of drift now Harry, 1 sample in 90.

This may be unavoidable, but can I confirm with you that you actually set the recorder to RECORD at 48kHz, and not at some other rate and then converted to 48kHz when you saved it? because that is just as bad. I don't think that is what you did but I just want to be absolutely sure.

Hi Gary

Yes it asks what record rate .....i put in 48000 mono ...

on save it asks the same but also asks to save as 8 bit 16 32 i think 64 i just kept it as 16 i was not sure on that one ?

Was the crystal dead on 400hz gary ? i can see with your software it looked like there was a tiny amount of drift as you say .

I will try and set up to record the reflective sensor.

[Harry Dalek]

That's getting pretty stable now Harry (see video).

I am now getting pretty convinced that the shape of the sync bar is due to irregularities in the encoder - can you record the pulses before they get to the comparator?

I think i can see this Gary on my scope as all the sync pulses together look like the sync bar i am getting wavy ..they should all line up neat then and i might get a nice straight sync bar ..

If i was doing a horizontal system thinking about it the sync bar would be neater but the video would still be a lot of lines not lined up as this is anyway .

Getting off track but if i make another encoder without cutting it out this time and do the pot on the light sensitive tranys emitter ..results should be better ...if any thing this camera is a great test bed for getting a future one better .

[gary]

on save it asks the same but also asks to save as 8 bit 16 32 i think 64 i just kept it as 16 i was not sure on that one ?

Yep that's fine.

Was the crystal dead on 400hz gary ?

No, as I said it varies with about 1 sample in 90, but what I really meant was it varies by 1 sample in 90 video lines - this is about 90 parts per million.

A typical crystal has an error of 100ppm (ish) and so we seem to be right on the money in terms of how accurate we can be. But no, you can't say it is dead on 400 Hz.

[gary]

If i was doing a horizontal system

Then the wavy part would be vertical but otherwise the same...

thinking about it the sync bar would be neater but the video would still be a lot of lines not lined up as this is anyway .

Well that depends on what is causing the waviness - if it IS the encoder then if the sync is processed the sync bar will be straight but the video wavy, and if not processed then the sync bar will be wavy and the video, or rather the picture, will be straight.

Getting off track but if i make another encoder without cutting it out this time and do the pot on the light sensitive tranys emitter ..results should be better ...if any thing this camera is a great test bed for getting a future one better

Well, that would make the pulses the same length (hopefully) but the waviness would still be there - instead of being a jaggedy wavy line it will be a smooth wavy line - anyway worth trying I suppose, it would eliminate another variable.

[Harry Dalek]

Yep that's fine.

Was the crystal dead on 400hz gary ?

No, as I said it varies with about 1 sample in 90, but what I really meant was it varies by 1 sample in 90 video lines - this is about 90 parts per million.

A typical crystal has an error of 100ppm (ish) and so we seem to be right on the money in terms of how accurate we can be. But no, you can't say it is dead on 400 Hz.

[/quote]


No worries Gary ..the crystal oscillator does have a trimmer .

Heres about 10 seconds of the reflective sensor .

It takes a fair while to get up to speed does it slowly i have to have a look what my voltage to the motor is any case as long as its doing the the job it can take as long as it wants !

You know that saying i used before the better the devil you know !..i listen so i will leave the lm317 voltage control alone for a while .

i think now i will look into the encoder side of things again .

[gary]

No worries Gary ..the crystal oscillator does have a trimmer .

Good! in that case it is an issue of how to calibrate it. You know, probably the easiest and best way would be to modulate a led behind the disk and adjust until the picture is perfectly steady. In fact I did a similar thing with just a 555 timer and it worked quite well.... for a while

No need for a fancy modulation circuit as contrast, brightness, and gamma are not important at all.

The beauty of doing it this way is it should account for any discrepancies in the sound card clock.

[Harry Dalek]

Well, that would make the pulses the same length (hopefully) but the waviness would still be there - instead of being a jaggedy wavy line it will be a smooth wavy line - anyway worth trying I suppose, it would eliminate another variable.

Well i had a good look at the encoder while running and what you were saying about the cutting out of the slots.

I think it could be this i have the encoder also a touch off center also i can see a wobble ..perhaps as you were thinking this part of the system is causing it ..

The pulse widths would be changing due to how well not well i should say i cut them out ..i can see they change..

perhaps this and a slight wobble is enough to cause this..

i will see if i can put the new one on a clear plastic dvd disk protector i can swap it with other nipkows then also.

[Harry Dalek]

No worries Gary ..the crystal oscillator does have a trimmer .

Good! in that case it is an issue of how to calibrate it. You know, probably the easiest and best way would be to modulate a led behind the disk and adjust until the picture is perfectly steady. In fact I did a similar thing with just a 555 timer and it worked quite well.... for a while

No need for a fancy modulation circuit as contrast, brightness, and gamma are not important at all.

The beauty of doing it this way is it should account for any discrepancies in the sound card clock.

Ok i will try that...Gary i will see how much play i have with the trimmer ,i am not sure they give much change but looks like its pretty close .

I will see if i can get it out again later tonight .

[gary]

perhaps this and a slight wobble is enough to cause this..

Yes - have a look at the sensor output attached, it's nice to confirm that it is indeed the encoder arrangement that is responsible for the wave. I think it is now highly likely that the wobble you mention is responsible for the waveiness. It is going to be difficult to align it perfectly - now I know why so many (including myself) don't bother with syncs

i will see if i can put the new one on a clear plastic dvd disk protector i can swap it with other nipkows then also.

Don't make the mistake I did and have the encoder under the clear plastic - it doesn't work, there is too much reflection off the disk itself.

[gary]

Ok i will try that...Gary i will see how much play i have with the trimmer ,i am not sure they give much change but looks like its pretty close .

Well it is probably not strictly necessary, however you seem to be so close it would be nice to close that gap.

[Harry Dalek]

Yes - have a look at the sensor output attached, it's nice to confirm that it is indeed the encoder arrangement that is responsible for the wave. I think it is now highly likely that the wobble you mention is responsible for the waveiness. It is going to be difficult to align it perfectly - now I know why so many (including myself) don't bother with syncs

Oh yes it is the encoder ! well thats good to know for sure .
Thats a very good experiment to prove this thing can cause it .
Now i know this i will think about how to adjust it if need be .
i will take note on this one .


Don't make the mistake I did and have the encoder under the clear plastic - it doesn't work, there is too much reflection off the disk itself.[/quote]

Ok yes that would be a problem ...the encoders nice and safe but
I was more thinking since the clear disk would cover the nipkow holes would it effect my light or focus and such i could cut it small i suppose .

[gary]

Yeah, you don't want the plastic covering the apertures that's for sure.

Small is good!

[Harry Dalek]

Yeah, you don't want the plastic covering the apertures that's for sure.

Small is good!

Yes i was worried about covering the holes with the plastic

Do you think a smaller encoder print out getting my head around if size of the encoder print out would make a difference in pulse accuracy..i know for sure now its getting it in the middle is a big one ..

[gary]

no should be fine - it's only the leading edge that is important - as long as the distance between successive leading edges is precisely the same it will work fine.

[Harry Dalek]

no should be fine - it's only the leading edge that is important - as long as the distance between successive leading edges is precisely the same it will work fine.

I will make one up tomorrow looks like both the nipkow and clear disk which is not as thick as the cd will both still mount ok ..

From what you say if i can get it to work without cutting out slots this time my only problem is centering it ,i will try and get this part moveable before sticking it on for good...cutting the center out a bit larger will make it moving about easy .