Photomultiplier Tubes camera build

[Harry Dalek]

Hang on sec gary i will check !

Sorry Harry I don't understand the difference between

nbtvcrystal.wav which is good
and nbtvencoderl.wav which is bad

Can you point out the different circuitry - I was under the understanding that whatever speed control you were using you were using the encoder. If you are not using it with the crystal circuitry what is controlling the speed of the motor?

You now have me absolutely and totally confused.

OH i was thinking the file was bad ...every thing plays ok
I get ya ..Gary same circuit just using either the mechanical sync 400hz pulse or crystal clock 400hz to feed via the mono to the video sync mixer so i can view either really just via that switch ! nothing apart from that switch was needed .

OK ...i have a switch ....3 connections center is to your 555 filter monostabe so when i switch the 400hz crystal position it feeds to your 555 mono it to the sync of the video sync mixer ....switch to the other position which is connected to the output of the Lm311 encoder circuit i get mechanical sync ....
The motor control is the same encoder circuit crystal circuit to the 4046 pll so all i am doing swapping the mechanical sync for the crystal clock via that switch to the video sync mixer in other words .
My idea was if the motor is synced to the crystal clock why use the mechanical sync where you see mistakes as the lines not matching up ...only problem i think is more framing ? as theres no missing sync pulse with the crystal clock ...what it looks like i am happy with rolling or not...i am sure others have used the idea ,i tried it on the fly spot in 2012 but i fed the clock right to the sync mixer and the pulse was to wide getting half a picture as i recall and i did notice drift back then as well...its in fact the same crystal come to think of it !
I will post up a schematic of the changes i have made since my first posting up might take me a few days need my notepad

[gary]

Ok, forget the 555 for the moment as that is just a pulse shaper, I think what you are saying is that the input to the *sync/video mixer* is either from the disk sync encoder or from the crystal sync generator?

The problem I have with that, Harry, is that THAT wouldn't effect the speed of the disk, and it seems it is - when you do that you are also changing how you control the disk speed somehow as well - perhaps there is a wiring fault? Or perhaps you are loading down the encoder signal somehow when you switch to that and need a buffer?

In any case it doesn't matter where you derive the mixing sync signal from because when the disk is *locked* to the correct speed they are both the same (at least the leading edge is), and that's all that matters.

The very fact that when you take the sync mixing signal source from the crystal generator the disk is "locked" to the crystal derived reference frequency - if the encoder pulses were wrong in the way they appear to be when you switch the sync mixing signal source to the encoder it wouldn't lock.

If you can't follow what I am writing I will have to draw a diagram - but in any case I don't think it matters, if you fixed whatever causes the speed to go awry when you switch the sync mixing signal to be from the encoder you will still have the slow roll - that is due to an inaccuracy in the crystal itself.

[Harry Dalek]

Ok, forget the 555 for the moment as that is just a pulse shaper,

The very one you sent me in 2012
i drifted away for a while and used monostable ics for a while on other projects but this one back to garys ..can pulse width change with it to so i used it .

I think what you are saying is that the input to the *sync/video mixer* is either from the disk sync encoder or from the crystal sync generator?

Yes

first via your mono theres a brake now between pin 7 ic 1 and the filter 10nf 10k center of that switch to that cap resistor other to pin 7 ic 1 other end of the switch to ic 3 pin 2 ....only changes for this .

The problem I have with that, Harry, is that THAT wouldn't effect the speed of the disk, and it seems it is - when you do that you are also changing how you control the disk speed somehow as well - perhaps there is a wiring fault? Or perhaps you are loading down the encoder signal somehow when you switch to that and need a buffer?

Take note that ic 4 is now the PLL not the 555 bistable i have changed nothing on the motor control notice pin 7 ic 1 is still feeding now the pll and pin 2 of ic 3 to the other pll connection so ..motor control clock is clocked via the ic 3 and mechanical feed back to the other pll connection ....
The video and the motor control do not effect either in the switching side of it

In any case it doesn't matter where you derive the mixing sync signal from because when the disk is *locked* to the correct speed they are both the same (at least the leading edge is), and that's all that matters.

I would say it for me is a very elegant solution

The very fact that when you take the sync mixing signal source from the crystal generator the disk is "locked" to the crystal derived reference frequency - if the encoder pulses were wrong in the way they appear to be when you switch the sync mixing signal source to the encoder it wouldn't lock.

Just that nothing is changed in the motor control only video side of it neat eh !

If you can't follow what I am writing I will have to draw a diagram - but in any case I don't think it matters, if you fixed whatever causes the speed to go awry when you switch the sync mixing signal to be from the encoder you will still have the slow roll - that is due to an inaccuracy in the crystal itself.

I still have not tried the trimmer cap or other caps resistors with it it might be be able to tune it a touch

below also my change to the video sync mixer
BTW gary the crystal clock used as sync start up video is below the picture in that post with all the other wav files i uploaded it again as i thought i made a mistake with the file its 14mb as it 2 and half min for the bounce to stop at i could see and wanted to show it all...its interesting seeing it start up on the crystal its like its an electronic camera or no mechanical side look to it until the bounce appears .
Also BTW i can switch live via recording if you want to see tomorrow between mechanical and crystal synced video !

[Klaas Robers]

Gary, about the roll in free running mode: I timed one line slip in about 40 sec. That will say that the X-tal frequency is 1 / (40 x 400) = 63 ppm too low or high. That is not impossible for a standard x-tal.
Harry, may be you can correct that with the 22pf trimmer capacitor. The problem is: what is your reference ? To start with I would put the trimmer to max. capacity.

Gary, becauseYOUR reference is the X-tal that defines the clock of the computer. That will have a comparable deviation in frequency. So, who is on frequency ? If one is too low and the other too high both deviations add.....
But if Gary switches sync "on" the slow rolling stops immediately. I would not be unhappy with 63 ppm frequency deviation.

And about your test frame Harry, I would like to see absolutely a horizontal line from the left edge to the right edge of the frame. Only then something can be visible of the accuracy of the apertures of the disc. And then too a diagonal line from the upper corner left to the bottom corner right and one from up right to down left would be very usefull. The usual test frames are also needed, but they are for different tests.

And then, can't you make a black level set, or a sync length adjustment? Now everything is grey, while the darkest areas should be black. A PMT has quite some dark current, so you have to make some correction for that. You will see that this improves your pictures a lot.

[gary]

Gary, about the roll in free running mode: I timed one line slip in about 40 sec. That will say that the X-tal frequency is 1 / (40 x 400) = 63 ppm too low or high. That is not impossible for a standard x-tal.
Harry, may be you can correct that with the 22pf trimmer capacitor. The problem is: what is your reference ? To start with I would put the trimmer to max. capacity.

Yes that is my take on it. In this case, as I have previously stated, if Harry simply wants a rock solid picture on TBP for purposes of analysing the myriad of OTHER anomalies that crop up in a camera like this then "tweaking" the slow roll out should be possible.

Gary, becauseYOUR reference is the X-tal that defines the clock of the computer. That will have a comparable deviation in frequency. So, who is on frequency ? If one is too low and the other too high both deviations add.....
But if Gary switches sync "on" the slow rolling stops immediately. I would not be unhappy with 63 ppm frequency deviation.

Absolutely, for all we know this may be the most accurate camera disk speed ever built, and the roll could be being introduced by the computer clock - there is no way to tell unless you have an even more accurate reference.

That's why, as far as I am concerned, the motor speed is working, and working very well within the parameters set by this kind of circuit. I know that I would be very happy with it.

The reason for tweaking, however, occurs when you consider the computer as a measurement instrument that is not adjustable - in this case, for perfection, it's a matter of "sender makes it right". There may be little to gain, but there absolutely nothing to lose.

You might say, "not adjustable? well, that is what the sync processing is for", and that would, in general, be correct, but in my experience it is much better to view the signal in "free running" mode so as to remove a variable from the equation.

The outstanding mystery, however, is why, when Harry switches the sync mixing signal from the xtal to the disk (encoder) does the performance change? - it shouldn't as the motor speed circuit is otherwise unchanged, I am convinced that the switch over is changing the encoder derived sync pulses in some way, and on top of that, to be perfectly correct, the sync pulses *should* be derived from the disk itself, *not* from the reference signal IMHO.

Having said that, keeping the camera in the other mode is OK by me - it works - and the time spent could be well applied to other issues such as noise etc (if Harry cares to do so).

[Harry Dalek]

HI Klaas and Gary i had a read of your posts ...

And about your test frame Harry, I would like to see absolutely a horizontal line from the left edge to the right edge of the frame. Only then something can be visible of the accuracy of the apertures of the disc. And then too a diagonal line from the upper corner left to the bottom corner right and one from up right to down left would be very useful. The usual test frames are also needed, but they are for different tests.

OK that seems like a easy test i will work on that for you .

Reading what you saying about the crystal oscillator i can record the 400 hz it says 400 hz on what i have to read it but you guys might have some thing more accurate to take readings .

Yes that is my take on it. In this case, as I have previously stated, if Harry simply wants a rock solid picture on TBP for purposes of analysing the myriad of OTHER anomalies that crop up in a camera like this then "tweaking" the slow roll out should be possible.

Well that would be good to stop the roll ..again as i recall same circuit different parts apart from the crystal which i have reused here ...that rolled slow as well .

Absolutely, for all we know this may be the most accurate camera disk speed ever built, and the roll could be being introduced by the computer clock - there is no way to tell unless you have an even more accurate reference.

That's why, as far as I am concerned, the motor speed is working, and working very well within the parameters set by this kind of circuit. I know that I would be very happy with it.

I am not sure i posted up a recording just of the clock to measure ...that might be an idea before i start to fiddle with it ...if you both would like a recording .

The outstanding mystery, however, is why, when Harry switches the sync mixing signal from the xtal to the disk (encoder) does the performance change? - it shouldn't as the motor speed circuit is otherwise unchanged, I am convinced that the switch over is changing the encoder derived sync pulses in some way, and on top of that, to be perfectly correct, the sync pulses *should* be derived from the disk itself, *not* from the reference signal IMHO.

ON the mechanical side i was thinking i can see a very slight wobble as the disk as it rotates so might effect the encoder as the opto sensor would see a slight change in size of the reflective area also i did just put the encoder disk on by eye if its not really centered correctly that would give the same result ..also i don't know if the size i am working at that is size of the raster on this scale that makes a nipkow that is small more prone to if theres a pulse width error from the encoder ...don't know .

Having said that, keeping the camera in the other mode is OK by me - it works - and the time spent could be well applied to other issues such as noise etc (if Harry cares to do so).

Yes i like the crystal idea much more as i said its only a matter of a flick of a switch if i wish ...The PMT as noisy thing i think they must be also picking up particles left right and center ....Noise that i see comes from light getting into the camera from a gap any thing that by passes the nipkow is not good shows up as static .

And then, can't you make a black level set, or a sync length adjustment? Now everything is grey, while the darkest areas should be black. A PMT has quite some dark current, so you have to make some correction for that. You will see that this improves your pictures a lot.

No i don't have a good Black level ...i do have a sync pulse width and pulse amplitude adjustments ...Yes could be very much improved here ...i have not been able to adjust any thing to get this better ,the gamma switch is on the best result its worse on the other setting ...any thing you can suggest here i am more than willing to do a circuit change .

[Harry Dalek]

I looked into do the line test my line is a bit thin and not very straight viewing i should of made a better larger bar in any case ...pushed for time tonight so best i can do today .

So linecrystal .wav is that with sync from the 400hz crystal

second is just some symbols as above

third is as above again but better centered

Fourth 32 symbol but switched from crystal to encoder sync

Fifth iron cross again switched from crystal to encoder so on ..

OK i have worked out the drift is mechanical its not drifting now but framed half way down it depends on how accurate the mechanical 400hz is so it will move to that up or down ...

[gary]

I looked into do the line test my line is a bit thin and not very straight viewing i should of made a better larger bar in any case ...pushed for time tonight so best i can do today .

So linecrystal .wav is that with sync from the 400hz crystal

second is just some symbols as above

third is as above again but better centered

Fourth 32 symbol but switched from crystal to encoder sync

Fifth iron cross again switched from crystal to encoder so on ..

OK i have worked out the drift is mechanical its not drifting now but framed half way down it depends on how accurate the mechanical 400hz is so it will move to that up or down ...

Harry, firstly these show that switching the sync mixing from xtal to encoder makes little or no difference to the disk speed or stability at all so the fact that previous tests showed the speed and stability worse for the encoder sync was a coincidence and a furphy.

Secondly, it is unfortunate that you chose to only have one line test because, for some reason, that happened to be the one with least stability (bounce), however it does seem to indicate that the accuracy of your disk apertures is not all that good (yes it's hard to produce accurate disks at home).

Have you ever tried this disk as a monitor? If so do you have any pictures?

You say the drift is "mechanical" - what do you mean by that? It is indeed mechanical in the sense that the disk is rotating at a slightly offset speed (at least as far as the recording device is concerned) but that is almost certainly due to their being a difference between your crystals clock rate and the recording devices clock rate - in any case it is relatively trivial.

Thirdly, there seems to be no point in switching to the encoder syncs for mixing until you correct the error in the trigger network of the monostable.

It's important to realise this makes no difference to the speed and stability of the disk as your encoder pulses, for the purpose of motor speed control, are extracted before that network.

[Harry Dalek]

Harry, firstly these show that switching the sync mixing from xtal to encoder makes little or no difference to the disk speed or stability at all so the fact that previous tests showed the speed and stability worse for the encoder sync was a coincidence and a furphy.

Yep no effect on the motor side of it swapping the sync ...it swaps over very neatly

Secondly, it is unfortunate that you chose to only have one line test because, for some reason, that happened to be the one with least stability (bounce), however it does seem to indicate that the accuracy of your disk apertures is not all that good (yes it's hard to produce accurate disks at home).

OH yes after that i adjusted the encoders LM311 trimmer ...so on the other videos to me on the original TBP it looked like the image was half way down but not moving ...The bounce seem to be when the disk is hunting just off either way 400hz ..

Have you ever tried this disk as a monitor? If so do you have any pictures?

I don't think so Gary this is one of the nipkows i made for the flying spot camera in 2012 i found this one the holes were to small for that just using a luxeon light for that one ...never been tried for a monitor .

You say the drift is "mechanical" - what do you mean by that?

As when my sync to video is on the crystal ...theres a bounce if the mechanical side is hunting around the 400 hz adjusting the encoder lm311 trimmer it can be increased to make the disk go out of sync or stop the roll altogether ...so i am thinking its this .

I am show you but a video wav recording if you like .

It is indeed mechanical in the sense that the disk is rotating at a slightly offset speed (at least as far as the recording device is concerned) but that is almost certainly due to their being a difference between your crystals clock rate and the recording devices clock rate - in any case it is relatively trivial.

Yes as mentioned its like a dog chasing its tail and the difference to catching it or almost .

Thirdly, there seems to be no point in switching to the encoder syncs for mixing until you correct the error in the trigger network of the monostable.

Yes i will add the diode and repeat results should be better you would think i will do that tonight .

It's important to realise this makes no difference to the speed and stability of the disk as your encoder pulses, for the purpose of motor speed control, are extracted before that network.

[/quote]

Yes it would just be a better sync pulse for the video i would think ..i will try and we will see tonight ..i'll post up a test then ,i have time now so i will add the diode .

[gary]

Harry the bouncing is (most likely) a damping issue not a frequency issue - getting the damping right with the 4046 cct is always an issue - if you can get it right by tweaking the trimmer I would call that a win - but have you read Karen's wonderful write up on adjusting that circuit in the newsletter? If not I thoroughly recommend it.

Harry, when I say the frequency is off I am referring to the slow roll NOT the bounce.

[Harry Dalek]

Harry the bouncing is (most likely) a damping issue not a frequency issue - getting the damping right with the 4046 cct is always an issue - if you can get it right by tweaking the trimmer I would call that a win - but have you read Karen's wonderful write up on adjusting that circuit in the newsletter? If not I thoroughly recommend it.

OH i don't have the newsletter Gary i should keep up my subscription I am finding it is temperamental on the bounce where as it is pretty stable in the mechanical sync side

Harry, when I say the frequency is off I am referring to the slow roll NOT the bounce.

Oh ok i can see on my pc scope its the slight hunting that causes the bounce when the hunting slows up there gos the bounce .

Gary the wav files here are not inverted correct my pc flips the signal so invert them
I recorded at 16 bits 48000 that ok
First file is out of pin 7 lm311 just before bounce stops
Second wav is again pin 7 but when bounce stops or best i can see
last is out of pin 3 on the monostable to video sync mixer
I will post the other tests now in the next post

[Harry Dalek]

Next tests...I was looking into the shading problem
The black seems better today mainly i think due to the amount of light,i have no auto light level on this so manually have to adjust as best i can to the light levels
The toy test seems blacker is showing up better here .
Redoing the line test with a thicker line with white square and movement of it line gets better and worse depending on its screen position angle.
last did a grey bar the lighter bars are hard to make out .
these will play i have inverted them

[gary]

Harry the bouncing is (most likely) a damping issue not a frequency issue - getting the damping right with the 4046 cct is always an issue - if you can get it right by tweaking the trimmer I would call that a win - but have you read Karen's wonderful write up on adjusting that circuit in the newsletter? If not I thoroughly recommend it.

OH i don't have the newsletter Gary i should keep up my subscription I am finding it is temperamental on the bounce where as it is pretty stable in the mechanical sync side

Harry, when I say the frequency is off I am referring to the slow roll NOT the bounce.

Oh ok i can see on my pc scope its the slight hunting that causes the bounce when the hunting slows up there gos the bounce .

Gary the wav files here are not inverted correct my pc flips the signal so invert them
I recorded at 16 bits 48000 that ok
First file is out of pin 7 lm311 just before bounce stops
Second wav is again pin 7 but when bounce stops or best i can see
last is out of pin 3 on the monostable to video sync mixer
I will post the other tests now in the next post

Harry, you have something wrong surely? The last one that you you say is from pin 3 on the monostable, but it is rock solid - in previous examples you have given us the sync is all over the place - what gives? That file looks like it's from the crystal not the monostable.

The other files look fine.

EDIT: Hang on is this before or after the diode mod?

[Harry Dalek]

Harry, you have something wrong surely? The last one that you you say is from pin 3 on the monostable, but it is rock solid - in previous examples you have given us the sync is all over the place - what gives? That file looks like it's from the crystal not the monostable.

The other files look fine.

EDIT: Hang on is this before or after the diode mod?

Gary I forgot i had the switch set to the crystal it was feeding to the mono whoops
but yes i do have the diode now in place ..
i will take a reading from the mono again with the encoder feeding it ...

[Harry Dalek]

Heres the encoder sync pulse out of the monostable circuit ...had a lot of trouble to night the switch is now effecting the motor speed increasing it depending on how long i have switched over to the crystal i have bent the board a touch and noticed the same effect of motor speed change so i think i had a bad soldering connection some where i will have to track down .