Photomultipliers in Practice

[Panrock]

From another thread....

I'd be happier leaving the potential divider resistors at 750K. Steve

Sure, I wasn't suggesting you change them, it's just a jumping-off point from where I would start. There actually are reasons to reduce the bleeder resistor values (within reason), but if you're happy with what you've got then there's no reason to change them.

This is certainly something I could better understand.

Presumably the final resistors in the chain are the most critical? As the electronic snowball gathers weight as it bounces from dynode to dynode, in the final stages it's load may become significant w.r.t the high impedance accelerating voltages. I presume that's why decoupling capacitors are specified at these points.

However, if the resistors were too high in the final stages, wouldn't this simply introduce negative feedback (varying though with picture content)? What is the effect of chain resistor values on gamma? Come to think of it, what is the natural gamma of a photomultiplier tube?

Steve

[Steve Anderson]

I presume that's why decoupling capacitors are specified at these points. What is the natural gamma of a photomultiplier tube? Steve

The natural Gamma of a photomultiplier is 1.0, however distortion (quite low) can be introduced along the lines you mentioned due to bleeder resistor values.

Attached is all you'll ever need to know about PMTs. Chapter 5 should elighten one, I'm concentrating on Chapter 4.3.7 concerning noise as with your 80-line camera you're going to have less signal unless a faster lens is used (not forgetting depth-of-field), the scene is lit better and attention is applied to noise.

The 931A for example is good to over 1MHz of light modulation, but it takes some special techniques to realise that. Your current arrangement is probably good to a few kHz, hence my suggestion of the transimpeadance arrangement, but I need some time to wring this through my cranial Z80.

Steve A.

[Panrock]

Attached is all you'll ever need to know about PMTs. Chapter 5 should elighten one, I'm concentrating on Chapter 4.3.7 concerning noise as with your 80-line camera you're going to have less signal unless a faster lens is used (not forgetting depth-of-field), the scene is lit better and attention is applied to noise.

I'll go away and look at this with great interest. Thanks.

Oddly enough, (I hope) sensitivity and noise is not likely to become a serious issue even at 80 lines, with only about 1/8 (or 3 stops) reduction of the light getting through at 30 lines.

In any case the new shorter lens that goes in will be faster - f2 rather than f2.8.

Currently, this full-colour camera is happiest operating in semi-twilight at f3.5. Certainly, it's so dark that I have trouble reading the channels' output level meters!

Stopping down for operation in daylight has meant vignetting became a problem. No, it's not been easy evolving a perfect optical condenser lens system using whatever is to hand! However I expect this problem should disappear with the smaller and less demanding 80-line field.

[Steve Anderson]

Steve, here's my initial thoughts...

Firstly the bandwidth required. Using 80 lines in a 4:3 aspect ratio results in 60 pixels/line, 4,800 pixels/frame. Add in an extra 5% for sync requirements (if used) comes to 5040 pixels/frame. At 12.5FPS = 63,000 pixels/sec. Bandwidth needed is about 32kHz.

The current arrangement has the anode of the PMT feeding a 1M pot which converts the anode current into a voltage with a source impedance of 1M. Now this combined with the PMTs anode capacitance (6pF) results in a -3db point of 27kHz. Add in some socket, pot, stray and amplifier input capacitance, we'll make that a total of (being very optomistic) 20pF, the -3db point comes out to be 8kHz. Not even quite up to the task for 32 lines.

Here's where transimpedance amplifier wins hands-down, ideally its input impedance is zero, and our -3db point zooms skywards. Below is my first suggestion.

The TL071/2/4 series are the low-noise versions of the TL081/2/4 series with less bandwidth, but more than enough for this application. As there is a lot of gain here I have split it into two stages. The first stage does the current-to-voltage conversion and neatly puts the waveform 'the right way up', the second is a conventional voltage amplifier.

R1, D1 and D2 are simply for protection of IC1 should any of the -1kV find its way here. C1 ensures stability and provides a -3db point of 75kHz for this section. C2 also rolls of the frequency response and the combined chain is -3db at 55kHz.

The first stage will handle PMT currents up to 100uA, from what data I have you're currently getting about 5-10uA. This arrangement will provide 1v/uA at the output, (max pot setting) values can be adjusted to suit.

I have shown a screen between the stages, this might not be needed with good layout, but I'm a belt-and-braces guy. What will be required is very thorough decoupling of the supply rails.

Hope this is of use.

Steve A.

[Panrock]

Thanks Steve. I really am most grateful for the intensive effort you're putting into this project. The entire circuitry will have been redesigned by your good self, apart from Pete Smith's standard club sync circuitry. Maybe you'd like to cast your eye over that too? I do get mild hunting on the disc.

I certainly will be incorporating this camera circuit and as a result it should work 100% better when we get to 80 lines.

I can see this project will go on for a few weeks yet, so I now have a corner of my workshop devoted to it and I shall be doing it in parallel with my paid work - whenever I fancy a change in fact.

I'll post developments here as they happen. Will you be coming to the convention to see the fruits of your labours?

Steve O

[Steve Anderson]

Thanks Steve. I really am most grateful for the intensive effort you're putting into this project. Steve O

OK, irrespective of 'Panrock', Steve O it is from now on, perhaps you could explain where 'Panrock' came from?

As for the effort, it's something I revel in. I'm in India at the moment working on a TV broadcast project, which is all fine and well, but I far prefer the component level stuff, it's just sad it doesn't pay as well.

Being far away from my own workshop in Bangkok is a real hinderance, I prefer to knock up the circuits I publish first, even if on breadboard, to see that I have at least got it mostly right.

However, thus far, by you being sorta 'remote controlled' we've come up with a success, or I like to think so. You've done the donkey work, it's easy for me to push a few buttons on a PC.

The chances of me being in the UK around convention time are most unlikely, in the past 20 years or so I have just spent a total of a few months in the UK, mostly in 2006. Even then I wasn't in the UK in April. However I would really like to attend, perhaps 2009? It all depends, like all of us, on our paid work, where we are and our commitments.

As for the Pete Smith motor drive circuit, I've seen so many variations of it that it seems the original has been lost in the depths of time. There are so many mechanical variabilites that unless a mechanical duplication is done, it's going to be less than perfect. There is nothing intrinsically wrong with the circuit, it all depends on mechanical time-constants and the like to correctly set the PPL loop-filter parameters.

Steve A.

[M3DVQ]

As for the Pete Smith motor drive circuit, I've seen so many variations of it that it seems the original has been lost in the depths of time. There are so many mechanical variabilites that unless a mechanical duplication is done, it's going to be less than perfect. There is nothing intrinsically wrong with the circuit, it all depends on mechanical time-constants and the like to correctly set the PPL loop-filter parameters.

Steve A.

I had a great deal of trouble getting the circuit to work at first, but I can get it to behave now with a lot of twiddling of variable resistors

alas I'm back at university now, so I can't look at the sync feedback mixer I added. I added a pot that varies the input to the FET from just a DC input derived trough a potential divider from the supply voltage to just the pulses from the PLL chip. the disc can then be set spinning at just under the right speed by varying the potential divider, and then the pulses can be brought up in amplitude to pull it into exact sync

[Steve Anderson]

I had a great deal of trouble getting the circuit to work at first, but I can get it to behave now with a lot of twiddling of variable resistors

Whimper.....

With no disrespect to M3DVQ (a real name would be nice instead of a serial number), why do people push on with the concept of DC brushed motors? They are simply unsuited to this class of service. M3DVQ, I admire your persiverence.

I published an item in the newsletter years ago about using DC brushless motors. However it seems to have been largely ignored. I got the feeling the drive electronics was 'too difficult'. Yet it was actually quite simple and yielded results that cannot be even considered by brushed motors. Within one frame the variation in the duration of the 10mS fed-back drive pulses was less than 20uS, 0.2%, more than likely due to variations in the accuracy of the rotor magnetic field. Long term there was no error at all...and this without any sync pulse holes being needed in either the camera or display apart from one per rev.

But this is outside the subject matter of this thread, if others wish to start a fresh one, please go ahead.

Steve A.

[M3DVQ]

I had a great deal of trouble getting the circuit to work at first, but I can get it to behave now with a lot of twiddling of variable resistors

With no disrespect to M3DVQ (a real name would be nice instead of a serial number), why do people push on with the concept of DC brushed motors? They are simply unsuited to this class of service. M3DVQ, I admire your persiverence.

I published an item in the newsletter years ago about using DC brushless motors. However it seems to have been largely ignored. I

Steve A.

I used a horribly unsuitable motor, it's an enormous heavy thing.
the reason I push on with using it, is cause it's what I have! and fiddling around with a few resistors is more my style than actually (shock horror) buying anything!

my whole televisor is built out of old odds and sods.

The name's Alistair by the way, the callsign just seemed the only username I could come up with when presented with that terrifying flashing cursor on the registration page

[Steve Anderson]

...my whole televisor is built out of old odds and sods.

..and long may that spirit continue..it's the backbone of all we do here, just we try to impove on the quality of the results we get from spinning saucepans and the like.

..sometimes it can get a little 'heavy', but most of the time we can be viewed as a bunch of typical English eccentrics, those overseas are honouray members.

As for the sign-up page, like you, I have only seen it once, can't recall what it looks like.

Anyway, welcome Alistair, good to see some young blood amongst us old f**ts. I hope you're working on your Degree in English Eccentricity, and eventual membership of the IEE. (Institute of English Eccentrics).

Motto: "You're never alone when you have yourself to talk to."

Steve A.

[M3DVQ]

Anyway, welcome Alistair, good to see some young blood amongst us old f**ts. I hope you're working on your Degree in English Eccentricity, and eventual membership of the IEE. (Institute of English Eccentrics).

Motto: "You're never alone when you have yourself to talk to".

Steve A.

indeed I am, the powers that be have renamed the institute though, the Institute of Eccentric Twits, someone must have caught us trying to get tv pictured out of biscuit tin lids

[Steve Anderson]

[Someone must have caught us trying to get tv pictures out of biscuit tin lids

Reminds me of the 80s when a certain borough councils' galvenized dustbin lids were a perfect porabola for downlinking satellite frequencies. Poor old Aunt Mable was always wondering where her dustbin lid had vanished.

And with that I would like to return to the original thread.

Steve A.

[Panrock]

... perhaps you could explain where 'Panrock' came from?

Well err.. since you are a moderator here and this board seems tolerant of OT posts compared with that 'other' vintage technology board (but what about starting a 'general discussions' section chaps?) I'll tell you...

I once had a very boring job as 'Distribution Clerk' at BBC Wood Norton in the late seventies. The job itself was supposed to consist of assembling technical publications and sending them to engineeers all over the Corporation, but in practice it simply involved sleeping for long periods in a corner on the floor, interspersed with short bouts of frenetic exercise as I rode my bicycle along a track I had constructed amongst the aisles between the storage racks... I digress. Long periods were also spent sitting in the Gents. It was here that 'Panrock' was born.

Panrock was the name coined because at that time I couldn't bring retrieve from my memory the name 'Kilroy'. Panrock spent many happy if tedious hours inscribing his name on the lavatory walls with poetic offerings with titles such as "A Paean of Praise from Old Panrock, Lord of Wood Norton Toilet Halls" or somesuch. No wonder the BBC went into crisis in the '80s.

Well - you did ask....! Mods - I promise no more OT posts.

As for the effort, it's something I revel in. I'm in India at the moment working on a TV broadcast project, which is all fine and well, but I far prefer the component level stuff, it's just sad it doesn't pay as well.

Being far away from my own workshop in Bangkok is a real hinderance, I prefer to knock up the circuits I publish first, even if on breadboard, to see that I have at least got it mostly right.

You sound like an interesting chap with a strong character. I've done a fair bit of gadding about the world too, but the jobs I took on the way were more of the type to fund my onward passage.

Cheers

Steve O

[Steve Anderson]

Well err.. since you are a moderator here and this board seems tolerant of OT posts compared with that 'other' vintage technology board Cheers, Steve O

Well..err.(to quote your own words), I can only assume Andrew has the same viewpoint, hands off is the way we prefer to play it. If you wish to input the humanistic viewpoint of your experience, as in the past tense. then we would all be interested.

I have spent far too much time for example at BBC White City, and what a crushing bore of a place that is.

Well - you did ask....! Mods - I promise no more OT posts.

What might seem an OT subject may well not be......here's one that is OT....the breakfast in the BBC Club is really good!

Stve A.

[Panrock]

Right, to work! Doing the camera will have priority this week, then it's back to my real job.

The top picture shows the camera, the bottom picture with the side cover removed. The right hand mid-level box houses the condenser block, behind it the PMTs and dichroic splitter, and above - the pre-amp bit I shall be changing! The eht unit is at the bottom (with another power supply tacked on).




Steve O