Used to maintain one of these in the BBC "Breakfast Time" studio in the early 1980s. It back-projected animated graphics onto a 5 foot screen for the weather man (Francis Wilson) to stand in front of and wave his arms around, which was quite a tall order at the time. For use in-shot it had to be bright enough to compete with the studio lighting, and we all know the television camera is very unforgiving of colour impurity or shading etc. The diffraction gratings created on the oil film required a very stable video input. If the input was disturbed or switched to a non-synchronous source, a wave of uneven oil would pass across the scanned area, taking about 20 minutes to clear. This looked like a series of bubbles or bright clouds sweeping diagonally across the screen and was very noticeable. On the weather map graphic it looked uncannily like a weather front approaching from the north Atlantic! The getter thing hanging out of the bottom was called the IHSP (Internal Hydrogen Sorption Pump), and it was the main thing that determined the lifespan of the light valve. When the titanium filament went open-circuit you were running on borrowed time because the vacuum would be slowly contaminated by gases from the oil. Then it was time to get out the cheque book and order a new light valve (£17,000 UK at 1980s prices). The other thing that could kill the light valve was explosion of the xenon bulb. This was a several-hundred watt high pressure lamp mounted in a reflector behind the light valve itself. If you tried to operate this lamp beyond its recommended service life the pressure could build up dangerously, until it failed. Bits of broken bulb and lens could then be thrown forward into the delicate input dichroics causing all sorts of damage. There was an hours meter on the projector but I don't recall any automatic shutdown when the bulb hours had been exceeded. Thank you for reminding me of this amazing device and showing its internal details. Even at the time it seemed unbelievable that the concept could work at all, let alone with reasonable consistency and reliability. The routine alignment procedure was long-winded and not at all intuitive, and had to be followed to the letter by a specially trained engineer. Any slip could result in rainbow colours and shading over large areas of the picture, and the only solution was to go back and start again. Casual adjustment was very much discouraged, to the extent that we engineers had to fit a padlock and chain to deter unauthorised fiddling!
@@mikeselectricstuff Yes that CRT base certainly had enough spare pins, though probably the designers figured out that when the getter filament was dead the drive filament would probably be well past it's best anyway, so a rebuild to replace the one will of course also involve replacing the other.
Most interesting teardown I've seen in a long time. Thanks for making the video. I'd love to hear the conversation between the engineers and business people regarding the cost/benefit analysis for that projector.
VERY COOL TEARDOWN! I worked with these projectors a bit in the 1980s, and always wanted to see the innards of the light valve. Unfortunately, the worn out ones were always sent back for rebuilding, so I could never get my hands on one. The rest of the projector is a similar bit of electronics geek porn, with all the required power supplies, scanning circuitry, interlocks and video processing circuitry on dozens of plug-in modules. The units needed 30-45 minutes from a cold start to warm up the oil before you could fire up the lamp and begin the image setup. GE stopped making them in the early 1990s, and the story that I heard concerning the reason had to do with the loss of corporate knowledge of how to manufacture the specialized oil. The key person involved in the production either died or retired and took the secrets with him...
@@tekvax GE is a massive company, withdrawing one niche product isn’t going to put it out of business, but that division certainly. By the early 90s presumably other technologies were reducing the market and/or the profitability needed to re-acquire the technological skills and knowledge required to keep making them.
I wonder if the oil filter meshes were also for stripping charge from it. Those gear-wheel oil pumps are counterintuitive. The oil doesn't go through the middle between the teeth, it gets displaced around the outside of the gear wheels and squeegeed off by the mating teeth.
would make sense .. the thing would run for hours so there would be some kinda build up unless there was done some kinda stripping the charge from it. But that oil, it must have some kinda crazy weird properties, must be some patent on that stuff.
Good idea - I'd think these would be made in clean-room conditions so hard to see they'd need filters to trap particles, and that arrangement would seem odd for a simple crud filter.
They might've expected a certain amount of mechanical wear too, for which filtering would be wise. But given the short lifetime, maybe that doesn't amount to much, and it's not like anything should ever be without an oil film in operation.
The oil pump and the filtering plates also helped even out the oil temperature/viscosity.Oil returning to the sump after being scanned would be warmed by the radiant energy of the lamp and the beam current of the scanning beam. Forcing all the oil through that massive metal structure of meshes and plates (filter and heat exchanger) before wiping it onto the image plate gave a more uniform film.
very cool! Back in the early 80's, I spent a few months repairing Talaria projectors that were being used on Harrier flight simulators being sold to the U.S. Marine Corps. I was just fixing the various circuit boards, but remember watching these warm up and seeing the initial film of oil on the displayed image as the disc slowly rotated. The fact that it used diffraction gratings to generate colors still amazes me!
@@beefchicken Getter ion pumps are amazing. A chamber that sputters gettery metal fom one end to the other. Reasonable starting pressure, good pumping and an ultimate vacuum.
Yes, it is a miniature Ion vacuum pump. Likely needed to scrub any off-gassing from the oil. The molecular sieves above it are to protect the Ion pump from the oil.
I integrated six of these (into one image) for Whitney Houston’s “I’m Your Baby Tonight Tour”, with pneumatic shutters (these have a low contrast ratio), and X Y alignment mirrors along with a huge mount for the six projectors. They take a while to warm up, cool down, and are difficult to tour with!
Props to you. Tell some more stories or they will be lost for all time when you forget/slip the mortal coil. At least here in youtube comments they will survive for eternity.
From an old Ebay Ad for one of these:- "The Talaria was a $100,000 projector, used most commonly in closed-circuit TV. …The Pope owned two, Hugh Hefner owned about a dozen and they were all over the Pentagon and in flight simulation. Microsoft co-founder Paul Allen had one, Steve Jobs had one, and we rented them to all the big rock groups for concerts, including eight for the Rolling Stones for their Steel Wheels tour"
Very nice! I had one of these at my work around 1982. It took 20-30 minutes to get a decent image after startup because the disk turned so slowly. You could see the image slowly become more complete as the disk turned. I remember scouring the service manual for how they made color in a single light valve. I recall that the electron beam deflection was modulated at two frequencies around 100MHz, I believe, for the color signals. The modulation created diffraction gratings in the oil. US patent 3437746 seem to cover it.
The Eidophor used velocity modulation, I think because altering beam current would change the depth of the charge within the oil - may have been something similar, though the tails on the cathode connection on the electron gun don't look like they're designed for 100MHz!
@@mikeselectricstuff Are you sure the tails are not also for transport? It looks like they clip on very roughly, and maybe there was a special cap that went on in the projector.
@@Spirit532 The main light source shone past the electron gun so it had to be robust and small as possible. Another comment above suggested the modulation was added to the x and y deflection signals, wiggles in X would generate patterns at one grating frequency and wiggles in the Y scan would generate patterns at the other grating frequency. I think the tiny X-Y deflection plates would have added those wiggles and the big X-Y plates would have provided the raster scanning.
From the wiki... "The single lens color projector (PJ5000 line) use dichroic filters to separate the white light of the xenon bulb in two channels, Green and Magenta. RGB color separation and processing is obtained using vertical wobbulation of the electron beam on the oil film to modulate the green channel and sawtooth modulation is added to the horizontal sweep to separate and modulate Red and Blue channels. The optical system used in the Talaria line is a Schlieren optic like an Eidophor, but the color extraction is much more complex."
It almost sounds like that tiny variation in the beam pattern causes a change in the direction of the reflected light. And that slight direction change of the beam gets amplified by the whole Schlieren optics thing, because it only needs a very tiny amount of change. Maybe that can separate the red and blue from the magenta light using a chromatic abberation effect? No idea. lol (and Green gets activated like a "monochrome" version of the tube, where the light gets deflected a tiny bit in the vertical direction?)
The lump on the side is certainly a vacuum pump - not entirely sure which kind, but by the looks of it, it's probably either an evaporable getter, or a titanium sublimation pump. The rings are heatsinks, to recondense the vapors(and either activate them, or capture the gas if it's a tisub). The beads are an oil filter - the pump can pump gases, but if it meets oil, there will just be lots of carbon decomposition. The oil is likely a silicone high vacuum oil with special light-valve-y electrostatic-y additives. Looks almost identical some diffusion pump silicone oils - goop at low temp, liquid at high temp, and low vapor pressure(for high vacuum operation). The ceramic electron gun is most likely because they were _very_ disposable - and while rebuilding the rest of the tube wasn't necessary, the gun likely failed quite often due to the extreme operating conditions(imagine how much power you'd need in this beam!). Ceramic vacuum assemblies are very easy to put into mass production - metal to ceramic brazing is fast and cheap, and ceramics can be prefabricated, along with the metal(likely kovar) rings.
Nice analysis! The oil consistency reminds me a bit of Santovac 5, which is a polyphenyl ether diffusion pump fluid. DC705 silicone oil is pretty viscous as well, but less so than Santovac. I've seen recommendations against using silicone fluids in vacuum systems with electron optics, due to the formation of insulating silica coatings on electrodes upon decomposition.
@@Mikkel324 That is a valid concern! I wonder what they're doing to avoid that. Perhaps that meshy filter thing is less for filtering the oil, and more for separating the dead oil from the still good bulk?
I really appreciate the literal breakdown of the device's and the descriptions of the tech it affords. I don't know what in my brain wouldn't allow me to do it myself I'm somehow unable to destroy such a wonderful assembly. Knowing it will never find useful work again I just couldn't do it so thank you.
Fascinating tech. I used to service 3 CRT video projectors, Barco and Sony mainly, as well as doing line-ups at the installations. Never saw anything like this! Some of the CRT projectors had a film of oil flowing on the front of the tubes for cooling, nothing more exotic than that!
@@mikeh4840 Tell us more, is this a legacy production line that keeps turning them out or is this a new plant to provide tubes for legacy applications? Where is this place? Will they let you make a factory walk through video if you ask nicely?
Some of the stuff online dates the Talaria to the early 1980's. I worked at a GE location from 1972-1980 which had at least one GE light valve projector in service during that time for rear projection of "outside" view for aircraft cockpit simulators. I suppose it might have been a development prototype. This was 5-10 years before the commercial introduction of Talaria in 1983. None of my colleagues called it "Talaria" at the time, just "light valve." I'm pretty sure the one used in the project I worked on for a time had more than 10k hours on it before I left.
The talaria replaced the eidophor in NASA's mission control in the early 80's and was replaced again in 1990 by Hughes LCD units, there's a paper on it on the NASA archive
Linotypes and Intertypes are also amazing tech, reputed to have largest number of parts for a machine at the time but most unlikely to be true as a ship would have more. However they were incredibly complex for a single machine. I had the pleasure of operating an C4 Intertype for a few years and repaired it based on available documentation a few times. What is beyond most people these days is the precision required for the Monotype Composition Caster moulds. The moulds had two main parts that should have been interchangeable but because they were lapped together at the factory one mould base would not work with another mould due to hard to measure difference on the mating surface geometry.
Nice breakdown. A friend on Facebook who also used to operate them posted your video. You got everything pretty right as far as I remember. So - color. Did you notice the correlation between the dichroics and the schlieren gratings? That’s because green is created - deformed in order to pass through the paired gratings - by horizontal modulation - a wave applied to the plates and red/blue by vertical modulation. The hard part (apart from dealing with the oil and temperature) of operating was tuning the front optical schlieren to block all the light from the lamp 2 hence all the thumbscrews on the front optics. Making a ‘darkfield’ we called it. Then the split green and magenta waves were applied to plates and that caused deflections in the oil that allowed either green or magenta light to pass through. The extra wrinkle was that video is a horizontally scanned signal, so a black video frame would actually produce a bright green frame. So - an extra cancellation wave was applied horizontally to make your ‘electronic darkfield’. That was a bit of an art form - it was never really perfect. I seems to remember a few adjustments to that. There was also some horrible ‘flatfield’ adjustments to compensate for the natural differences in deflection across the oil surface, to get a sort of flat color field. That also never really worked and was very temperature dependent! Side fact - the talaria op was also first at work to turn on the oil heater for an hour in the mornings!
What a fascinating piece of obsolete technology. I would love to see how these were made. I bet that somewhere in GE's archives, there's video (probably a film) documenting the construction process. Companies are really thorough at documenting their patented processes and if anyone knows where these archives are kept, they might be able to make a few connections and get a copy. People will just have to keep asking around until they happen upon someone with access. It would be a shame to allow this historical information to be lost.
When I was in military service we had one of those in the education room. I had to work really hard on persuading the responsible technician until I finally got him annoyed enough to allow me to take a peek inside that thing.
Just when I thought I seen it all.... Super impressed with this video. Though the destructive teardown made me sad for such a work of art. It was done in the name if SCIENCE! Thanks Mike!
The device that is seen on the outside is a "sublimation vacuum pump", for removing the impurities from the vacuum itself, not the oil, fascinating technology that is still used today, think modern day air filters, similar system of using a very pure gaseous metalloid to attract gaseous impurities to itself, they don't last long however, so 1000 work hours is a bit too long for a system like that, we'd estimate 600 hours before it dips below its peak, over 800 hours and its probably working at less than half efficiency, its also a "off the shelf" unit, probably to keep costs down, usually with custom equipment they commission their own "filters" to be built specific for purpose, in which case we would recommend two "torus", "flat plate" or "disc" units, it would use far less power overall and would make the unit more streamlined, something tells us that costs were tight when it came to this!
I simply can't imagine the man hours, grey hairs and fainancial resource required to turn this thing from a concept that worked on a bench to a (reasonably) robust commercial product!! (and the EOL setup, calibration and test required to get working units out the door of the facility that made them)
The pellets between the getter pump and the main tube look like zeolite. They also help to improve the vacuum in the assembly by absorbing particular gases, water in particular.
Hi Mike, great video. I really enjoy what you do, and as a bonus, the sound of your voice and accent gives me memories of my childhood family doctor(I am in Canada), and it's quite soothing.
My dad always tells the story of one of these projectors holding up the start of a corporate show in London, that it was the day he learned the best and quickest thing to do when there's a problem at was just leave the specialist tech alone to do his job.
Thanks for the look at this Mike… the development and production costs must have been extreme on that… I’d guess the odd getter has to do with the oil breaking down and releasing something that needed to be removed… probably a VOC of some description…
yeah I think that side thing is a fancy getter assembly to keep the vacuum clean. The oil breaking down might cause contamination, or maybe it can off gas over time. those discs look like a large version of a regular tube getter, and there's a bunch of them stacked up.
Very interesting. The vacuum conditioning device reminds me of early Crookes x-ray tubes where the vacuum wasn’t perfect and pressure could be controlled by heating a gas trap. Could there actually been a controlled gas pressure rather than a complete vacuum. I have a feeling there were tubes that used a slight pressure for focusing. Also seen heater devices to release hydrogen in thyratrons.
absolutely love these videos. i understand the english and the mechanics but the electronics are still highly mysterious, but still cool to hear everything explained by someone who knows what they is talkin aboot
That electron tube with white pellets at the end looks like a gas filter I have . I'm absolutely fascinating with this kind of stuff , I have war time vacuum tube circuits and hi kv radar magnetrons that I love to geek out over
@@BlackWolf42- the thing we're good at is war and destruction of our own environment - i'm sure we can teach you a thing or two:) thank you for Elvis though! Hihi:)
It was a Dr. Gretener at the Zurich polytechnic (now known as the ETH), whose name became part of his company’s name: Gretag, which would eventually become part of Gretag-Macbeth, and eventually X-rite, still around today in the field of color calibration. Gretener’s device was the Eidophor, the older product mentioned in the video.
In 1985, when I was working at Swiss national TV, I heard that name "Talaria".... maybe they had one in the studios for back projection. But that's all I know. What I know is they certainly had an Eidophor projector. That one was located vis-a-vis of the aprentices workshop but it was not used anymore and was removed soon after. If you wanna know how Swiss TV looked in 1990 watch this: th-cam.com/video/Z-0Bt3NoaCM/w-d-xo.html
Comeing down the same road over Eidophor to the Talaria I was about to get a tube en ebay, too, but now I don‘t have to anymore as Mike just satisfied my curiosity :-)
@@tookitogo Thank you for your question and Sub. I'm aware of the coincidence, but if you look at my logo, you'll see a fishing boat - a Brixham Beamer, I live in Devon, UK, and I enjoy those local boats. Alas though when anybody TH-cam searches for me, they have to scroll through millions of projector and BMW sites :) All the best, Beamer.
@@beamer.electronics Oh? I don’t see the boat anywhere. The avatar is your face, and the channel background is electronics parts. (I’m on the iPad app, not sure if anything else shows up in a browser.) But that’s lovely, I’m jealous of people who live near the sea! (Switzerland is just a little bit landlocked, and lakes just aren’t the same!) I’ll check out your videos soon, many of them looked interesting, hence the sub. :) Cheers!
That bottom assembly looks like a stack of getter rings. They would contain a coating to absorb oxygen and contaminants from the tube. I imagine the oil would be a source of gasses as it was heated.
..... Well done Mate, I was thinking of customs scratching their heads looking at the X-ray, Checking the address to see if the recipient is Flash Gordon ! !
Think that was designed to service, just take the unit, place in a jig, and bake in an oven at 250C to get the vacuum epoxy above glass transition, and it would fall apart. Likely first break the gas seal, which is easy to do with a hot wire and thermal stress, though likely GE did have the whole assembly in a vacuum when doing this, so as to allow them to not contaminate the oil, or have it catch fire when they allowed it to come back to room temperature. Got to be some real special oil, to not evaporate any amount in a high vacuum., and still act like an oil.
th-cam.com/video/PjKzCRkc7Ew/w-d-xo.html&lc=Ugw6pO06_lVZCZnqAZp4AaABAg gives nice insight :P Supposedly they lost the way to manufacture the oil in the early 90s.
Presumably the electron gun is in ceramic because of the external heating from the lamp it is in front of, coupled with the relative isolation from any form of heat sync, and the filament working temperature, I would imagine it was easier to package it in a housing which withstand the heat. Also being opaque would eliminate a light path around the electron gun avoiding the filters at the back of the tube.
When i was a kid, a friend's Dad had a shed full of decommissioned industrial, medical and scientific equipment. We used to spend weekends tearing stuff down. His dad had an old ammo box full of stuff that would set the Geiger counter off. :/
I used to work with these in the 90's and have always been fascinated with how they work, until today none of the available documentation really helped clear it up. Than you so much for sacrificing this one, everything now makes sense. However, as you say, the colour mixing is still a bit mysterious, but having seen it I feel as though I could solve that challenge if I was charged with designing the thing, whereas the oil valve thing totally bakes my noodle.
Really!!! I would love to know what the most common failure was on these devices. (Besides end of life for the valve) when i started in the av field 20 ones years ago we were still hanging Barco and Sony 3 gun crt projectors. But the company owners and senior managers always talked about light valves in that way people do when they are commiserating how far the industry has come in thier lifetime.
Me too - spent a week at GE's PDPO learning how to set one up - we had one of the first Talaria PJ5055 in the UK, after Samuelson and Anna Valley, who were the importers.
I would like to point out that even though this is a critical component of the projector shown here, it had many other aspects to it including the Xenon Lamp, remote control, digital signal processing (probably worth investigating considering the era), rasterization circuits, chassis, power supplies, and a built-in spanner wrench. I remember reading the specs that indicated how many G forces the projector would operate under.
I suspect the colour comes from the 3 coated glass plates (green on inside) and how it interacts with the pattern on the grating... Maybe controlling the height of the oil interruption mixes the colour?
These were amazing projectors, but they were pretty much killed off by the big Barco LCD "Light Cannon" models (5000 lumens). Which were then killed of by smaller LCD projectors a few years later.
I think this might be one of the valves from a slightly later version of the projector - the page you've linked mentions they made a single-tube version, a two-tube version with one tube for green and one tube for red/blue, and a three-tube version. If this is a red/blue combo tube, that'd make sense...
No... the single light valve PJ (ge had different models, we had a pj5055 that utilized the same "clamshell" (as ge referred to it) as in this video. The "mlv" (gets multi light valve) was used to get higher output/brightness.
Yes looks like a getter assembly and vacuum gauge all in one, heat up the filament to measure the vacuum, and then toast the getter disks to get a higher vacuum and remove the oil vapours.
Good point. It is always a risk when tearing down rare and exotic equipment that they could contain lethally toxic gasses or substances within, and have been assembled at a place or time when warning labels weren't a legal requirement.
at 22:40 wouldn't that encoder still spin even if the inner disk was stuck since its mounted to the motor? or do you think the magnetic coupling is strong enough to stall the motor?
Damn x-ray tubes and photomultipliers tubes look like they are just annoying junk now 🤓 The first projector I seen in action in my life was a huge ceiling mounted 3-CRT Barco, somewhere in the mid 90's, I never saw one of those running sadly.
Used to maintain one of these in the BBC "Breakfast Time" studio in the early 1980s. It back-projected animated graphics onto a 5 foot screen for the weather man (Francis Wilson) to stand in front of and wave his arms around, which was quite a tall order at the time. For use in-shot it had to be bright enough to compete with the studio lighting, and we all know the television camera is very unforgiving of colour impurity or shading etc.
The diffraction gratings created on the oil film required a very stable video input. If the input was disturbed or switched to a non-synchronous source, a wave of uneven oil would pass across the scanned area, taking about 20 minutes to clear. This looked like a series of bubbles or bright clouds sweeping diagonally across the screen and was very noticeable. On the weather map graphic it looked uncannily like a weather front approaching from the north Atlantic!
The getter thing hanging out of the bottom was called the IHSP (Internal Hydrogen Sorption Pump), and it was the main thing that determined the lifespan of the light valve. When the titanium filament went open-circuit you were running on borrowed time because the vacuum would be slowly contaminated by gases from the oil. Then it was time to get out the cheque book and order a new light valve (£17,000 UK at 1980s prices).
The other thing that could kill the light valve was explosion of the xenon bulb. This was a several-hundred watt high pressure lamp mounted in a reflector behind the light valve itself. If you tried to operate this lamp beyond its recommended service life the pressure could build up dangerously, until it failed. Bits of broken bulb and lens could then be thrown forward into the delicate input dichroics causing all sorts of damage. There was an hours meter on the projector but I don't recall any automatic shutdown when the bulb hours had been exceeded.
Thank you for reminding me of this amazing device and showing its internal details. Even at the time it seemed unbelievable that the concept could work at all, let alone with reasonable consistency and reliability. The routine alignment procedure was long-winded and not at all intuitive, and had to be followed to the letter by a specially trained engineer. Any slip could result in rainbow colours and shading over large areas of the picture, and the only solution was to go back and start again. Casual adjustment was very much discouraged, to the extent that we engineers had to fit a padlock and chain to deter unauthorised fiddling!
Thanks- one of the things I love about these kinds of teardowns is getting info from people that used them!
Pity about the pump filament being the limiting factor - they could easily have put a few more spare filaments in there!
@@mikeselectricstuff Yes that CRT base certainly had enough spare pins, though probably the designers figured out that when the getter filament was dead the drive filament would probably be well past it's best anyway, so a rebuild to replace the one will of course also involve replacing the other.
Most interesting teardown I've seen in a long time. Thanks for making the video. I'd love to hear the conversation between the engineers and business people regarding the cost/benefit analysis for that projector.
Mike seriously needs to send the oil for you to analyze!
Ben, this is exactly the kind of crazy thing you would figure out how to DIY...
VERY COOL TEARDOWN! I worked with these projectors a bit in the 1980s, and always wanted to see the innards of the light valve. Unfortunately, the worn out ones were always sent back for rebuilding, so I could never get my hands on one. The rest of the projector is a similar bit of electronics geek porn, with all the required power supplies, scanning circuitry, interlocks and video processing circuitry on dozens of plug-in modules. The units needed 30-45 minutes from a cold start to warm up the oil before you could fire up the lamp and begin the image setup. GE stopped making them in the early 1990s, and the story that I heard concerning the reason had to do with the loss of corporate knowledge of how to manufacture the specialized oil. The key person involved in the production either died or retired and took the secrets with him...
I wonder what that oil is worth per litre... :P
Isn’t that always the way… only one guys knows how it works; when he’s gone, the company goes out of business…
@@tekvax GE is a massive company, withdrawing one niche product isn’t going to put it out of business, but that division certainly. By the early 90s presumably other technologies were reducing the market and/or the profitability needed to re-acquire the technological skills and knowledge required to keep making them.
@@ohnoitisnt Not much for it's original purpose anyway...
I wonder if the oil filter meshes were also for stripping charge from it. Those gear-wheel oil pumps are counterintuitive. The oil doesn't go through the middle between the teeth, it gets displaced around the outside of the gear wheels and squeegeed off by the mating teeth.
would make sense .. the thing would run for hours so there would be some kinda build up unless there was done some kinda stripping the charge from it. But that oil, it must have some kinda crazy weird properties, must be some patent on that stuff.
Good idea - I'd think these would be made in clean-room conditions so hard to see they'd need filters to trap particles, and that arrangement would seem odd for a simple crud filter.
Woah, the big boi himself, fancy seeing you in these parts
They might've expected a certain amount of mechanical wear too, for which filtering would be wise. But given the short lifetime, maybe that doesn't amount to much, and it's not like anything should ever be without an oil film in operation.
The oil pump and the filtering plates also helped even out the oil temperature/viscosity.Oil returning to the sump after being scanned would be warmed by the radiant energy of the lamp and the beam current of the scanning beam. Forcing all the oil through that massive metal structure of meshes and plates (filter and heat exchanger) before wiping it onto the image plate gave a more uniform film.
very cool! Back in the early 80's, I spent a few months repairing Talaria projectors that were being used on Harrier flight simulators being sold to the U.S. Marine Corps. I was just fixing the various circuit boards, but remember watching these warm up and seeing the initial film of oil on the displayed image as the disc slowly rotated. The fact that it used diffraction gratings to generate colors still amazes me!
I wonder if the device on the side was a continuously active getter for gathering impurities and maybe helping to maintain the vacuum.
An active getter. It certainly looks like one.
@@beefchicken
Getter ion pumps are amazing. A chamber that sputters gettery metal fom one end to the other. Reasonable starting pressure, good pumping and an ultimate vacuum.
Yes, it is a miniature Ion vacuum pump. Likely needed to scrub any off-gassing from the oil. The molecular sieves above it are to protect the Ion pump from the oil.
Actually I was wrong. I didn't finish the video before commenting. It's actually a titanium sublimation vacuum pump.
@@hightechstuff2
Really ?
Never heard of it. Please tell us how it works.
I integrated six of these (into one image) for Whitney Houston’s “I’m Your Baby Tonight Tour”, with pneumatic shutters (these have a low contrast ratio), and X Y alignment mirrors along with a huge mount for the six projectors. They take a while to warm up, cool down, and are difficult to tour with!
Props to you. Tell some more stories or they will be lost for all time when you forget/slip the mortal coil. At least here in youtube comments they will survive for eternity.
that things mad! the oil slivering around looks cool. hope you make some eggs in that pan after
From an old Ebay Ad for one of these:-
"The Talaria was a $100,000 projector, used most commonly in closed-circuit TV. …The Pope owned two, Hugh Hefner owned about a dozen and they were all over the Pentagon and in flight simulation. Microsoft co-founder Paul Allen had one, Steve Jobs had one, and we rented them to all the big rock groups for concerts, including eight for the Rolling Stones for their Steel Wheels tour"
The one which cost $100,000 used three light valves, one for each of the primary colours, and a 10kW xenon short-arc lamp.
I love ingenious weird old tech like this and I love the attitude "Here's what we want, unlimited budget, make it happen"
Very nice! I had one of these at my work around 1982. It took 20-30 minutes to get a decent image after startup because the disk turned so slowly. You could see the image slowly become more complete as the disk turned. I remember scouring the service manual for how they made color in a single light valve. I recall that the electron beam deflection was modulated at two frequencies around 100MHz, I believe, for the color signals. The modulation created diffraction gratings in the oil. US patent 3437746 seem to cover it.
The Eidophor used velocity modulation, I think because altering beam current would change the depth of the charge within the oil - may have been something similar, though the tails on the cathode connection on the electron gun don't look like they're designed for 100MHz!
@@mikeselectricstuff Are you sure the tails are not also for transport? It looks like they clip on very roughly, and maybe there was a special cap that went on in the projector.
@@Spirit532 The main light source shone past the electron gun so it had to be robust and small as possible. Another comment above suggested the modulation was added to the x and y deflection signals, wiggles in X would generate patterns at one grating frequency and wiggles in the Y scan would generate patterns at the other grating frequency. I think the tiny X-Y deflection plates would have added those wiggles and the big X-Y plates would have provided the raster scanning.
mike, you have the coolest teardowns on youtube. everytime i see a new upload i know im going to learn something new
Couldn't have said it better myself! He showcases amazing things :)
Thanks for another fascinating tear-down and explanation Mike.
From the wiki...
"The single lens color projector (PJ5000 line) use dichroic filters to separate the white light of the xenon bulb in two channels, Green and Magenta.
RGB color separation and processing is obtained using vertical wobbulation of the electron beam on the oil film to modulate the green channel and sawtooth modulation is added to the horizontal sweep to separate and modulate Red and Blue channels. The optical system used in the Talaria line is a Schlieren optic like an Eidophor, but the color extraction is much more complex."
It almost sounds like that tiny variation in the beam pattern causes a change in the direction of the reflected light.
And that slight direction change of the beam gets amplified by the whole Schlieren optics thing, because it only needs a very tiny amount of change.
Maybe that can separate the red and blue from the magenta light using a chromatic abberation effect? No idea. lol
(and Green gets activated like a "monochrome" version of the tube, where the light gets deflected a tiny bit in the vertical direction?)
And then of course I see Mike uploaded a follow-up vid, nine hours ago. lol
The lump on the side is certainly a vacuum pump - not entirely sure which kind, but by the looks of it, it's probably either an evaporable getter, or a titanium sublimation pump. The rings are heatsinks, to recondense the vapors(and either activate them, or capture the gas if it's a tisub). The beads are an oil filter - the pump can pump gases, but if it meets oil, there will just be lots of carbon decomposition.
The oil is likely a silicone high vacuum oil with special light-valve-y electrostatic-y additives. Looks almost identical some diffusion pump silicone oils - goop at low temp, liquid at high temp, and low vapor pressure(for high vacuum operation).
The ceramic electron gun is most likely because they were _very_ disposable - and while rebuilding the rest of the tube wasn't necessary, the gun likely failed quite often due to the extreme operating conditions(imagine how much power you'd need in this beam!). Ceramic vacuum assemblies are very easy to put into mass production - metal to ceramic brazing is fast and cheap, and ceramics can be prefabricated, along with the metal(likely kovar) rings.
Nice analysis! The oil consistency reminds me a bit of Santovac 5, which is a polyphenyl ether diffusion pump fluid. DC705 silicone oil is pretty viscous as well, but less so than Santovac. I've seen recommendations against using silicone fluids in vacuum systems with electron optics, due to the formation of insulating silica coatings on electrodes upon decomposition.
@@Mikkel324 That is a valid concern! I wonder what they're doing to avoid that. Perhaps that meshy filter thing is less for filtering the oil, and more for separating the dead oil from the still good bulk?
The light output came from an external light source, the e-beam was only needed to scan the oil film which amplified the signal massively.
I really appreciate the literal breakdown of the device's and the descriptions of the tech it affords. I don't know what in my brain wouldn't allow me to do it myself I'm somehow unable to destroy such a wonderful assembly. Knowing it will never find useful work again I just couldn't do it so thank you.
Having built several Schlieren setups in the lab, this device blows my mind! I'm glad you tore it down, this is really really cool tech!!
Fascinating tech. I used to service 3 CRT video projectors, Barco and Sony mainly, as well as doing line-ups at the installations. Never saw anything like this! Some of the CRT projectors had a film of oil flowing on the front of the tubes for cooling, nothing more exotic than that!
Im the guy who is stuck still making those Sony and Panasonic tubes!
@@mikeh4840 Tell us more, is this a legacy production line that keeps turning them out or is this a new plant to provide tubes for legacy applications?
Where is this place? Will they let you make a factory walk through video if you ask nicely?
Some of the stuff online dates the Talaria to the early 1980's. I worked at a GE location from 1972-1980 which had at least one GE light valve projector in service during that time for rear projection of "outside" view for aircraft cockpit simulators. I suppose it might have been a development prototype. This was 5-10 years before the commercial introduction of Talaria in 1983. None of my colleagues called it "Talaria" at the time, just "light valve." I'm pretty sure the one used in the project I worked on for a time had more than 10k hours on it before I left.
I can confirm that I found documentation of exactly this setup online related to a Boeing flight simulator... possibly for military training.
The talaria replaced the eidophor in NASA's mission control in the early 80's and was replaced again in 1990 by Hughes LCD units, there's a paper on it on the NASA archive
I like the demo at the end of the oil towards the end, makes it much easier to visualize how this might work.
1:10 - loved your eidophor talk, i discovered the linotype machine shortly afterward and had fell down a similar rabbit hole
There's a good documentary on the linotype - think it was a kickstarter a few years ago
Linotypes and Intertypes are also amazing tech, reputed to have largest number of parts for a machine at the time but most unlikely to be true as a ship would have more. However they were incredibly complex for a single machine. I had the pleasure of operating an C4 Intertype for a few years and repaired it based on available documentation a few times.
What is beyond most people these days is the precision required for the Monotype Composition Caster moulds. The moulds had two main parts that should have been interchangeable but because they were lapped together at the factory one mould base would not work with another mould due to hard to measure difference on the mating surface geometry.
Nice breakdown. A friend on Facebook who also used to operate them posted your video.
You got everything pretty right as far as I remember. So - color. Did you notice the correlation between the dichroics and the schlieren gratings? That’s because green is created - deformed in order to pass through the paired gratings - by horizontal modulation - a wave applied to the plates and red/blue by vertical modulation.
The hard part (apart from dealing with the oil and temperature) of operating was tuning the front optical schlieren to block all the light from the lamp 2 hence all the thumbscrews on the front optics. Making a ‘darkfield’ we called it.
Then the split green and magenta waves were applied to plates and that caused deflections in the oil that allowed either green or magenta light to pass through.
The extra wrinkle was that video is a horizontally scanned signal, so a black video frame would actually produce a bright green frame.
So - an extra cancellation wave was applied horizontally to make your ‘electronic darkfield’. That was a bit of an art form - it was never really perfect. I seems to remember a few adjustments to that.
There was also some horrible ‘flatfield’ adjustments to compensate for the natural differences in deflection across the oil surface, to get a sort of flat color field. That also never really worked and was very temperature dependent!
Side fact - the talaria op was also first at work to turn on the oil heater for an hour in the mornings!
The "heater" thing you've mentioned is probably a combination of hit wire to break down larger molecules and ion pump to absorb smaller molecules.
Indeed. Looks like a getter stack.
What a fascinating piece of obsolete technology. I would love to see how these were made. I bet that somewhere in GE's archives, there's video (probably a film) documenting the construction process. Companies are really thorough at documenting their patented processes and if anyone knows where these archives are kept, they might be able to make a few connections and get a copy. People will just have to keep asking around until they happen upon someone with access. It would be a shame to allow this historical information to be lost.
When I was in military service we had one of those in the education room. I had to work really hard on persuading the responsible technician until I finally got him annoyed enough to allow me to take a peek inside that thing.
Holy heck, that's crazy! I would of loved to of seen that thing in action! Optical gating for image control using charged oil, that's just genius.
Just when I thought I seen it all.... Super impressed with this video. Though the destructive teardown made me sad for such a work of art. It was done in the name if SCIENCE! Thanks Mike!
The device that is seen on the outside is a "sublimation vacuum pump", for removing the impurities from the vacuum itself, not the oil, fascinating technology that is still used today, think modern day air filters, similar system of using a very pure gaseous metalloid to attract gaseous impurities to itself, they don't last long however, so 1000 work hours is a bit too long for a system like that, we'd estimate 600 hours before it dips below its peak, over 800 hours and its probably working at less than half efficiency, its also a "off the shelf" unit, probably to keep costs down, usually with custom equipment they commission their own "filters" to be built specific for purpose, in which case we would recommend two "torus", "flat plate" or "disc" units, it would use far less power overall and would make the unit more streamlined, something tells us that costs were tight when it came to this!
Are those beads molecule sieves then ?
The 2 separate pumps may be to account for that; one for the first 500hr and the other for the second 500hr of the tube's life.
I simply can't imagine the man hours, grey hairs and fainancial resource required to turn this thing from a concept that worked on a bench to a (reasonably) robust commercial product!!
(and the EOL setup, calibration and test required to get working units out the door of the facility that made them)
You have NO idea how long I have been waiting for a video on light valves :) VERY cool
The pellets between the getter pump and the main tube look like zeolite. They also help to improve the vacuum in the assembly by absorbing particular gases, water in particular.
Hi Mike, great video. I really enjoy what you do, and as a bonus, the sound of your voice and accent gives me memories of my childhood family doctor(I am in Canada), and it's quite soothing.
I get a great joys from seeing how things work, this tear down and the obscurity of the things is amazing.
Love this channel. He does the most interesting teardowns on youtube.
in a world of embedded processors and leds, this is area51 technology
I'm glad to see you still making these sweet videos, Mike. Thanks a lot!
My dad always tells the story of one of these projectors holding up the start of a corporate show in London, that it was the day he learned the best and quickest thing to do when there's a problem at was just leave the specialist tech alone to do his job.
Thanks for the look at this Mike… the development and production costs must have been extreme on that… I’d guess the odd getter has to do with the oil breaking down and releasing something that needed to be removed… probably a VOC of some description…
yeah I think that side thing is a fancy getter assembly to keep the vacuum clean. The oil breaking down might cause contamination, or maybe it can off gas over time. those discs look like a large version of a regular tube getter, and there's a bunch of them stacked up.
Very interesting. The vacuum conditioning device reminds me of early Crookes x-ray tubes where the vacuum wasn’t perfect and pressure could be controlled by heating a gas trap. Could there actually been a controlled gas pressure rather than a complete vacuum. I have a feeling there were tubes that used a slight pressure for focusing. Also seen heater devices to release hydrogen in thyratrons.
absolutely love these videos. i understand the english and the mechanics but the electronics are still highly mysterious, but still cool to hear everything explained by someone who knows what they is talkin aboot
DAMN never thought I'd see one of these getting a teardown on youtube
Those beads on top of the "electron gun" are definitely a vacuum rated descicant, it will take out water vapor and maybe some other small molecules.
Thing under may be an ion pump or something similar. They are used in closed systems that need consistent pumping to maintain a high vacuum.
A truly fascinating and unique device. Thanks for sharing such a detailed and insightful teardown.
I like the "Hot wire getter strip" assembly .. great video Mike 😎👍👍
Never heard of this technology, thanks for the real interesting teardown!
This obviously is an engineering masterpiece! Thank you for sharing this!
That electron tube with white pellets at the end looks like a gas filter I have . I'm absolutely fascinating with this kind of stuff , I have war time vacuum tube circuits and hi kv radar magnetrons that I love to geek out over
I think its an ion pump. Its purpose is to degas the tube, continually. Regular tubes have a passive getter. This is an active getter device.
Mike don’t dispose the oil. Send it for analysis so we can know the composition!
I was worried that is is a PCB based oil. Who knows.
I absolutely love the crazy tech you look at on your channel. Thanks!
King of obscure teardowns !
And at the end, everything come together again :D Very interesting piece of equipment. Now I must know how it works.
That is an alien technology! Whoever invented this was a bloody genius!
Ikr?! I was about to write " ...so that's where a part of the technology they got in Rosewell went!"
@@sanches2 come on guys, give your own species some credit. We're capable of things that would blow your mind.
@@BlackWolf42- the thing we're good at is war and destruction of our own environment - i'm sure we can teach you a thing or two:) thank you for Elvis though! Hihi:)
It was a Dr. Gretener at the Zurich polytechnic (now known as the ETH), whose name became part of his company’s name: Gretag, which would eventually become part of Gretag-Macbeth, and eventually X-rite, still around today in the field of color calibration.
Gretener’s device was the Eidophor, the older product mentioned in the video.
In 1985, when I was working at Swiss national TV, I heard that name "Talaria".... maybe they had one in the studios for back projection. But that's all I know. What I know is they certainly had an Eidophor projector. That one was located vis-a-vis of the aprentices workshop but it was not used anymore and was removed soon after.
If you wanna know how Swiss TV looked in 1990 watch this: th-cam.com/video/Z-0Bt3NoaCM/w-d-xo.html
Comeing down the same road over Eidophor to the Talaria I was about to get a tube en ebay, too, but now I don‘t have to anymore as Mike just satisfied my curiosity :-)
Wow these things actually exist?!? I had the concept in my head, but always assumed it was a strange dream. I cant belive it really works/exists!
this is insane man! are you a programmer by any chance? have you seen Category Theory?
A brilliant dissection of alien technology. All the best, Beamer.
Is it a coincidence that your channel name, Beamer, is the common word for “video projector” in German, Dutch, and a few other languages?
@@tookitogo Thank you for your question and Sub. I'm aware of the coincidence, but if you look at my logo, you'll see a fishing boat - a Brixham Beamer, I live in Devon, UK, and I enjoy those local boats. Alas though when anybody TH-cam searches for me, they have to scroll through millions of projector and BMW sites :) All the best, Beamer.
@@beamer.electronics Oh? I don’t see the boat anywhere. The avatar is your face, and the channel background is electronics parts. (I’m on the iPad app, not sure if anything else shows up in a browser.) But that’s lovely, I’m jealous of people who live near the sea! (Switzerland is just a little bit landlocked, and lakes just aren’t the same!) I’ll check out your videos soon, many of them looked interesting, hence the sub. :) Cheers!
Потрясающая технология, спасибо за очередное интересное видео
That bottom assembly looks like a stack of getter rings. They would contain a coating to absorb oxygen and contaminants from the tube. I imagine the oil would be a source of gasses as it was heated.
..... Well done Mate, I was thinking of customs scratching their heads looking at the X-ray,
Checking the address to see if the recipient is Flash Gordon ! !
16:50 is an active degasser.
The weird electro gun thing that you thought was a vaccum gauge reminds me of a diffusion vaccum pump of some sort...
Think that was designed to service, just take the unit, place in a jig, and bake in an oven at 250C to get the vacuum epoxy above glass transition, and it would fall apart. Likely first break the gas seal, which is easy to do with a hot wire and thermal stress, though likely GE did have the whole assembly in a vacuum when doing this, so as to allow them to not contaminate the oil, or have it catch fire when they allowed it to come back to room temperature.
Got to be some real special oil, to not evaporate any amount in a high vacuum., and still act like an oil.
th-cam.com/video/PjKzCRkc7Ew/w-d-xo.html&lc=Ugw6pO06_lVZCZnqAZp4AaABAg gives nice insight :P Supposedly they lost the way to manufacture the oil in the early 90s.
Presumably the electron gun is in ceramic because of the external heating from the lamp it is in front of, coupled with the relative isolation from any form of heat sync, and the filament working temperature, I would imagine it was easier to package it in a housing which withstand the heat.
Also being opaque would eliminate a light path around the electron gun avoiding the filters at the back of the tube.
When i was a kid, a friend's Dad had a shed full of decommissioned industrial, medical and scientific equipment. We used to spend weekends tearing stuff down.
His dad had an old ammo box full of stuff that would set the Geiger counter off. :/
Dichronic filters.
Could the crt parts be some kind of getter?
It reminds me if a oil delay guitar effect that used oil to store audio.
Cool teardown.
I have this projector in an Anvil Case that was used for one concert. The case is why I bought it. Should I keep it, use it? You want it ?
Please consider donating it to a technology museum. Where are you located?
hi skypuddle, definitely have interest to buy it from you. Where are you located? Thanks! Dennis
I live in Price Utah USA.. This projector works and has less than 100 hours on it.
@@skypuddle what did you end up doing with it? i'd love to see one in action somewhere.
I used to work with these in the 90's and have always been fascinated with how they work, until today none of the available documentation really helped clear it up.
Than you so much for sacrificing this one, everything now makes sense.
However, as you say, the colour mixing is still a bit mysterious, but having seen it I feel as though I could solve that challenge if I was charged with designing the thing, whereas the oil valve thing totally bakes my noodle.
I am/was certified by GE to repair and operate Talaria.
Really!!! I would love to know what the most common failure was on these devices. (Besides end of life for the valve) when i started in the av field 20 ones years ago we were still hanging Barco and Sony 3 gun crt projectors. But the company owners and senior managers always talked about light valves in that way people do when they are commiserating how far the industry has come in thier lifetime.
So tell us more, please!!! And if you have technical docs, service manuals, etc, please share them!!
Me too - spent a week at GE's PDPO learning how to set one up - we had one of the first Talaria PJ5055 in the UK, after Samuelson and Anna Valley, who were the importers.
Never seen anything like that beast.... amazing!
I would like to point out that even though this is a critical component of the projector shown here, it had many other aspects to it including the Xenon Lamp, remote control, digital signal processing (probably worth investigating considering the era), rasterization circuits, chassis, power supplies, and a built-in spanner wrench. I remember reading the specs that indicated how many G forces the projector would operate under.
playing with alien tech in the garage again are we ? .. would be interesting to see such a thing running.
Wow! That's some amazing tech... Back from the days when we used physics to get things done )))
I suspect the colour comes from the 3 coated glass plates (green on inside) and how it interacts with the pattern on the grating... Maybe controlling the height of the oil interruption mixes the colour?
These were amazing projectors, but they were pretty much killed off by the big Barco LCD "Light Cannon" models (5000 lumens). Which were then killed of by smaller LCD projectors a few years later.
That electron gun looking protrusion on the side might be an electrostatic/getter pump for maintaining vacuum.
Or a "titanium sublimation pump"...
That tech is absolutely mental! Love it!
I think this might be one of the valves from a slightly later version of the projector - the page you've linked mentions they made a single-tube version, a two-tube version with one tube for green and one tube for red/blue, and a three-tube version. If this is a red/blue combo tube, that'd make sense...
No... the single light valve PJ (ge had different models, we had a pj5055 that utilized the same "clamshell" (as ge referred to it) as in this video. The "mlv" (gets multi light valve) was used to get higher output/brightness.
Those pellet looks like molecular sieve, the gauge like object could be a getter pump
Yes looks like a getter assembly and vacuum gauge all in one, heat up the filament to measure the vacuum, and then toast the getter disks to get a higher vacuum and remove the oil vapours.
Rule #1 of operating a Talaria... don't ever ever ever ever ever tip it on its side...!!!!
Mike’s eclectic electric stuff👍
Cool teardown. I hope that oil isn't polychlorinated biphenyl (PCB).
Good point. It is always a risk when tearing down rare and exotic equipment that they could contain lethally toxic gasses or substances within, and have been assembled at a place or time when warning labels weren't a legal requirement.
No, certainly not. Way too low of a vapor pressure. This is likely silicone oil, since it operates at high vacuum.
Why did You destroy it? I now restore two of these projectors using this tube...... Maybe need a spare one
11:35
do you repull the vacuum on the tubes? I imagine there's quite a bit of air that got into the tubes through the glue bond over the decades.
Thats a great teardown, such an interesting bit f kit.
beyond amazing tech here!!! wow wow thank you for sharing this teardown --- I wish so much to find stuff like this! love it- m0ar please!
at 22:40 wouldn't that encoder still spin even if the inner disk was stuck since its mounted to the motor? or do you think the magnetic coupling is strong enough to stall the motor?
Is the bit after 25:49 just leftovers from editing? Can probably trim it in the YT editor rather than reupload it if so.
Agreed looks like an ionization vacuum gauge, pretty fancy built in diagnostics. Ohh watched more, is it some kinda getter?
Thank you. This is a technology I didn't know existed.
How does somebody imagine and design this?
They say that some people at MIT are 1000x smarter than other people at MIT...
My induction stove makes patterns in the cooking oil.
Nice to see you back. Love your videos :)
Yessssss another Mike's teardown!!
Hmm, i was wondering how the schmoo was handled for transportation the last time I saw a pic of this with no description.
The electron gun look alike appears to be be a getter stack..
I still have manuals for this, the TLV and the MP for VGA. remember VGA?
Wow, didn't know disturbances in the force could be used to make images :)
Damn x-ray tubes and photomultipliers tubes look like they are just annoying junk now 🤓 The first projector I seen in action in my life was a huge ceiling mounted 3-CRT Barco, somewhere in the mid 90's, I never saw one of those running sadly.
Interesting (to me) to see what looks like flat-flex on the temperature sensor at 7:31. How long's that stuff been about?
Probably the 1960s.