Building a CRT pip-boy because why not?(Also for school project because why not?)

[cj7hawk]

Here's a good article about the evolution of gas plasma screens. The technology evolved a lot over the years and devices like the Plato screen differ is some ways from more modern screens capable of things like, well, grayscale.


Obviously they're not literally composed of many thousands of independent little bulbs, and most monochrome screens do use an "open cell" design. Color screens, however, need to have the different phosphor colors subdivided by little glass ridges... in other words, they're pretty nightmarishly complex despite not literally being a bunch of tiny bulbs. FWIW, as the article explains, plasma screens don't *have* to be red. Color screens use a xenon plasma that emits ultraviolet light, which in turn stimulates a phosphor. So strictly speaking you can make a plasma screen whatever color you want.

Anyway. As also is made clear by the article, those 1960's plasma displays were *horrendous* power hogs. I don't think you'd be carrying one around on your wrist unless you had a very long extension cord for it. Frankly I think the only real-world tech that even remotely makes sense in the "Pip Boy" would be electroluminescent. Dot matrix EL displays were demonstrated as early as 1965, and non-dot-matrix but pretty complex implementations of it were used for things like the Apollo guidance computer. And notably many of these early EL displays are roughly the same color green as the pip boy display.

Well, if practical were the objective, a 4" OLED with every second line unaddress to create the visual of "scan lines" would suffice would it not?

Though I'm not sure that our concept of practical and @CompaqSniffer 's idea of practical are the same.

I like his project. I think it's a fun idea and I really hope he succeeds -

 

[Eudimorphodon]

Though I'm not sure that our concept of practical and @CompaqSniffer 's idea of practical are the same.

My definition of "practical" here is something a human could actually wear for more than an hour without having it broken the first time you bump into a doorframe. An EL display can be *exceedingly* rugged and they're quite common in military equipment; that is *very much* not true of CRTs.(*)

(* Okay, you'll find them in vehicles of all sorts, but that's not your *wrist*.)

If you *really were* to do something like a "pip boy" with a CRT you'd do it with a viewfinder CRT and hold it up to your eye to work it. I think if I dig around a bit I could actually find an example of a piece of military equipment that works that way. But yeah, it doesn't match the game.

(* edit: when I say “hold it up to your eye” I’m probably picturing something helmet or headband mounted with the CRT over your ear projecting onto a mirror in your peripheral vision field. Picture a prehistoric Google Glass. HID displays like this actually exist.)

 

[cj7hawk]

A good source of viewfinder CRTs is old video cameras, that are old enough that they no longer have value, but not quite old enough that people collect them yet.

Usually they run off of a low DC voltage and have a composite video type input, though in monochrome mostly ( some were color LCDs )

However 1" is a lot smaller than I think @CompaqSniffer was thinking to make this...

Up until a few years ago, it was still cheaper to take old CRT viewfinders and use them for monocular purposes than it was to buy something equivalent.

 

[CompaqSniffer]

I figured since I would already be including a Geiger counter in the case, what harm would it be to include another High-voltage hog?

Also I have heavily considered using a CRT viewfinder and magnifying it with a lens, I have a professor here on campus who is absolutely obsessed with optics, and he would probably love to help.

I'm mostly confused on what the right camcorder I should get for the viewfinder CRT. I know some of them are pretty good on length, but some of them are atrociously long.

If any of you have a good viewfinder CRT suggestion, I could go with it. Maybe I could also build it up like one of those CRTs in back projection TVs with the massive lenses filled with glycol, minus the glycol(and minus the X-rays).

Speaking of X-rays I had my friend measure X-ray emissions from my CRTs and surprisingly the Compaq Portable I CRT actually has a few coming off of it. We could only measure them when the Compaq was open and on the ground though, so as long as you're not standing directly over your Compaq while using it and using it every single day, you should be fine.most X-rays from CRTs are reflected at 90 degree angles from the screen, so you won't get any from being in front of one. Plus most readings are only slightly over ambient. It's cool what you can measure with an old Soviet pancake tube.

 

[cj7hawk]

I figured since I would already be including a Geiger counter in the case, what harm would it be to include another High-voltage hog?

Geiger counters IIRC are around 250 to 400v - Much lower than the 5000+ found in a CRT.

over about 20,000v they start to produce XRAYs, but at least your geiger counter will detect them.

Also, I think the PIP boy would have used an Ion Chamber instead of a geiger counter if it was modeled on the older style of detector that was common. I have both here - geiger counters and ion chamber detectors.

Geiger counters are more sensitive but can saturate so unless you detect that, you detect zero radiation once you get above a certain level.

The older Civil Defense equipment that was made in the US around the 50s was most commonly based on ion chambers.

Also, a slow to read a dosimeter would be cool - again you can buy Civil Defense dosimeters still I think - You charge them up and just read off your dose via a lense system built into the tube itself, so having a slot for one would be cool!

edit: one of these - https://www.ebay.com.au/itm/316121402115

Also, it might not be XRAY's you're detecting. Glass will attenuate xrays, but not eliminate them.

I'm mostly confused on what the right camcorder I should get for the viewfinder CRT. I know some of them are pretty good on length, but some of them are atrociously long.

I used to just find old ones at second hand stores and pawnbrokers. All you can do is but some and try some.

Also, there were some flatter screens in door monitors that used conventional CRTs IIRC. They might also be an option.

 

[cj7hawk]

Oh, and don't forget to find a nice ADM-3A terminal - I've seen those a LOT in the Fallout series. They would have supported the graphics necessary too if they had the graphics expansion.

 

[CompaqSniffer]

Also, it might not be XRAY's you're detecting. Glass will attenuate xrays, but not eliminate them.

X-rays are from the breaking radiation of the electrons smashing against the glass, just did a lab on it a month ago, they call it bremstrauhlong radiation.

I had heard that higher voltage CRTs give off X-rays but never researched it myself. It makes sense, since at higher voltages, the electron has higher kinetic energy, actually as I'm writing this I'm pretty sure voltage is related to how high of kinetic energy your electron has. The more kinetic energy you have, the more the electron has to lose when smashing against the screen. That kinetic energy is released in the form of a photon, in a CRTs case, an X-ray.

Also X-rays are really the only form of radiation (gamma) that can penetrate the shielding in a CRT.

 

[CompaqSniffer]

I used to just find old ones at second hand stores and pawnbrokers. All you can do is but some and try some.

Do you know any good models to search for?

 

[CompaqSniffer]

I think I might have found one, does anyone know if this uses a CRT? It has a color viewfinder:

https://www.ebay.com/itm/267138830115?mkcid=16&mkevt=1&mkrid=711-127632-2357-0&ssspo=-8_i6q-yRoG&sssrc=4429486&ssuid=iglnetqatze&var=&widget_ver=artemis&media=COPY

I think it might be, I found a reddit post about someone who had broken v-hold definitely a CRT

 

[cjs]

If we're going to play this game then *any* kind of electrical device is a "memory". If I flick a light bulb on and off the filament is going to stay warm for a while, do we want to call that a "memory device"(*)?

I guess it depends on whether the devices is intentionally designed to preserve state. A light bulb isn't: nobody cares if it doesn't glow for a bit after being turned off. But CRTs are: they are specifically designed to have phosphors that keep glowing for a while after the electron beam that stimulates it has passed. (Quite a long time, actually, relative to the amount of time that the electron beam is impacting that pixel of phosphor.)

So yes, CRT tubes (at least those intended to be visible to humans, and even many not) are memory devices, and you yourself have pointed out ways other than a Williams tube that they can be refreshed and read.

 

[cj7hawk]

I guess it depends on whether the devices is intentionally designed to preserve state. A light bulb isn't: nobody cares if it doesn't glow for a bit after being turned off. But CRTs are: they are specifically designed to have phosphors that keep glowing for a while after the electron beam that stimulates it has passed. (Quite a long time, actually, relative to the amount of time that the electron beam is impacting that pixel of phosphor.)

So yes, CRT tubes (at least those intended to be visible to humans, and even many not) are memory devices, and you yourself have pointed out ways other than a Williams tube that they can be refreshed and read.

Hmm... That's not entirely accurate - Tubes come from Crookes Tubes long before memory or digital concepts like that existed, and they aren't designed to keep glowing - they are designed to rely on persistence of vision - and the glow is a problem and different phosphors have different characteristics such as how long the glow lasts and what it's decay looks like. That someone noticed this could be used as a form of memory - like like anything that stored data... I've heard of CRTs that used a spring and mechanical transducers to store data, and if you thumped the desk hard enough, you'd get random video characters. People started looking everywhere for ways to store data... TVs sometimes used ultrasonics and a tuned gap to store PAL color from the previous scan. Mercury was already mentioned. Undersea fiber if a form of FIFO memory and they use this in modern gyroscopes.


@CompaqSniffer - sorry - I don't know what to recommend - I just walk into old pawnbrokers where that stuff is cheap, and whatever is available and cheap is the good model to search for

 

[cjs]

Hmm... That's not entirely accurate - Tubes come from Crookes Tubes long before memory or digital concepts like that existed, and they aren't designed to keep glowing....

CRTs are indeed based on principles first developed Crookes tubes (electron beams), but I wouldn't call those memory devices, and, unlike CRTs, they are not only not designed to keep glowing, but many don't glow in the first place. But a CRT is designed to have a persistent glow where excited; it's not just a generic Crookes tube.

- they are designed to rely on persistence of vision -

Actually not. As Wikipedia says:

Since its introduction, the term "persistence of vision" has often been mistaken to be the explanation for motion perception in optical toys like the phenakistiscope and the zoetrope, and later in cinema. This theory has been disputed since long before cinematography's breakthrough in 1895. The illusion of motion as a result of fast intermittent presentations of sequential images is a stroboscopic effect, as explained in 1833 by Simon Stampfer (one of the inventors of the stroboscopic disc, a.k.a. phenakistiscope).​

...and the glow is a problem and different phosphors have different characteristics such as how long the glow lasts and what it's decay looks like.

It's sometimes a problem, depending on whether the glow lasts for a very long time, but more often it's a solution. As I mentioned, a certain amount of persistence is absolutely required: the beam on a CRT excites a chunk of phosphor for well under 100 ns on a typical 15.7 kHz CRT. That's simply not enough for a human to see. The standard P4 phosphor used in black and white televisions continues to glow after that excitement period for milliseconds, tens of thousands of times as long, and P4 is considered a fairly fast phosphor for visual purposes. And, note, longer than typical DRAM refresh periods, too. (The exact figure

I think where we're at cross-purposes here is that some people don't consider something to be a memory device unless the machine that wrote the data can also read it back. Which makes a printer and its paper not a memory device, despite the fact that it's clearly storing information. (Unless you have a scanner, in which case it perhaps is a memory device now?) And by that definition most "storage" CRTs are not memory devices, either.

I use a slightly wider definition of anything that can store information to be read later, regardless of whether it's the same device reading it. It might be just a human, which for me makes a Tek 4010 a memory device.

 

[cj7hawk]

It's sometimes a problem, depending on whether the glow lasts for a very long time, but more often it's a solution. As I mentioned, a certain amount of persistence is absolutely required: the beam on a CRT excites a chunk of phosphor for well under 100 ns on a typical 15.7 kHz CRT. That's simply not enough for a human to see.

The human eye can detect something as small as a single photon... So, let's consider "red" light or the longest at 600nm. A single photon is 600nm long. Smaller still if it's a particle, and I'm simplifying thing here, but it travels at exactly 299 792 458 m/s...

So the period during which your eye sees the photon is approximately 2.3 x 10-15 seconds. That's 2.3 femtoseconds. It's probabalistic, but it happens.

Yes, your eye can see that. And it takes time for that cell in the retina to "reset" to see another photon. that's around 1/10th of a second. So you don't need to see anything after that.

The standard P4 phosphor used in black and white televisions continues to glow after that excitement period for milliseconds, tens of thousands of times as long, and P4 is considered a fairly fast phosphor for visual purposes. And, note, longer than typical DRAM refresh periods, too. (The exact figure

Now if I take my copy of Image Tubes by Illes P Csorba off the shelf, page 180 lists most of the contemporary phosphors used in different tubes. P4 is listed as 20us at 460nm (Blue peak) and 60us at 560nm. (Green peak) for decay to 10% brightness - So persistence of vision is definitely stopping this from making the P4 look yellow, since the blue receptors are finished pretty quickly.

Just to note though - that's Microseconds and not Milliseconds... And while you can make the argument for *some* glow, since it is an exponential decay, it's pretty minimal by the time you reach 1ms, since for each 20us it decays by an order of magnitude in the blue, and for each 60us in green.

P12 would be better for memory if you're slow scanning an entire screen at 50hz... it lasts for 200 milliseconds. A bit slow to change, but it depends on what you're doing with it. Maybe with faster phosphors, you could change the scan rate, or step the deflection. I'm not really sure how they did it when they used a screen for memory. I've never read up on it's characteristics. Maybe glare was a problem? I'm also not sure how they synchronised the scan rates.

My favourites are P-31, P-39 and P-40... P-40 is like an upgrade of P-4 - 150uS at a peak of 440nm ( so it looks slightly bluer ). I have some nice tubes in each.

P-26 (orange) lasts for 10 seconds ! They use it in radar. I don't have one.
P-47 is 80 nanoseconds. - I don't have one of these either.

There's a LOT of phosphors used in technology - even some special screens like P41, which has an 80ns 380nm response making it ideal for light pen applications, and a slower response closer to the visual range.

Image tubes are one of my other technical hobbies. Though not CRTs in a sense. Just analog amplifiers.

I use a slightly wider definition of anything that can store information to be read later, regardless of whether it's the same device reading it. It might be just a human, which for me makes a Tek 4010 a memory device.

I like your definition -

 

[Eudimorphodon]

I guess it depends on whether the devices is intentionally designed to preserve state. A light bulb isn't: nobody cares if it doesn't glow for a bit after being turned off. But CRTs are: they are specifically designed to have phosphors that keep glowing for a while after the electron beam that stimulates it has passed. (Quite a long time, actually, relative to the amount of time that the electron beam is impacting that pixel of phosphor.)

I’m sorry, no. That is just moving the goal posts *way* too far.

A television CRT does not have a useful “memory effect”. If you scan a single frame of information onto a P4 phosphor display someone staring at it will see a flash but it’s going to decay away far too fast to usefully interpret. And sure, you can turn the argument to long persistence phosphors, which *are* designed to leverage a “memory effect” by lasting seconds for use by devices like radar, but they still need to be refreshed… *with brand new information”. This is why your argument comparing a CRT to a DRAM that you’re not refreshing is completely specious: the DRAM is designed to hold its contents indefinitely as long as you refresh it, and that act of refreshing does not require any outside knowledge, IE, to "refresh" it you're not redrawing it again with information stored elsewhere. (And this childish "well, what if I don't refresh it, it's still a memory thing" is dumb. No, it's not a memory if you're not refreshing it. Or at least not behaving as such. The most you can say is it is at least *capable* of doing the thing you're not doing, unlike a bare CRT tube.) Needing to be periodically redrawn with new information not stored in the tube *is* the case with every kind of CRT tube except for specialized devices like those bistable memory tubes in the 4010 terminals or something like a Williams tube with an external charge sensor. QED: you have to add tech well beyond a basic CRT for it to hold memory indefinitely, and the definition I was using for memory was *that*. And that's why I made the light bulb analogy; by your initial definition any operation that has an aftereffect, no matter how fleeting, could be called a "memory".

Again, those Plasma displays that were referenced are completely bistable, and once a pattern is entered into them they will hold it indefinitely until disturbed. Even the storage CRTs in the Tektronix terminals with their flood beams can’t say that, the image slowly degrades over time, but sure, we’ll let that slide. So…

I use a slightly wider definition of anything that can store information to be read later, regardless of whether it's the same device reading it. It might be just a human, which for me makes a Tek 4010 a memory device

This is exactly the definition I was using for the plasma displays. I’ll award partial points for long persistence radar tubes, but no, normal CRTs are not “memory devices”. Maybe you can think of the long persistence ones as "FIFOs" for converting bits of information presented in one form (instantaneous blips on a wire) into a "cached" form that our eyes can consume at our lower effective data rates, but they're not "memories" in the indefinite sense.

 

[cjs]

Now if I take my copy of Image Tubes by Illes P Csorba off the shelf, page 180 lists most of the contemporary phosphors used in different tubes. P4 is listed as 20us at 460nm (Blue peak) and 60us at 560nm. (Green peak) for decay to 10% brightness - So persistence of vision is definitely stopping this from making the P4 look yellow, since the blue receptors are finished pretty quickly.

Seems the reference I was using is completely smoking crack. Or at the very least, the guy is using "ms" to mean μs, and not milliseconds. Yeah, page 12 of this HP document says that P4 goes down to 10% in 60 μs.

P12 would be better for memory if you're slow scanning an entire screen at 50hz... it lasts for 200 milliseconds.

That sounds rather too slow to me: at 50 Hz a screen refreshes every 20 ms., so it's going to take quite a few scans to turn a '1' back into a '0'.

 

[Eudimorphodon]

FWIW, here's a "CRT" that has nearly indefinite passive memory: Skiatron. These devices used a "scotophor", not a phosphor, that turned dark when hit by the electron beam. That dark mark would stay there indefinitely until the face of the tube was heated above a critical point, at which point it'd go back to light. These were also used in radar displays (*very* slow scan), and were also the first form of trace storage oscilloscopes.

These tubes even store their traces with the power off, so it'd be fair to lump them into the same category as devices like digital paper.

 

[cj7hawk]

Seems the reference I was using is completely smoking crack. Or at the very least, the guy is using "ms" to mean μs, and not milliseconds. Yeah, page 12 of this HP document says that P4 goes down to 10% in 60 μs.

Well, it's not incorrect, since he used weird terms like "Not Over 7%" and sub-1% is definitely lower than 7% and you want a fast phosphor so you don't get some kind of trail when objects are moving quickly - Interestingly theres a lot of information collected there. I wouldn't go just throwing it away, but it is somewhat misleading. I sometimes do calculations about tube phosphor, and Illes Csorba is the "king" of that kind of data in my field so it did sound a bit long the way you phrased it. P4 wasn't one I was familiar with, so I looked it up. His was the first book I bought on the topic about 20 years ago when I started playing with image tubes.

If the data we rely upon is bad, then that doesn't help us much... Worse, because we reach other conclusions from that information, it can mess up our other assertions as well.

I wonder if we oscillated the tube up and down at a fast rate if our eyes would see the rolling shutter effect They probably would !

That sounds rather too slow to me: at 50 Hz a screen refreshes every 20 ms., so it's going to take quite a few scans to turn a '1' back into a '0'.

I don't know how they worked at a detailed level, but we're talking about a phase locked scanning tube picking it up and repeating it through an amplifier aren't we? So there needs to be enough intensity to ensure that it continues to propagate and 200ms is the time taken for the intensity of P-12 to drop by 90%... So if you have an amplifier of sufficient pickup sensitivity and subsequent amplifiers, then the question would be what is the threshold at which the level change transitions? I don't know if this is a large value (eg, 50%) or a low-value (eg 1%) - and I have no idea what the noise floor is for that kind of memory either.

But I'm guessing a reasonably long-lasting phosphor would have been the objective so that it didn't suffer from noise. What they used? I have no idea.

 

[Eudimorphodon]

But I'm guessing a reasonably long-lasting phosphor would have been the objective so that it didn't suffer from noise. What they used? I have no idea.

Wait, who is using this? Is this back to Williams Tubes, or did this suddenly spin off into applications like Slow Scan Television? In the former case the characteristic of the phosphor used isn't really going to be its "persistence", in the optical sense, it's going to be its secondary emission characteristics, since it's a static charge, not luminescence, that the sensor on the tube is meant to pick up...

Oh. You're calling "50hz" a "slow scan" television. No. "Slow Scan" television is measured in seconds per frame, not frames per second. In the era before frame buffers on the receiving end then, yes, you needed a long persistence tube for this. But by "long" we mean *really* long, IE, those radar tubes with persistence times in the single, or even double, digits of full seconds.

 

[cj7hawk]

Wait, who is using this? Is this back to Williams Tubes

Yes, data storage tubes - I don't have a lot of knowledge of how the different ideas worked, how they refreshed them or how they read the data out. I'd have to read up more on them to do that - I only have a rudimentary idea of that kind of memory that came from conversations rather than actual study. Don't read too much into my guesses there.

 

[Eudimorphodon]

Yes, data storage tubes - I don't have a lot of knowledge of how the different ideas worked, how they refreshed them or how they read the data out. I'd have to read up more on them to do that - I only have a rudimentary idea of that kind of memory that came from conversations rather than actual study. Don't read too much into my guesses there.

A: William Tubes: Have you ever run your hand over a CRT screen and noticed how static electricity tends to build up on them as they run? That's because of the buildup of charges from the the electron beam scanning behind the glass. The Wikipedia page doesn't do a terrible job explaining how adding a metal plate and a sensor to the front of the tube made it possible to sense if particular spot on the screen had scanned been "recently", IE, within the last fraction of a second, by writing that spot *again* and observing charge; an area that still had a decaying charge from the previous activation would behave differently from a "blank" area. Because the act of reading the tube erases it the result is a "write after read" memory (magnetic core has that same property), and because the charge decays the whole thing needs to be constantly scanning through and refreshing the individual bits in-between the computer accessing them.

FWIW, They made a prototype of a system like this that used optical photosensors to measure the decaying glow of a phosphor instead. The designers of this system were trying to address a serious weakness in the Williams Tube, that being it was not reliable at all because the little static charges it was trying to measure were easily disrupted by electrical noise nearby. Work on this system was abandoned when magnetic core came along.

B: The Direct-View Bistable Storage Tube, as used in Tektronix terminals and storage oscilloscopes, is *not* very well explained by the Wikipedia page. Start with a "conventional" CRT tube with a fairly long persistence phosphor (like you'd find in an oscilloscope) and embed a grid of electrodes (fine enough to see through) under the face. The phosphor you coat the face, in addition to being long persistence, has the property of having some degree of hysteresis, IE, it has a "threshold" energy that switches it to glowing brightly from glowing dimly; this is necessary because the CRT has, in addition to the normal "write" beam that's steered like a vector display, a secondary "flood beam" that constantly lights the whole tube with a low level of electrons.

When the write beam activates the phosphor to glow brightly secondary emission from the phosphor essentially creates a "circuit" with those embedded electrodes in the glass, and this makes the lit phosphor "more attractive" to the electrons from the flood beam, which allows the tube to be scanned *once* by the vector hardware in the terminal and store/display the image semi-indefinitely for as long as the flood beam is maintained. I say "semi-indefinitely", because over time the lit areas will tend to spread/blur out. Because this system should store the image for as long as the flood beam is turned on this tube isn't a "dynamic" memory like the Williams tube, it's static, but it's also inherently lossy.

As I mentioned earlier by pointing out that Tektronix made a printer that could make a hardcopy of the contents of the screen, the contents of these tubes can be read out through the sustainer grid using a method similar to that used by the William's tubes, IE, you can scan the write beam at a low power through the screen and, depending on whether the area being scanned is already lit or not, you'll pick up a detectable difference in the charge. The fidelity wasn't great but it was good enough to be useful at the time.