EEVblog

Another mystery teardown.
Forum: www.eevblog.com...
SPOILER:
A Philips PM5639 Colour Analyser Teardown
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
Support the EEVblog through Patreon!
/ eevblog
EEVblog Amazon Store (Dave gets a cut):
astore.amazon.c...
T-Shirts: teespring.com/s...
💗 Likecoin - Coins for Likes: likecoin.pro/@...

[@hardwareful]

Come on Dave, put some more effort into it. Dielectric filters look like the complementary color in reflection and when you do test them, use an incandescent lamp.That diffusor plate and tinted glass filter will also have to do bandpass filtering (plus there's fungus growth on them). That's because the dielectric filters become ineffective towards IR and UV and Si has peak quantum efficiency in the NIR. You made spectroscopy look like a silly thing with colored glass discs. I thought that with the advent of DLP projectors and their dichroic color wheels at least the general appearance and spectral properties of dielectric filters very similar to these presented here should be known. Calculating the dielectric multi-layer structure is a different thing altogether. I come here to see specialized equipment teardowns and these inevitably require some degree of understanding of the underlying working principles. Otherwise it's just breaking stuff.

[@leisergeist]

To be fair, it's not even near his field at all, and really not common knowledge The mold growth was interesting, someone else said that's common with these kinds of lenses? Weird!

[@EEVblog]

I already put enough effort into it with what I had to hand to the show an effect. I was not going to delay the video by another two days (yes, two days) just to make it look a bit more professional. I can always do another video if there is enough demand. I'm extremely busy at the moment and time is very limited.

[@stevelupton2533]

hboy007 - if you're going to be critical, you might want to be accurate yourself. It's dichroic, not dielectric!

[@hardwareful]

read en.wikipedia.org/wiki/Dielectric_mirror

[@stevelupton2533]

[hangs head in shame]... sorry!

[@djrbaker1]

5 AM here in Brazil. I was just getting ready to go to bed but then this showed up. Damn you, Dave Jones.

[@EEVblog]

You're welcome.

[@djrbaker1]

EEVblog do you plan on following up on the batteriser debunking videos?

[@EEVblog]

If I can muster the enthusiasm

[@aaronbrandenburg2441]

@@djrbaker1 he should do more Niche application videos they probably be pretty interesting if he can find things more teardowns of equipment like that as well Niche stuff can be very interesting at times and yes the batteriser I'm actually surprised Google actually got that right anyways but for some things as he said there could be an application also is that could actually be a form of energy harvesting that in a way sort of kind of maybe or for example you know how that you're able to use a partially dead battery that won't power something else is it doesn't have enough juice available but would still operates a a remote control four months stick what on that battery and even longer use that might be a way that things like remote controls couldn't even run even longer there would be some applications if you have a power source. Would work like that battery dead That was supposed to improve and combine that circuit with that it's possible you could make an energy source that's not quite as useful War useful possibly an idea like a jewel thief or something like that same idea I've seen some old car jump starter things that use D cell batteries years and years ago and they claim to use power extraction to extract to get enough power out of the cells when actuality it was a dc-to-dc converter going from a higher voltage to a lower voltage but increase current really wasn't expecting anything it was just converting are horrible tudge to a lower voltage at a higher current dc-to-dc converter in other words if the circuitry from a batter Riser could be used from a solar cell where did Big reminder for recharge will battery or maybe even a super cap even things that seem as ridiculous as that device that I'm talking about do have their merits somewhere a lot of times but not always but he did mention it may have its applications no matter how niche it is there is at least one? But yes I could think of a few. But Niche products and devices are ugly fascinating and sometimes pretty spectacular on teardowns I've seen on other channels Mike's electric stuff and others as well including a v e Yallakora cev blog but that's being redundant since I'm watching one of his videos right now LOL

[@AllanSavolainen]

I don't think the filter is RGB but CMY

[@jerzyk007]

afaik this is why they have separate model for the LCD

[@ABaumstumpf]

The colors really do match with CMY.

[@jipihorn]

No, these filters mimic the x,y,z curves of the CIE observer (probably the 1931 2°). All colorimetric devices uses CIE coordinates (although there is a CIE r,g,b that is obsolutely different from RGB). RGB (or CMY(K))are device coordinates and are useless in physical measurements. This is why the "red" filter has two colors on it. This is the x curve.

[@ABaumstumpf]

www.fujifilmusa.com/products/semiconductor_materials/image-sensor-color-mosaic/cmy/img/index/pic_02.png That is a CMY filter and the curves look pretty much like what we saw coming through with his spectrum analyzer. I think getting a CMY filter and processing the data with the CIE standard is what they are doing.

[@laurikangas7809]

Interference RGB filters look like CMY because they transmit R, G or B and reflect whatever remains, which would then be C, M and Y, respectively. If you block light from your direction and look through the filter, you see RGB, just like Dave did there.

[@olivialambert4124]

The red filter really isn't a problem. It lets through the red light plus the full spectrum of the blue filter. So it simply needs to subtract the value from the blue filter and you're left with the brightness of the red alone, job done.

[@billa8671]

you don't know shit about electronics. go back to the kitchen and make a sammich

[@glenwoofit]

I've been after one of these units for some time so it distressed me to watch it being taken apart. I used to use them to line up broadcast monitors and now I want one to line up my amateur television station...

[@zx8401ztv]

It might be old tech, but its lasted, unlike modern junk :-)

[@PatrickHansen101]

Compare the pricing of the old tech and modern junk :p if you don't cheap out, and buy proper stuff, it'll last.

[@joshsamuelson1793]

In 2000 CRT had way better resolution than LCD did at the same price. Its not unreasonable that test equipment for CRT screens would still be in use.

[@EwanMarshall]

One needs to calibrate LCDs too if they want that level of colour accuracy: www.xrite.com/categories/calibration-profiling/i1basic-pro-2

[@ChlorideCull]

Ewan Marshall Or, if you don't want to break the bank, a ColorMunki. Made by the same company and claim to be equipped with the same hardware and color engine technology. It's just aimed at hobbyists instead of big businesses.

[@EwanMarshall]

+Sebastian Johansson thanks, I'll look into it. Yeah, not my usual area, I just know the big player in that area. Wouldn't mind calibrating my monitors properly.

[@EwanMarshall]

+fragwits Eizo are very high end, making only IPS panels, and designing them more for people working at companies where colour is more important, like in mastering movies. Some panels have some basic calibration so they can do some auto-tuning (VGA had to sync up the horizontal and vertical refresh rate, and so some early LCD monitors had some calibration as part of that). In most cases built in calibration is not as good as proper calibration via colorimeter, even on higher end IPS panels, ones that are that good are tuned with a colorimeter at the factory and then a profile loaded onto them and cost a few thousand each.

[@tookitogo]

Ewan Marshall You don't have to spend that much to get that type of factory calibration: every Retina iMac has its display so calibrated. (You just happen to get a computer inside the display while you're at it.)

[@loughkb]

Geez, a lot of guys jumping right on CMY colors, probably thinking of print. They didn't bother to think for a minute about what a CRT has in it's face. Red, Green, and Blue phosphor. So of course the filters would pass RGB, not Cyan, Magenta, and Yellow.

[@GoldSrc_]

2:20AM here, I was about to go to bed and then Dave puts out a video... sleep can wait.

[@SquidOfOverkill]

Even a mute can type! (provided they have hands)

[@rarbiart]

Using an incandescent light source would have been a lot more suitable to test the filters.

[@TechGuyCharlie]

G'day Dave, our second TV was a 21 old Sony CRT TV which we replaced last year. The only time we had it open was to adjust the video position on the screen, other than that never had any issues. They don't make stuff as durable like they used to.

[@marko99butter]

I spent much of my career doing color science for cameras displays and LED lighting. The RGB filters are CIE tristimulus functions x-bar y-bar z-bar. They try to mimic the human color responses of the three cone types in the retina. They are broad band. The instrument is used to set the primary RGB points of 255 0 0 0 255 0 and 0 0 255 then you need to set the white point 255 255 255. You can also use it to get the linearity (gamma) of a ramp from 0 to 255. For those interested you can google CIE 1931 color space. Different industries use different color math but they are all derived from the work done back in 1931. Konica Minolta made many instruments similar to this one.

[@BlankBrain]

There is a community of people who restore vintage vacuum tube televisions (non-solid state), just as there are people who restore old radios and audio equipment. This test gear would be a good find for them!

[@UpLateGeek]

Yeah, my broadcast engineer friends still use monitor colour calibrators to calibrate the monitors on their video editing and graphics workstations. Datacolor Spider series is the one I've seen, and I gather the software is pretty well automatic, so it's pretty much a matter of a few clicks and it's calibrated. I guess colour calibration is more important when you've got footage shot at different locations, indoors and outdoors, on different cameras, and when you're working with graphics and visual effects. The colour grading all those things need to match, otherwise the flow of the scene will be jarring as they cut between camera angles that are different colour temperatures, brightness, contrast, etc. And it should match between scenes in the same location, time of day, etc. So colour calibration is probably a lot more important in the film and television industry than for most vloggers and other KZbinrs. They'd tend to only have the one camera and lighting rig, and use the same one or two monitors, so near enough is probably good enough to get their colour grading pretty close.

[@goyabee3200]

Mystery teardown. Tells us what it is in the first 30 seconds.

[@AssGoblin]

I've been sick with the flu and your videos have been keeping me going this week. Keep up the awesome work!

[@hakemon]

Wow, I use a modern version of these. It's just a USB based probe that I stick onto the tube of the TV (or panel if it's LCD, the probe works on both via different profiles). Do they work? Yes, actually, very well actually. Every one of my monitors on my computer go through these same colorimeters, and they literally all match, it's amazing that they all look the same.

[@kallitokaco]

@Dave nice video, a visual computing guy here. The blue component in the red filter is not important because you can substract blue from it(due to the very similar range as the blue and every signal in the blue filter will be in the red too.). Thus the primary color for red is: RF-BF. Probably the coating of the lens was cheaper in this way and its better to have a known error in the filter that you can eliminate than an unknown. Dont you have an old non led flashlight with filament inside? They are close to black body emitter. If you put an thin packing paper in front you have a nice diffusion too and you dont have to handle with the amplitude spikes. Greetings!

[@EEVblog]

If the sensor is a broadband one, then you'd have to substract out in software from the blue sensor. Pretty yucky stuff. Incandescent are very low in the blue. Direct sunlight is best for a wide flat spectrum.

[@kallitokaco]

you are right, its far from perfect hardware but has only a computing disadvantage without much loss in accuarcy/size of the gamut. And the size of the gamut is certainly the critical part compared to the computing time for calibration. They are low but close to constant light source and they emit photons of every energy level in the visible spekturm. This is a big benefit for measurements compared to the sun if you dont need absolute values and only the change caused by the filter("absorbtion"). It allows you to calibrate to your source spekturm and only show the differences caused by the filter. Futhermore there are no big spikes in the spekturm of filaments compared to led/phosphor-based lightsources and its easy to imagine that the error is linear increasing.

[@berni8k]

There is no red because your flashlight doesn't seam to have a lot of red content in the light in the first place.

[@mikeselectricstuff]

The filters look different from the rear as you're seeing the reflected colours, Cyan,Magenta & Yellow

[@firefly618]

Isn't the spectrum analyzer also showing CMY curves? At 14:38 we're looking through the C filter letting through B+G light; at 15:36 it's the Y filter, showing a G+R plot (but his light source has very little red); and at 16:17 I think that's the M filter showing R+B light, that's why there's a dip in the green region.

[@EEVblog]

Spectrum is identical either way, I tried it.

[@fsantini81]

Those are dichroic filters. Very narrow bandwidth. Nice stuff (not too expensive, probably $10 each or so). And they are indeed RGB. And the sensor could be a Hamamatsu S2386 photodiode.

[@feynthefallen]

I love teardowns. Not so much the chip numbers, I'm not THAT electophile. But the whole fooling around with the parts is just great.

[@Anamnesia]

I still have my old CRT vacuum TV. I must have gotten it right around the transition time between CRT & Plasma/LCD TV's. I purchased mine in early 2004 & even though it doesn't have any HDMI input, I've added an Android media player - works fine! 😎

[@erpalma]

FYI on the front glass there is mildew. You can find it on old camera lenses for example

[@T3sl4]

Those'll be dichroic filters: layers of glass spaced for interference colors. They can be very sharp, but can't do much more than an octave, for obvious reasons. Good enough in a visual-spectrum (only about an octave anyway) filter!

[@Shadowdncer]

And dichroic filters would be directional *hint hint*

[@EEVblog]

Wouldn't a dichroic filter show a different response in the other direction. I got the same response in both directions.

[@T3sl4]

No, not in/out direction (which can't be different, else you'd have an amplifier in there!), direction as in, a small tilt off the axis. You can kinda see this in the video already, because of just how weird they look -- the iridescent colors at different viewing angles. That's the other property, too: whatever light isn't transmitted, is reflected. (Just like an ordinary electronic filter!) Tinted glass, filter gels, pigments and such, are all absorptive type filters.

[@philcrawley]

I've used a 5639 most days for 25 years! CRTs are almost gone in pro-TV but they did a subsequent LCD head. It is a colourimeter (so only RGB tristimulus, not a spectralradiometer). In the BBC these were often referred to as "Elvis detectors" and they had the annoying habit of falling off the tube face when you were half-way through your calibration. Dave - nobody has perfect colour memory; it's not how human vision works!

[@Vokabre]

Studio CRT monitors are finding a second life now, they're the best things to connect to vintage computer and gaming tech. We use them at our travelling retro tech museum as much as we can. And even at home most enthusiasts prefer to have a CRT TV around to connect to some vintage tech as well.

[@zorktxandnand3774]

Yes I know these, I used to use these a lot when I worked as a technician in broadcast, in 3 different company's so it is sort of the default calibration tool for broadcast CRT's (at least here in the Netherlands, but Philips, so yea..) It is common practice to regularly calibrate monitors in a broadcast environment, especially reference monitors, witch are used in any editing suit, OB truck or studio control room ect. to determine how the actual picture looks, unlike at home on a crappy TV with all settings al over the place :). Also every monitor was calibrated after a repair, even the 6 and 9" viewing monitors. On a broadcast reference monitor all the settings (brightness, contrast, saturation ect.) are normally in the preset position, so you know what you look at. If you can ever can get your hands on a broadcast CRT monitor, like a Sony BVM or a Barco CVM It would make for some very nice tear down material (as would most broadcast equipment up to +/- 2005 ). Unfortunately I cant help you with that as I now moved on to a different industry.

[@SilverSpoon_]

I used one of these to calibrate my HP LP2465w and 2475w monitors! it's really a professional tool.

[@km5405]

9:57 here, good way to start my Saturday morning :P .... also nice to see a Philips made quality item

[@arsenixa]

ah Philips, reduced to a label on crappy chinese USB sticks, TVs and light bulbs. Tricking people into believing it's the company they knew in the past. How the mighty have fallen

[@makomk]

I'm pretty sure that Philips still design and make their own LED light bulbs, possibly even with their own LEDs inside. Other items possibly not so much.

[@km5405]

Don't forget the shaving irons :P also don't they still make stuff like diodes for RF too? we had a scrapped one in the school lab a while ago.

[@mjouwbuis]

Whats left of their diode line-up is made by NXP.

[@markbell9742]

Got rid of my crt tv in about 2002, it was a monster felt like 150 lbs. When I tried to lift it the plastic case crumbled into a million shards. Cheers, Mark

[@phpn99]

You need to be aware that the spectra multiply: The spectrum of your light source is correlated to the spectrum of the filter. High-end measurement devices typically will use smooth-spectrum tungsten sources, so as to de-correlate the spectra accurately. Also; the shape of the filter bands are typically tuned to the photometric response of the human visual system that is implicit in the measurement metric used, such as CIELUV. There is a fundamental difference between photometric and radiometric measurements.

[@Forssa1]

The large teal window is a common IR filter for all senors. LED is not good for this sort of test. Better to use incandescent/halogen for a much wider and continuous spectrum.

[@EEVblog]

incandescent is actually pretty bad down in the blue.

[@Forssa1]

It gets more even if you drive it harder than normal

[@itsGeorgeAgain]

Why didn't you place the analyzer to the place that the sensors would be, and shine the light from the front through the diffuser and the glass? Like it's supposed to be used.

[@Manderby]

The three filter spectra do not need to be narrow-band at all and it does not matter if there is any blue in the red filter. The important thing is that these filters are long-term stable and they are distinctly different from each other. In radiometry, you multiply the three measured spectra with the so called standard observer curves (industry standardardized by CIE in 1931. For everyday use take the 2 degree observer) and take the integral of the result which results in three values which are kind'a but not exactly like normal RGB values which you know from computer screens with the importand distinctions that they are linearlily rotated and shifted in three dimensions to accomodate every possible perceyvable color by the human eye in the positive definite octant and are not linearized (no gamma correction). These values are now (again) rotated and shifted using a transformation matrix (definedby the calibrated known filter characteristics) resulting in XYZ values which you then normalize by dividing by (X+Y+Z) resulting in the Yxy chromaticity diagram (image search for it, you probably have seen it before). In colorimetry (in contrast to radiometry), you look at these values relative to a specific white reference, for example a D50 or D65 daylight reference, (which you deduct very much the same way using three filtered spectra) resulting in densitometric XYZ values (roughly) between 0 and 1. This is called color density. Note that there are different mathematical routes you can go. You can premultiply the spectra with the filter spectra or just rotate the values with a 3x3 matrix defined by the filter calibraction. If this would be a true radiometer, then no filter would be required and only one sensor which measures the whole spectrum. Or you can go (as with this device) the colorimetric way and just compute and compare the XYZ values. This device uses three filters, hence it is a densitometer. Probably, it is a simple 3x3 matrix computation. The computation of the daylight temperature... Ah, that is just some quirky formula you can whack the x and y values in and bingo, you got some Kelvin coming out. Not too physical, but industry-standardized, and that's the point of it. All other color theory including Lab, LUV Yuv, RGB, HSV, HSL, LCH, ... are deducted from that. If you like to see all that in action, have a look at my application Color Machine (only available on the Mac) or go see the website of brucelindbloom.

[@ayratg5640]

Those are interferience filters, the same you can find in DLP projector wheel. Very narrowband filters that reflect everything except one color, so thats why they look like yellow cyan and magenta.

[@materialsguy2002]

"In spec, ship it." LOL. Thanks Dave.

[@martinda7446]

Someone who may like this, Shango066, who rescues old CRT TVs. I used to dabble with TV and much of this set up was done by eye, a pattern generator and colour bar generator was about all you used - and to a high enough standard for consumer sets, and often much better than it started out. I remember looking in TV showrooms and seeing how bad all the sets were set up out of the factory - All different colours, choose one to go with the curtains! Illuminant 'D' (6500) was the BBC recommended colour temperature 'board' for monitor set up. Much of the real work for television standards was done by the BBC, and the like. Same for audio. The technical department was a world leader in the early days. Philips was of course a big supplier of colour TV, and test gear. They owned Mullard in the UK, who made wonderful CRTs, which I had one preserved until 1995 ish in a DECCA? Anyone who has better memory than me, sorry if I say anything silly.

[@flex209]

The Datacolor Spyder5 and the X-Rite i1 Display Pro are the most common display calibration solutions nowadays. And the weird blue filter that filters all of the light going into the sensor is still a thing!

[@uwezimmermann5427]

the blue plate is an IR filter - same that you find in digital cameras in front of the CMOS/CCD sensor. Otherwise your silicon photodiodes will pick up near-IR wavelengths which are transmitted by the interference filters. You donät see any on your spectrometer because your LED flashlight does not emit any...

[@3fus]

2:30 Maurits van Tol was working for Philips in the Nat.Lab. Eindhoven NL, after that he started his own consultancy firm, he's still active according to linkedin.

[@nictamer]

The red filter lets far blue (violet) in because that's how the eye works; that's how you can see violet: it stimulates both blue and red sensors, while near blue only stimulates the blue sensor.

[@IC225]

i still have a CRT telly in my room thats still in use for CCTV, also have one downstairs hooked upto a digibox. only LCD units i have is a 7inch handheld/portable telly, laptop & an external monitor

[@TheMightyKinkle]

Oh man. There's bacteria growing on front lens. It's acting like a Petri dish

[@georgegonzalez2442]

Your typical silicon sensor has a response curve that is like 10 times more sensitive in the red than in the blue, so you need to slope all your curves strongly to the left. For instance the blue peak on the red filter is only a tenth as high once it's been sensed by silicon.

[@JouMxyzptlk]

Last time I used a CRT? Last week. In our "install PC departement" we still have some CRT monitors. Since we are a computer company those things still exist until they die.

[@crocellian2972]

This sort of equipment is in daily use in photo editing shops (prepress guys). In fact, if you don't use something like calibrated test exposure cards then use those data to test the monitor, you will get stuck with "as is" clauses in your print contracts. Critical tech - although we use the master exposure cards and the correct detectors for our displays. Millions of dollars have vaporized for want of calibration.

[@rtesimpson]

yeah...I'm playing along at home...

[@FensterstockHias]

After watching this the first time i keep hearing some sort of inner voice every once in while shouting FOOTLAMBERT!!! And then i have to go and rewatch this. Just as i have done right now. Cheers!

[@tmdrake]

Love these teardowns!

[@jrausa1]

Dave, you were wondering why the red filter had blue in it. Someone mentioned it already, but just to emphasize, the "red" sensor in the human eye passes some blue, and these filters are made to mimic the human eye. That's why they don't look like conventional RGB colors.

[@ManWithBeard1990]

It looks to me like the filters are CMY, not RGB. Another clue that this may be the case is that the plastic mold has four holes, not three, so possibly there is a higher end model that measures full CMYK. I can't really think of a reason why, though.

[@LoneWolfZ]

you had much less red to start out with. It makes sense that it would measure low through the filter on the spectrometer

[@SproutyPottedPlant]

"this LED isn't the best' but...but...but it's a Nichia 219b the Lumintop tool is a high CRI torch, one of my favourites !

[@DavidTelesPortugal]

I still have a working 30 inch crt tv at my kitchen. Since in Portugal every fiber connection contains a UHF modulator for analog tv signals I don't see why I should replace it, if it is stil working. Just remember I have another one but this time a 32 inch on the outside lounge. But I also have LCD's at home

[@williamsquires3070]

Hey Dave, note the block diagram for the SED1330F has two diodes connected back-to-front near the character generator ROM block, and the dot/refresh counter block. Weird - why would then need to connect a diode between the two internal busses? Or maybe it's a red/green bi-color LED on the die for testing! :)

[@IvanKowalenko]

It's a colorimeter! We have that exact same model in the office for calibrating displays! That model is built for use on CRTs, and has, unfortunately, low sensitivity on LCDs, but you can coax it to work, but you really have to blast the backlight. Last used the thing about a year or so ago on some TVLogics. I'm not an expert in colorimetry, but I can tell you that's quite a sophisticated tool. Cool to know what's inside the thing. I'm surprised that has RGB sensors, since video signals exist within the YCbCr domain. That may explain why the filtering is so wonky.

[@JoelHudson]

I thought it was a color analyzer, BTW they work with LCDs too. They use them for High quality TVs, and Monitors used for displaying Photography

[@josephtorelli]

I work with flight sims and we used to have four 25" crt's pointing down into a beam splitter the going into a mirror facing the cockpit. We had to color match the crt's so the view thru the window looked the same from one channel to the next. i could aways tell who adjusted the system last when I walked in. Some would have a green, or yellow "tint" to the scene depending on who adjusted them. I was the blueish guy. Back in the early 90's we had a rep who was trying to sell us one very similar this one. He was demonstrating it to us while the display was on a cart. It worked ok. Every body was excited. Until I told them that once you rotate the display 90deg to face downwards it would lose it's purity. And it would hav to be degaussed and recolor balenced. Plus the beam splitter added a little green to the image. The sensor would not fit in the area once the crt was mounted. So I popped the salesmans bubble. He wasn't happy with me. Such is life!

[@17hmr243]

mold growth?

[@TheAmmoniacal]

yep, it's a typical problem on optical lenses www.4photos.de/camera-diy/Lens-Fungus.html

[@CornishMiner]

To me, it looked like a lens/filter made from a salt. Common in spectrometry.

[@Shadowdncer]

My thought as well. A good reason to throw that stuff out and swab everything that has been around it with alcohol.

[@Vokabre]

Best offer! Free Petri dish included with all PM 5639 colour analysers! (/irony)

[@tohopes]

Someone used it to measure the color of their leftover milk after eating a bowl of Fruity Pebbles.

[@Petertronic]

I worked repairing monitors in the CRT era and always wanted one of these in the workshop, but they were stupidly expensive.

[@mysticalrockhoundingfossil5464]

Love your vids, Notifications work here . Those are nice sensors , PIN foto diodes i think , Same kind of sensor as what was used in the rad.. detector kit you build in one of your mailbags. these ones are metal canned versions of that . (I hope Dave will try to make the odd display work from last christmas teardown , it was beautiful.)

[@Psychlist1972]

These days, there are little pucks you attach to your display, and then PC software for doing the calibration/reading. CalMan is an example.

[@TheDefpom]

No notifications at all, saw your twitter post !

[@EEVblog]

The work experience kids at KZbin again...

[@FennecTECH]

I got it it popped up in my youtube bell

[@toysareforboys1]

Unsubscribed me for the second time :(

[@voltlog]

I got my push notifications Desktop (Chrome) and mobile ios youtube app.

[@wotid]

What time did you post this video? My email notification came in at 19:09 AEDT

[@Boffin55]

Dave, these were pretty common in the Graphics arts industries too (to match your screen to what you end up printing....)

[@nathantron]

That would make a good raygun prop to dress up. XD

[@AntiProtonBoy]

These are dichroic glass filters, more specificaly, Bayer filters. They model the tristimulus response curve of the human eye, which explains why there is a blue peak when measuring light through the red filter. The photoreceptor cells in the retina that is responsible for red perception is also somewhat sensitive in the blue and violet parts of the spectrum. Hence the reason why we can perceive violet.

[@george8bitsworth]

About why there would be some blue light coming through the "red" filter, there is one logical explanation. It isn't a red filter it is a magenta filter. Magenta is the mixture of red and blue light. In that RGB-CMY "discussion" below one comment said that the "only" time you would use CMY filters is in printing (or general color photography). Why subtractive colors would be used in this equipment is a puzzle. It looks that is what they did, though. Maybe someone here knows why. If there is a comment about it, I missed it.

[@LasVegasVocalist]

The Duck's Guts??? LOL

[@Nicktrance1]

Yeah, you can get some decent color analyzers for LCD monitors in the $100-$200 range.

[@WafflesASAP]

21:25 -- "...that's not the microchip symbol." Nope, that's the Tri-Force.

[@SidneyCritic]

My 1979 Rank Arena CRT broke down in 2005, and that's when I switched to my first LCD(using it to watch this now). It would be good if you can get a better light source for that analyser, because that was interesting.

[@kokodin5895]

filters are crazy, they look like cyan , magenta and yellow on one shot and rgb on another :P

[@floriandaler5327]

kokodin Those are, as I think, dichromatic mirrors. They let one Color trough and reflect the Rest of the spectrum. So you can See 2 diffrent colors, depending where you are looking from.

[@TheCaitlinlopez]

No they still used, forgot publish house, but they are normally smaller and in very expensive monitors

[@TheCaitlinlopez]

I need more coffee before put a comment

[@feieralarm]

I wonder how accurate this is compared to a modern colorimeter you can already get for about $100.

[@ColinJonesPonder]

My Illudium Pu-36 Explosive Space Modulator! That creature has stolen my Space Modulator!

[@Designandrew]

I think those photosensors are totally custom probably specific to this unit.. the fact they are hot snot attached is a bit of a give away

[@jackwhite3820]

If they were totally custom, they would have found a better solution than to use a standard photo diode package and hot snot it down. Also what would be the reason for using a custom photo sensor, since all the magic is happening ion the filters.

[@Designandrew]

Yes good points, I meant to really say that the stands the custom, like the angle and position and hence glued to the board, but the actual component was not custom

[@ThePetaaaaa]

That sensor reminds me a lot of a photodiode I used to work with: OPT301 (Integrated Photodiode and Amplifier) Brass casing, big window and a large piece of semiconductor.

[@mitchese1]

This seems to be just something which detects the colour of light appearing when you stick it to a surface, any idea why they would need a different model for an LCD (and why the LCD looks like a big pencil rather than a flat disc)?

[@USWaterRockets]

They probably used a software filter to trim off the blue from the red sensor output and only look at the red peak.

[@whitcwa]

An incandescent lamp would have made a better light source. It has continuous spectrum. An LED light was the worst choice. The filters are RGB in transmission. They appear CMY in reflection.

[@einars899]

The sensors may be BPW21. They have a specral range similar to the human eye. And the sensor chip looks quite like them.

[@pileggitech]

3:16 am in New York. Still up here.

[@AnthonyShuker]

Pileggi Technical Services of course you're still awake. How would you have commented?

[@pileggitech]

LOL

[@MicobyteMichael]

Hi Dave- I was just wondering, What do you do with the things you recieve to tear down after you tear them down? Do you throw them away, keep them stockpiled somewhere just in case, or what?

[@RobertShaverOfAustin]

Do you know of an inexpensive spectrometer that I could use to test the CRI of video lighting equipment?

[@shelydued]

HAHA, my family used a crt television up until 2016. we finally replaced it when the tube lost it's focus and took about 3 hours before the tube had a somewhat presentable picture, but you still couldn't read a tickeer tape on the news or any text.

[@Razor2048]

With how simple those colorimeters are, I wonder why these companies like to price gouge so much for the. Even for modern ones, it is not uncommon to see them go well above $150, with the spectrometers getting significantly more expensive. These devices should really be in the sub $20 range.

[@mbirth]

10:05 … I don't think it's a trigger. Looks more like a suction release valve for easier removal after calibration.

[@rarbiart]

those flat cables directly soldered on the front LCD PCB is strange. Why did they go into so many different ways of interconnecting the piggyback? routing problems? inherited design of LCD front from another project? BTW: the power regulator with this massive inductor looks like mid 1980ies, plus those ROMs alongside with discrete TTL-Latches(?)

[@robbieaussievic]

..... The close up of the three lenses reminded me of the Martians, 'The war of the worlds (1953)'. The contaminated round lens was from Errol Flynn's doctor. (A result of Errol's promiscuity ?).

[@budude2]

Hey Dave - I must have missed your mini light analyzer you are using from earlier videos - what model/type is it? I could use that myself!

[@Mr.Unacceptable]

I'd love those sensors and filters for my light table.

[@fabimre]

at 16:20---- your detector is off. At all colours it show the same extra blue!

[@lachee3055]

Hey Dave. I was wondering if you where going to do any more videos on your Nixie clock. I was enjoying the design process and how you are going your own way instead of getting a kit.

[@Danandrea737]

You should definitely try the spectometer with the original led to see a realistic response! :) The logo on that TC500A looks like Olidata symbol! (or the Triforce from Zelda games!) :) No idea if they actually produce components, though

[@michellesperling2726]

Hey Dave, just applied for you job! I swear Im not a stalker lol, just doing my research and hopefully getting a jump on other applicants! Interesting

[@WolfmanDude]

Wow that thing is cool! I actually have a use for something like that since I restore old crt TVs!

[@willproctor7301]

Nice toys Dave, would the spectrometer be sensitive enough to gather a spectrum from star light through a telescope for example?

[@yomamsie4438]

Will Proctor the detector would most likely need to be cooled to allow sensing of the extremely low light levels, like liquid co2 pumped through a heatsink or a colder fluid. Maybe a peltier

[@luisdanielmesa]

do you think the green and red glass let in a little of the blue as well and they subtract it later taking the measurement from the pure blue?

[@ZomB1986]

Still have my CRT TV and CRT computer screen. CRT computer screen has visibly more brilliant bolors than the TFT next to it. Especially the red on the TFT looks almost orange compared to the CRT. Bought two of them for $10 each so I have a spare. They come from Universität Koln, so say hi when you went there and remember those CRTs. (I bought them in NL btw.) The TV will go when analog broadcast goes, and that shouldn't be too long I expect. (Yes, I have digital HD tv too but not in my bedroom since the provider's package only comes with one decoder box included :( And unlike in Japan, the TVs don't come with built-in digital decoders. )

[@JasperJanssen]

I reckon if you hooked that suction cup onto an LCD and gave it a few good squeezes you'd almost certainly find it was way out of spec -- permanently :)