Been for another trip to the lock-up yesterday, and one of the things I came back with was this! I'd forgotten about it, I picked up a box I didn't recognise and there it was sat in the bottom of it. Thought I'd bring it back to the workshop, it looks useful, but probably isn't, unless you happen to set up professional colour monitors for a living!
It looks like it's for setting up the white balance, you stick the probe on the tube face (then it falls off...) and the amount of red green and blue is displayed on the LED displays on the unit, It might be useful for setting up old colour TV's after repair, so that they don't all end up different when switched on together.
When I tried it out, all I got on the front was 'E01' on the top display, which quickly went off, along with the 3 green arrows dimming. I'm sure it used to do more than that before it was put away!
So off came the covers, and I spied a memory battery, NiCad 3.6V, with a nasty looking crustyness around it on the PCB. I thought this must be the culprit, so removed it and gave the board a good clean, and re-soldered it's connections where it had eaten the copper away. I put it back together and fired up only to see it do the same thing! I had a good look at other solder joints which looked dodgy, but that made no difference either. The power supply section has the output voltages marked on the PCB, which is handy, so I gave them a check. There needs to be +8V, -8V, and +5V. The first 2 were fine, but the 5V would come up for half a second on power up, then drop to 2V. I checked all the caps on the tester, and they all passed with flying colours. At this point, I decided to try injecting 5V into the thing with a current limit set on the bench PSU. I found that it needed about 650mA, and the thing came on and stayed on, I could even run the self-calibration and it would happily work.
I turned attention to the 5V regulator, a big one on a heatsink, LM340KC-5.0, a 3 terminal linear type, very much like the 7805 types. With it out of circuit and rigged up with a meter and bench PSU it was fine, up to 20V input, and a nice steady 5V out... But put a decent load on it (I used an EF91's heater to load it!) and it was fine with an input voltage up to 8.5V, then it would keel over and drop to 2V. In the unit the voltage fed to the regulator is about 10.5V, the datasheet I found online for the regulator said it should work with an input up to 35V, so something was wrong with it. All I have available currently are some 7805 regulators, so I've tacked one of those in, and it works. Probably runs a bit hotter than the bigger one, but it should be able to cope with it. At least until I decide if it's worth spending money on the correct replacement!
Next thing I notice is that running it in the 'mode analyser' setting just displays 0 in all the displays, no matter what the probe is looking at, it does respond to light, as after running the calibration the 0's flash when the probe is in the dark, and if you shine a really bright light into the probe the displays all go to '---'. In the 'chroma' mode the displays show numbers, which change in response to different coloured light, which tells me it's sort of working, but the bar graph meters to the side don't change at all.
Powering up with the probe disconnected makes it go into what I think is some sort of test mode, where the display alternates between everything lit up and a bunch of codes on the number displays. All the ones I've looked at on Google and eBay show different codes to mine, no idea if that is significant or not though!
So, is this likely to be a useful 'thing' or just a fancy gadget that has no real use? Anyone here used such a thing before?
Here's some shots of the chips on the main board.
Regards,
Lloyd.
Hi Lloyd,
I have a similar device made by Barco. Their main use was setting up video walls to get the colour consistent when video walls consisted of large numbers of CRT/rear projection monitors closely stacked together. Not a job you'd take on by eye!
Ah, yes! I'd forgotten about video walls! Something you don't see anymore. My main thought for this was to set up those 2 KV-1330UB's that I fired up the other day, the colours are totally different between the 2 sets. I did try to set them up by eye, but even though I thought I'd got them both very close, when I looked away for a bit then looked back, they both looked completely different again!
Regards,
Lloyd
A problem I remember well: matching four colour monitors in a lighting gallery perfectly to each other. Not a job undertaken lightly.
Lloyd said
It looks like it's for setting up the white balance, you stick the probe on the tube face (then it falls off...) and the amount of red green and blue is displayed on the LED displays on the unit, It might be useful for setting up old colour TV's after repair, so that they don't all end up different when switched on together.
Yes, that's the idea.
So off came the covers, and I spied a memory battery, NiCad 3.6V, with a nasty looking crustyness around it on the PCB.
Probably powers memory for data and options, which you'll lose if the battery is dead or made open circuit.
There needs to be +8V, -8V, and +5V.
I turned attention to the 5V regulator, a big one on a heatsink, LM340KC-5.0, a 3 terminal linear type, very much like the 7805 types.
When I worked for National Semi we sold both, but didn't tell customers they had the same die inside and we're tested to the same specs, the LM340 sold for slightly more.
In the unit the voltage fed to the regulator is about 10.5V, the datasheet I found online for the regulator said it should work with an input up to 35V, so something was wrong with it.
Not necessarily, while the input can safely be 35V the device only needs a minimum delta between input and output to maintain regulation. For this type you would only need about 3V more on the input to get out 5V.
However, you also need to consider worse case contions, namely low AC input, high output current, and any ripple on the input due to weak (or cost cutting) capacitors on the input side.
Newer regulators require much less delta, sometimes less than one volt, hence the term LDO; low drop out.
Powering up with the probe disconnected makes it go into what I think is some sort of test mode, where the display alternates between everything lit up and a bunch of codes on the number displays. All the ones I've looked at on Google and eBay show different codes to mine, no idea if that is significant or not though!
Perhaps find or buy the manual, to decode the error messages
Here's some shots of the chips on the main board.
An early 'microprocessor' based instrument. Quite a leap forward in hardware design.
Congrats on getting this far, and for sharing your story here.
Lloyd said
Ah, yes! I'd forgotten about video walls! Something you don't see anymore.
Oh, but you do... they've just evolved!
Video starcloths, anyone? http://www.starcloth.co.uk
They have evolved a little bit haven't they! the last place I saw a CRT video wall was in a shop in Stratford-upon-Avon, I can't remember what the shop was actually selling, I just remember the 9 or so large CRT monitors sat on the floor playing something.
Now there's a random pondering, what did they use with the monitors to split up the video signal to feed into them so it looked like one big picture, and wasn't just a load of small pictures of the same thing?
I have just found a user-manual that covers this model, looks like I might have to have a play with this thing again, and hopefully it'll function correctly, if I follow the procedure for storing the different values for each colour on it, rather than just stabbing at buttons to see what they do! I also noticed that looking at these on eBay, they are not cheap! Another unit would cost me around £100, so all the more incentive to get it working. I'll report back later as to whether or not it works after setting it up. I'll have to find a new memory battery too, I have some Lithium 3.6V's that look like AA cells, I'm just worried the unit will try to charge them, and make them go pop! Oh, and the 'E01' code is 'memory battery flat'.
Regards,
Lloyd.
Lloyd said
Now there's a random pondering, what did they use with the monitors to split up the video signal to feed into them so it looked like one big picture, and wasn't just a load of small pictures of the same thing?
Effectively, multiple frame stores - you take the video in, digitise it, split the picture up and then interpolate each section in a frame store so each TV sees a complete 'picture' - except of course, it's only a portion.
Last one I had any dealings with used to form the backdrop to Match Of The Day when it was based at Television Centre - that involved several hulking great Christie projectors concealed in the set and a clever computer-based control system (the name of which escapes me) that divided the picture up into strips between the projectors. The curved MDF projection wall had the unsurprising quality of having a focal point for speech, which was quite disarming the first time you found it!
As regards your memory cell: there are now NiMH versions of those NiCd memory retention cells.
Good news! This interesting little test-thingy actually works! After reading the manual, and setting it up, it's now displaying correctly on the RGB mode, and not all sitting at zero. So other than the regulator falling over, and the memory battery being dead as a dodo, there's nowt wrong with it. I found a replacement memory battery, in an old BT cordless phone, it's a bit bigger than the original, but is 3.6V and holds a charge. I've temporarily bodged it in, with a view to eventually replacing it properly one day. I guess that the original battery must have had some life in it when I first got the unit, as I remember it working before I put it away.
It probably wants setting up correctly, but just to get it going I displayed some colour bars on the Rediffusion Mk 4, which also returned from storage on Tuesday, and used that as default values. Maybe I should use my little Sony 9" PVM monitor as the defaults, someone must have calibrated that at some point in it's life!
Regards,
Lloyd.
Lloyd said
Good news! This interesting little test-thingy actually works!
Well done!
After reading the manual
My Dad used to say "If all else fails, read the manual" A coworker would always say "RFM" - she didn't explain further..
Can you explain what we're looking at in your PIX?
The first one is marked R-G-B and also R/G B/G and G, but what are the units?
The second one may be marked x, y and X, what does this mean?
FordAnglia said
An early 'microprocessor' based instrument. Quite a leap forward in hardware design.
Not that early really, there are ICs with date codes from 1989 and 1993. The Z80 would definitely have been seen as an "old hat" device for use in commercial equipment by then too.
John
Yes, the manual is very useful! Took a while to find it as I was searching for the exact model number, TV-2140, and getting no results, but just searching for 'TV-color analyser II' bought up the manual for the base model TV-2130, and it explains that the 2140 is the same with a few extra memory channels. Here is the manual I found for it.
In the first pic the number display is showing the percentage of each colour from the colour info stored in the unit. So when you set it up on a white screen all 3 displays should read '100', which is 100%. If you then move to another monitor showing the same white screen the numbers will tell you the difference, so red is 100%, green 90% and blue 110% on the next monitor, red is the same, and green and blue need a 10% adjustment. Or at least that's how I understand it! The manual explains it better.
The bar-graph displays are called 'deviation display', the manual says: ' Deviations of the currently measured colour-balance values (R/G, B/G) from the colour-balance values of the selected white standard or reference colour are displayed by the top and middle LED arrays. The bottom LED array shows deviation of the currently measured G beam intensity from the memorized G beam intensity. Each array expresses deviation in percentage terms.' It then shows you how to read the percentage off these displays, the 2 green LED's in the centre of the displays are the 0% marker.
The second pic shows the unit in Chroma mode, and the displays show chromacity co-ordinates on the top 2 displays (x, y), and the bottom one shows luminance (Y, in cd/m² or FT-L) of the colour currently being measured. I think I had it set to cd/m².
The bar-graph displays are doing effectively the same thing they do in the other mode, except that the top 2 are for the deviation of the x&y values from the stored white standard, and the bottom one is deviation of the luminance from the stored luminance value.
Regards,
Lloyd
Lloyd said
The second pic shows the unit in Chroma mode, and the displays show chromacity co-ordinates on the top 2 displays (x, y), and the bottom one shows luminance (Y, in cd/m² or FT-L) of the colour currently being measured. I think I had it set to cd/m².
That would make sense - the Minolta meter I use to set camera line-up lights can show x,y for chromaticity but can also show colour temperature in Kelvin; it will also show the absolute level of light in lux (cameras are lined up to 600 lux).
From memory, monitors are set to 88 cd/m².
Lloyd said
Yes, the manual is very useful!In the first pic the number display is showing the percentage of each colour from the colour info stored in the unit.
As you replaced the battery powering the stored data I assume the data was lost, and the unit is now referenced to a non-volatile data set from the factory?
So when you set it up on a white screen all 3 displays should read '100', which is 100%. If you then move to another monitor showing the same white screen the numbers will tell you the difference, so red is 100%, green 90% and blue 110% on the next monitor, red is the same, and green and blue need a 10% adjustment. Or at least that's how I understand it! The manual explains it better.
This is very useful for matching several monitors, as in the case of a video wall.
The second pic shows the unit in Chroma mode, and the displays show chromacity co-ordinates on the top 2 displays (x, y), and the bottom one shows luminance (Y, in cd/m² or FT-L) of the colour currently being measured. I think I had it set to cd/m².
The bar-graph displays are doing effectively the same thing they do in the other mode, except that the top 2 are for the deviation of the x&y values from the stored white standard, and the bottom one is deviation of the luminance from the stored luminance value.
It would be interesting to measure a black and white set, as 'white' in that context is tinted blue due to the P4 "TV White phosphor"
You could adjust a colour set to match P4 (not sure why, just an interesting experiment) Somewhere I saw a chart of x and y co-ords for various Luminant standards, so with this meter you could set you colour TV to match. Again, just a nice experiment with no practical value.
The Y (absolute brightness) meter would be very helpful in setting up a monitor (or TV) The monitors in a TV control room or technical area (Gallery) are noticeably not very bright compared to a home TV picture.
I have used a different brand light meter (Tektronix J16) in every job with CRTs, and I have a few of these for a future restoration project. Like a lot of kit with built-in NiCad cells, that eventually leak, it makes a mess of the internal parts, particularly PCBs. I'll post separately on this topic as I need some help cleaning up a regrettable mess in a Sinclair Microvison set I'm fond of.
Congrats again on rescuing a nice piece of kit, learning its operation, and putting to use in your workshop.
jjl said
FordAnglia said
An early 'microprocessor' based instrument. Quite a leap forward in hardware design.Not that early really, there are ICs with date codes from 1989 and 1993. The Z80 would definitely have been seen as an "old hat" device for use in commercial equipment by then too.
Agreed. I smiled when I saw the Z80 hardware in Lloyd's PIX.
While MCU-based instruments didn't appear overnight they did make a big impact on what could be done (compact, functionality, and lower power and manufacturing cost), and how these same instruments were maintained and repaired. I recall some of the early examples (late 70's, early 80's) were a nightmare to work on.
In the modern world we find MCU cores 'just about everywhere'. If you have an iPhone with the Lightning connector you may already know there's several digital ICs *inside* the charging cable's plug. That's my day job...
FordAnglia said
As you replaced the battery powering the stored data I assume the data was lost, and the unit is now referenced to a non-volatile data set from the factory?
It would be interesting to measure a black and white set, as 'white' in that context is tinted blue due to the P4 "TV White phosphor"You could adjust a colour set to match P4 (not sure why, just an interesting experiment) Somewhere I saw a chart of x and y co-ords for various Luminant standards, so with this meter you could set you colour TV to match. Again, just a nice experiment with no practical value.
When the battery goes flat, all the data is lost. The unit then needs to be set-up again with a 'standard' monitor. When it's switched on after the battery has died, it displays 'E01' to let you know, and there is no factory data, everything is at 0, and it refuses to measure anything in the 'analyser' mode, the 'chroma' mode does show measurements, but the deviation LED's all stay at one end. This is what made me think it had another fault after replacing the 5V regulator, as I couldn't get it to measure anything in analyser mode, I didn't know at the time that I needed to store some default data to it first.
I might actually try measuring a black and white set! and see if I can set up a colour set to look like a black and white one!
Regards,
Lloyd.
Lloyd said
I might actually try measuring a black and white set! and see if I can set up a colour set to look like a black and white one!
By all means measure a black and white TV, but don't try and get a colour set to match it - it'll be way too blue.
Look up CIE Chromaticity Chart, and in particular look for the coordinates for D65 - otherwise known as Illuminant D. That corresponds to a colour temperature of 6500K.
By all means measure a black and white TV, but don't try and get a colour set to match it - it'll be way too blue.
.
The Philips G6 Dual standard CTV changed the greyscale to a much more blue shade on 405, not sure if it also did this on BW 625 signals. They were trying to match BW set pictures, it reverted back to Illuminant D on colour signals.
The early sets used to be set to illuminant C but that soon changed to D, C having more red content. Whether the change was studio lighting, cameras or better phosphor in the CRT I dont know.
Frank
I'm not sure where I read it, but I'm sure that early colour tubes had problems with some of the colour phosphors not being very bright, it could have been the red that was dull, I'll have to see if I can find that again!
Regards,
Lloyd.
The TV installation film avaiable in this sites library dates from 1967 and uses illuminant C.
The International Committee on Luminants (CIE) adopted a new series of illuminant standards in 1967 based on average daylight. The CIE standards of CIE 1931 were expected to be replaced by the new standards. Specifically the CIE1931 Illuminant C would be replaced by the new iluminant D6500.
Obviously the 1931 standard predates colour TV and the CIE standards are used by many industries.
I presume CTV started to use the new standard in around 1969 so had nothing to do with phosphors etc, all to do with following standards.
To be honest the paper I found the information in is way above my comprehension but is available as a download from.
It was published in the " Journal of the Optical Society of America" Volume 58 number 12, Dec 1968.
Have fun if you decide to read it, you can explain it all to me then.
Frank
