|
9" XG
Nov 28, 2015 3:57:52 GMT -5
Post by Casethecorvetteman on Nov 28, 2015 3:57:52 GMT -5
Hi all, Gábor and I have been chatting about the creation of a 9" XG.
I reckon the best option is to put all the XG hardware into a 909 chassis, as this will retain the standard tube mounting and lens arrangement, fitting all the XG boards into the 909 chassis should not be too difficult.
The tricky part will be making the yokes work with the XG boards, which should be possible, there seems to be nothing significantly different in terms of electrical design from a quick look.
|
|
justin
Junior Member
Posts: 60
|
9" XG
Nov 28, 2015 7:04:14 GMT -5
via mobile
Post by justin on Nov 28, 2015 7:04:14 GMT -5
this sounds great casey. I had been thinking about doing this with a bg808... as the hardware is cheaper than a 909? and more available.
I have a cnc machine I was thinking about using to machine aluminium plate and fold it to fit the 909 case.
really though. its the hqf lenses and lug tubes that make the late model 9" barco so great? not so much the ability of the boards? is that correct?
I presume the xg can resolve 1080p at 60hz which is really all we do with the cine 9/909? I dont see anyone driving it to its limits.
does the xg 909 combo retain the unreliable barco quad?
|
|
justin
Junior Member
Posts: 60
|
9" XG
Nov 28, 2015 7:06:29 GMT -5
via mobile
Post by justin on Nov 28, 2015 7:06:29 GMT -5
also. the neck board pcb isnt huge. surely it could be reverse engineered and a bunch of new pcbs could be made?
|
|
|
9" XG
Nov 28, 2015 9:13:46 GMT -5
Post by gjaky on Nov 28, 2015 9:13:46 GMT -5
this sounds great casey. I had been thinking about doing this with a bg808... as the hardware is cheaper than a 909? and more available. I have a cnc machine I was thinking about using to machine aluminium plate and fold it to fit the 909 case. really though. its the hqf lenses and lug tubes that make the late model 9" barco so great? not so much the ability of the boards? is that correct? I presume the xg can resolve 1080p at 60hz which is really all we do with the cine 9/909? I dont see anyone driving it to its limits. does the xg 909 combo retain the unreliable barco quad? Well the XG's bandwidth is a delicate topic, some are good some are bad (related topic: link), but I at least working on a solution which would cure the problem for those which are bad. On the other hand nowadays it is impossible to reproduce any neckboards from scratch because key parts have become obsolete and most of them very hard to find. But still there are a lot of used part that could be reused by the ever shrinking CRT camp.
|
|
|
9" XG
Nov 28, 2015 9:39:27 GMT -5
Post by Casethecorvetteman on Nov 28, 2015 9:39:27 GMT -5
this sounds great casey. I had been thinking about doing this with a bg808... as the hardware is cheaper than a 909? and more available. I have a cnc machine I was thinking about using to machine aluminium plate and fold it to fit the 909 case. really though. its the hqf lenses and lug tubes that make the late model 9" barco so great? not so much the ability of the boards? is that correct? I presume the xg can resolve 1080p at 60hz which is really all we do with the cine 9/909? I dont see anyone driving it to its limits. does the xg 909 combo retain the unreliable barco quad? The convergence on the Barco is the biggest killer, its terrible compared to any NEC.
|
|
justin
Junior Member
Posts: 60
|
9" XG
Nov 28, 2015 10:58:52 GMT -5
via mobile
Post by justin on Nov 28, 2015 10:58:52 GMT -5
please excuse my ignorance as I dont mean to be rude. what parts of the neckboard are hard to find? the plastic plug looks like an easy part to have plastic injection molded? and is there a reason why the board is attached to the neck and not remotely allocated elsewhere out of the way with wires to the neck pins? and use a slide on plug on each wire instead of the plastic plug part?
I figured the convergence was the key reason for doing this. I notice you mention it allows alot more adjustment. does this mean it can make for a sharper or just as sharp picture with hqf and 9" lugs as the barco 909s? or are the boards on the barco 909s what help make the set the sharpest?
if its not, what is the key reason for you persuing this? you have peaked my curiosity. is it availability of the boards and longevity of parts? the necs look so much smaller than the barco 909. not having owned an nec before, are the less boards? is it possible to reproduce the nec boards and have a frankenstein of a set that could last for years (or as long as tubes are made available)
|
|
|
9" XG
Nov 28, 2015 16:58:11 GMT -5
Post by gjaky on Nov 28, 2015 16:58:11 GMT -5
The CRT driver circuit (aka neckboard) is a very special design. The CRTs need (relitively) high voltage to drive them, the difference between full on and off state is about 100V in the driving signals, also we want this transition to happen very fast, at hundreds of MHz's (this is the bandwidth). This is a very-very hard task to achieve. So hard that only special breed of components could achieve this performance, with the death of CRTs these parts were also phased out since they were developed specially for CRTs. Neckboards need transistor to stand both high voltages but they are fast in operation, now you can't find any transistor meet these requirements. Now you can only find either high voltage transistors or fast ones, but not both at the same time, they are simply not produced anymore. This is why it is impossible to build new boards. Of course there are remained stocks here and there, but you can't be sure if they are original... Many projector makers built their neckboards on so called hybrid-ics (mostly made by sanyo), the story ended similarly here too, without demand the ic production was shut down. If we look it that way these CRT amplifiers were spectacular electronic desigs since there is no circuit which is copmarable to them.
The NEC projectors are so small because they are highly integrated, but they have just as many boards as other brands. Generally, convergence/focus/astigmatism driver boards contain amplifiers, similar circuits to hifi amplifiers, but while a hifi amplifier has a passband of 20-20kHz, these convergence amplifiers have a passband like 0-3MHz. Again these are very good electronic stuff. For a hifi amplifier you need two channels, in home theater say 7 channels. Now count with me. In a CRT projector each colour has individual convergence vertically and horizontally, has a dynamic and static focus and at least 2 amplifier for astig that is 6 amploifiers for only one colour. So in a CRT projector you have 18 high bandwidth high power amplifiers. And this is where NEC could make the biggest size shrinking. Normally these amlpifiers could be built from discrete part like transistors and resistors. In the Electrohome projectors the discretely convergence and astigmatism amplifiers are filling the full backplane of the machine. In a NEC the convergence amplifiers would fit in your hand. They use special integrated circuits for this, in one package there are 3 amplifier channels built in. Using these integrated apmlifiers allow better heat tracking between channels resulting -> less drift, also the key part are preselected in factory, more accurate performance is possible. But there is a downside too, if one channel blows up you have to change the whole package unlike with discrete amplifiers where you may end up changing a few part and you are done. An other downside of these integrated circuits that these are also phased out. Also since these circuits are relatively small the generated heat is very concentrated on them they need good thermal contact, fortunately if these circuits kept cool they last just as long as the discrete amplifiers.
|
|
|
9" XG
May 31, 2016 12:17:41 GMT -5
Post by gjaky on May 31, 2016 12:17:41 GMT -5
Casethecorvetteman: gjaky: So checked the XG coil more closely and looked up my notes, and it seems the XG has some exceptionally low inductance deflection coil. Each half of the H-Def coil has about 80uH inductance (in series mode that is 160uH, in parallel mode that is 40uH). The Marquee coil has 270uH windings if I'm right (S: 540uH P: 135uH). This means if a Marquee coil would be used in its stock form in a stock XG the retrace times would be awfully increased on the given ranges. This could be lowered by decreasing the retrace capacitance which in turn would increase the retrace peak voltage that is limited by the MOSFETs. Now the MOSFETs both in the XG and the Marquee can hold up to 1800V so this seems to be solvable problem. I also have data on the Barco 909 coil which is not splitted by stock form, but might be splitted probably, but nevermind, the barco's coil has even higher inductance 1.6mH in series that would mean two 800uH windings. That means even higher retrace voltage peaks than the Marquee, not a surprise barco engineers had to build a deflection board that can hold up to 3000V peaks! This is not a way for XG. I don't have data on Barco 120x deflection coils, but the deflection board used there I think has a 2000V blocking capability and has a 1.25us retrace time, which would indicate a fairly low (at least similar to Marquee) inductance deflection coil that could be used. Sony G90 (1292?) yokes should also have very low inductance, because their deflection system can only hold up to 1500V, but there is no data on their retrace time. But they are not splitted and I don't know if they could be made splitted... EDIT: MT do you still have your parts G90 to check the deflection coil?
|
|
|
9" XG
Jun 1, 2016 3:53:02 GMT -5
Post by Casethecorvetteman on Jun 1, 2016 3:53:02 GMT -5
Good work there mate!
|
|