CurtPalme.com Home Theater Forum

[cmjohnson]

I'll say it again: NECK BOARDS ARE NOT RESPONSIBLE FOR FOCUS QUALITY. Not directly. If a board is softer in focus than another one, then there's something wrong with the softer board as it relates to the minimal circuitry that connects to certain electron gun elements.

If you want to compare how one neck board may alter focus compared to another one, you need to be looking at the sharpness of individual scan lines. Set up a 1 line on, 1 line off pattern and use that for the basis of comparison.

Do not mistake resolution for focus. They are not the same. But you can't make the best use of higher resolution if you can't focus sharply enough to use it.

Matsushita (Panasonic, VDC, Clinton, Lexel, Mikado) CRTs work best with the Kanto-Denshi made magnetics systems that were
designed specifically for those tubes for Barco and Sony.

The stock magnetics in a Marquee are Thomson products designed for Thomson CRTs which Electrhome stopped using back around 1994 or 1995. They elected not to switch to K-D yoke designs like every other manufacturer did due to cost/benefit reasons. They found the Thomson focus yoke to be "good enough" but it has always held the Marquee back on its ultimate sharpness potential.

Thomson magnetics on a Panasonic tube are WRONG. Oh, they work, but they are far from optimized for the task.

In fact, Thomson CRTs are SIMILAR to Panasonic tubes, but they are sufficiently different that EVERY aspect of the systems that operate them are substantially DIFFERENT. they require entirely different neck cards and even have a different anode voltage spec. (40KV) They're totally different, why would you expect that the Thomson magnetics would be right for Panasonic tubes?

[greg9518lc]

CM please do a G90 vs Marquee sharpness comparisson using standard yokes so we can see your ability
to dial in the Marquee before and after mod with pictures of center and all 4 corners. Have read but never
seen any documentation of these claims by you. Then describe what we are seeing and why the g90 yokes are better
on a Marquee. I only ask this because I have never seen any improvements demonstrated by you and Voldermort and
MP know they are useless. So the question is why are they better again????????

[cmjohnson]

My test environment doesn't include a screen. It's a (currently) very messy garage.

I'm not ready to show the results yet because I don't HAVE finished results yet. Until I get my
impedance matching transformers made up and installed, I can't run more than one G90 deflection
coil in a Marquee at once and even that should not be run for long because it draws too much current
and will eventually burn out the HDM in that channel. That's why I need the transformers.

My initial test run showed what I was hoping to see, in the ONE channel I tested it on: Very clean, very
sharp focused scan lines with great resistance to blooming at high contrast. Just like you see when you
run a Cine 9 or G90 at high contrast.

I will stop short of claiming that this makes the scan line/spot size any smaller than the Thomson focus yoke
when it's at sharpest focus, because I'd have to measure that on a ruled screen but I don't even use a screen
at this stage of the tests. But the stability of focus at higher IRE levels is what has me thinking this is very much
worth while.

Once I HAVE my first set of matching transformers and I have a full stack of G90 magnetics running, then I will
be able to directly compare, side by side, G90 and stock magnetics in the same projector. For that comparison
I'll use a red tube and a green tube (or two greens) and be able to run them side by side and document their
performance using ruled paper on the screen so there's actually an objective method of comparing scan line performance.

Mike's mods work. I've been running an older set of his modded neck cards and VIM for quite a
few years. The difference in low IRE detail between stock cards is easy to see.


Anyway, when it comes to the Thomson CRT/Panasonic CRT and magnetics subject, I am perfectly,
factually accurate on what I'm saying.

If Electrohome had made the RIGHT decision to maximize picture quality, they would have K-D design
a magnetics system for the Marquee. They chose not to spend the money because the Thomson yokes
were good enough for the needs of their simulation customers. (Which was the lion's share of their market.)

As a result, the Marquee is the only projector out of the whole industry that used Panasonic tubes not to use
K-D magnetics designed for the application. And it's the one with the worst focus.

Since K-D won't be designing a Marquee application magnetics system, to run them they need to be adapted to
the projector. I'm doing that.


The designer of the "new" VDC neck cards said to me that there's no way that the newer cards can possibly be sharper
or have more bandwidth than the legacy 2039 and 2038 cards, which are STILL the best. But all those cards are at least
10 years old and may need an overhaul. So you could be getting the impression that the newer cards are better because
you're comparing newer boards to boards that might be 20 years old and have not received an overhaul that may be needed
to restore them to full performance specs. There are several surface mounted tantalum capacitors on that board, and
they have a limited service life. Change them.

If you say the new cards are sharper, post pictures comparing SCAN LINE sharpness between a new card and any randomly chosen,
apparently functional legacy card.

I'm pretty sure that the newer cards don't have much of anything to do with scan line sharpness.

At this moment, thanks to a recent purchase, (I bought some neck cards) I now have a newer VDC neck card to test with.
I'll stick it in my test projector the next time I'm doing tests with it. So if your claims are correct,
I should see sharper scan lines, right?

Well, that's a test I can do today if I'm so inclined.

[greg9518lc]

Cm You are welcome to ask for proof and documentaion on the link in my signature
if you want to pursue this as Voldermort has a G90 and LCP and LUG Marquees as well has set up
peoples machines with frankeyokes as well has seen the barco yokes and states there is no evidence that
they make a difference but I see you believe this and we don't so look forward to you proving this
point. It has been shown the need to defocus the Marquee on 1080p 72hz or otherwise you see scan lines.
So either your Marquee is not setup right or you have other problems. After all who wants to see scan lines??????[img][/img]





Look forward to your response pic is a LCP Marquee that some people say they cannot focus

[gjaky]

[cmjohnson]

So why would it even matter that the image at 1080p is 72 or 60 Hz? That won't affect scan line sharpness,
but it CAN have an effect on horizontal resolution if you're already nearing the edge of your linear bandwidth capacity.

It is for this reason that I recommend 60 Hz as being demonstrably superior to 72 Hz refresh rates. If you don't see
flicker, that's as fast a refresh rate as you need.

Just look at the raw numbers: 1920*1080*60=124416000.
And for 72 Hz refresh: 1920*1080=149299200

That ignores the additional bandwidth required to account for the non-display information that is in the total signal,
but you can see that by going to 72 Hz you're increasing your bandwidth needed by 20 percent and NOT seeing an
improvement in picture quality. More likely you'll be running out of bandwidth, and if anything I'd expect the picture
definition to start degrading if you don't have a really good set of neck cards.

Mike has shown the "new" neck cards cleanly resolving the pattern at a pixel clock of 266 MHz, which is 133 MHz video bandwidth.

But how are they at 150 MHz bandwidth?

Actually when you include a 20 percent allowance for sync and non-displayed information in the video chain,
the real truth is that you need 150 MHz for 1080P-60 and you need 180 MHz bandwidth for 1080P-72.

So don't use 72 Hz refresh. All you're doing is pushing the video bandwidth requirement to or beyond the limits of the neck cards.

Incidentally, I personally have never had any problems with getting sharpness and good focus out of a Marquee equipped
with standard LCP tubes. Scan lines are apparent at 1080p, but the MTF value is not amazing.

[gjaky]

[cmjohnson]

Oddly enough, judder doesn't seem to bother me at all.

So there's that.

I'm not sure to what extent this is true, but I've been told that these days movies are often, if not usually, recorded digitally at 30 or even 60 frames per second, and if this is the case then there's no reason to use 72 Hz as there would be no judder to deal with.

[greg9518lc]

So 1080p 72hz 178mhz pixel clock running at almost 3x what a standard crt will resolve
and your point with the yokes is not needed. Look forward to your documentation of the
yokes being better than stock ought to be entertaining for all to see......

[cmjohnson]

Actually I'm recalling that when I started these yoke experiments, many years ago. Scott (Tse) tried out a sample frankenyoke (one I'd modded) and he did note that it was sharper than the stock yokes. But not so much that VDC had any interest in trying to get new ones made particularly when K-D was already out of the yoke business.

I will not amplify upon any claims of greater sharpness than stock yokes until I can demonstrate it to a reasonable standard.

Let this be sufficient for now: If the complete G90 magnetics stack is merely more stable than the stock set, with varying
conditions including contrast level, then that alone justifies the upgrade. If they actually yield a smaller effective spot/line size,
allowing the possibility of running at still higher resolutions, then that is a desirable bonus.

[gjaky]

[cmjohnson]

That would defeat the point of upgrading for the fun of doing it.

[gjaky]

[cmjohnson]

Oh, I could do that! But there's no picture quality improvement to be had by doing it. The simpler implementation of Scheimpflug in the Marquee LC assembly may not be as precision engineered but it works fine. I've actually looked into doing that, but the major problem is that it adds almost two inches to the height of the projector overall, which means you won't be putting covers on it.

[cmjohnson]

Let me put this question to the group:

Please only answer if you have personal experience with fully set up and running 9500LCs AND G90s OR you have experience with 9500LCs and Barco 909s or Cine 9s.

If you have direct experience with the 9500 and either or both of the others, answer me this:

Is the 9500 truly as sharp as the G90?
Is the 9500 truly as sharp as the Cine 9/909?

My prediction is that the answer will be a clear NO.

With the same tubes.

With, as many people say, the best video chain being in the Marquee.

So what's left that controls sharpness?

Yep...the magnetics.


The focus yoke in particular, but the magnetics are designed as a system and efforts made to keep the
systems together are likely to yield benefits in sharpness and stability.

[Curt Palme]

I've run through literally 100s of Marquees here, early, late model ,etc. About 7 years ago I got in a bunch of Marquee Ultras with LCP tubes, but with HD10E lenses. I saw a CLEAR difference switching from the stock GT17 lenses to HD10E. That's when the 10E/HFQ900 lenses were going for $1000/set.

Since then, people here have said that the 10E lenses are no better than GT17, but to me, the 10E lenses are still the cat's meow.

Now, to be clear, I did NOT do any A/B comparisons, which is why I always harp on people here that post about how good mods are, etc to show A/B pictures with the same shot, camera angle, exposure, etc, but very few people want to do a true scientific comparison, which is why I don't get involved in those discussions.

My claim though regarding the 10E lenses is simply that I've seen 100s of 9500 and 9500 Ultra images, and the 10E lenses showed an improvement. That's what I base my claim on.

[Curt Palme]

I've run through literally 100s of Marquees here, early, late model ,etc. About 7 years ago I got in a bunch of Marquee Ultras with LCP tubes, but with HD10E lenses. I saw a CLEAR difference switching from the stock GT17 lenses to HD10E. That's when the 10E/HFQ900 lenses were going for $1000/set.

Since then, people here have said that the 10E lenses are no better than GT17, but to me, the 10E lenses are still the cat's meow.

Now, to be clear, I did NOT do any A/B comparisons, which is why I always harp on people here that post about how good mods are, etc to show A/B pictures with the same shot, camera angle, exposure, etc, but very few people want to do a true scientific comparison, which is why I don't get involved in those discussions.

My claim though regarding the 10E lenses is simply that I've seen 100s of 9500 and 9500 Ultra images, and the 10E lenses showed an improvement. That's what I base my claim on.

[nidi]

I've been running my 9518LC Ultra (2004 mfg) with HFQ900, G90 yokes and magnetics and MP modded boards
for 6 years now.

The biggest boost in sharpness was clearly the HFQ 900 lenses, having tried GT17, HF 10F lenses and finally
HFQ 900 lenses.

THE HD 10F ones had a very wide mfg tolerance only one out of the 3 did come close to the HFQ 900's.

the next boost I got with the G90 yokes carefully setup by a pro.

the Contrast blooming was gone (no blooming up to 100).

also the MP modded boards did a little in boosting the sharpness.

unfortunatly, the hasen't been set up since then and has drifted out of it's optimum.

Hopefully, a guy from Denmark comes by and sets it up in the next few months.

Also, a good processor and input card will help in the visual experience a bit.

switching from 1080P 60 to 72 Hz (or 75 Hz) minimises the sharpness a bit.

But I hope , the latest mods from MP and Gjaky will help in that regard.

also the positioning of the G90 yokes is very important as well, still hopeing CMJohnson will have

instructions of how to get the outer plastic housings smaller so the yokes can be moved nearer to the front.


Michael

[VideoGrabber]

[VideoGrabber]

[cmjohnson]

Actually I only use test patterns when I'm putting a projector through the rebuild and repair process.

Once it's actually installed and it's time to watch movies. that's exactly what I do with it. Test patterns may be able to tell you about the characteristics and setup of the whole system, but they're boring to watch.

I use no more test patterns than necessary to set the projector up once it's in its room it will be installed in. After that I prefer to make final tweaks on actual content in the movies themselves. Such things as setting electronic astig seem to go better with an appropriate image out of a movie to work with rather than just a test pattern crosshatch.

I'm admittedly rather surprised, and dismayed, to learn that even now most movies are still recorded at 24 FPS. Really that makes no sense at all since movies today are not expected to EVER be shown on film, although that is actually a pretty recent development. It may be that producers and engineers will soon figure out that recording content at 30 or 60 FPS simply makes a lot more sense today.


As for the G90 yokes, nidi is reporting exactly what I've observed in my tests: Focus is very fine, and yes, I do think it's slightly sharper than with the stock yokes, and more importantly focus retention is excellent up to full output. That ALONE makes the change worthwhile.

As for positioning the G90 focus yokes, the wish to move them forward is PRECISELY why I'm working on adapting the G90 deflection yokes to run in a Marquee as well. I can do that now, but not on the green channel and to get the H. width down to within the limits of the CRT face the HDM needs a component value changed. After this is done I can demonstrate that the focus yoke is perfectly located when touching the deflection yoke.

But an interstage transformer is needed. The inductance (and thus impedance) of the G90 deflection yoke is too low, current draw is too high, and the need for a broadband impedance transformer between the HDM and the G90 deflection yoke is very clear.
I've got a transformer specialist company working on a set of prototypes for me now. I'll update my own Marquee mod thread with updated information as soon as I've got the prototype transformers in hand and under evaluation.