CurtPalme.com Home Theater Forum

cmjohnson · Joined: 03 Apr 2006 Posts: 5180 Location: Buried under G90s

Those of us with Marquees soon learn that the finest steps possible in converging a Marquee are fairly coarse, really coarser than they should be when attempting to get perfect convergence in full HD.

I've found that with a well aligned projector, there's usually a lot more total convergence range than you need.

So, I'm figuring that if you were to modify the convergence circuits so as to reduce the convergence amplifier gain, this would make convergence steps smaller. Thus, finer. Basically just scaling down the total effects of the convergence circuits, trading excess capacity for finer control.

My own 9500LC has roughly twice the convergence range that it actually needs. So if amplifier gain were reduced by half, then presumably the convergence step size would be twice as fine as it is now. Which I think would be just about right.

I just checked my service manual and there are six identical convergence amplifiers, all fed + and - 24 volts from the LVPS. I haven't determined exactly the best way to modify the system to reduce the overall gain of the amplifiers yet, but whatever would get the job done would need to be done on all six amps.

Thoughts, comments, suggestions? Anybody ever tried this?

gjaky · Joined: 05 Jun 2010 Posts: 2802 Location: Budapest, Hungary

This sounds to be a good way. The convergence amplifiers -in general- use a current feedback topology, so I'd modify the gain right on the CLM where the convergence signals are generated. Easy, simple.

mp20748 · Posted: Sun Feb 07, 2016 11:03 am Post subject:

Sounds interesting CJ

cmjohnson · Joined: 03 Apr 2006 Posts: 5180 Location: Buried under G90s

gjaky · Joined: 05 Jun 2010 Posts: 2802 Location: Budapest, Hungary

cmjohnson · Joined: 03 Apr 2006 Posts: 5180 Location: Buried under G90s

We always have to deal with SOME heat causing drift anyway. As long as the drift is repeatable and the projector is stable when it reaches a stable operating temperature, we can certainly handle that. And in fact, we do.

I'm pretty sure that every CRT projector exhibits a little drift during warm-up. We expect it and let it warm up and find that as long as we did the same thing last time, no further tweaks are needed.

My 9500LC has about a 20 minute warm-up time. If it's perfectly converged when it's warmed up today, if I don't mess with anything it'll be perfectly converged next week after another warm-up period.

I expect this to be consistent for most any CRT projector.

gjaky · Joined: 05 Jun 2010 Posts: 2802 Location: Budapest, Hungary

Ok.

Centering is carried out on the HDM and VDM, perhaps you should modify them too.

cmjohnson · Joined: 03 Apr 2006 Posts: 5180 Location: Buried under G90s

I'm not sure that there'd be any benefit to that. The HDM and VDM also provide the muscle for all geometry adjustments, so we would not want to restrict their range. Especially when you consider that most CRT enthusiasts prefer to run with maximized rasters for all the benefits they provide, which also means that a lot of geometry adjustment is required too.

gjaky · Joined: 05 Jun 2010 Posts: 2802 Location: Budapest, Hungary

I don't think you'd have to sacrifice geometry range with lowering the centering sensitivity, as in both VDM and HDM an independent circuit does these adjustments. At the VDM U4 and at the HDM U7 and U12.
But I'm done, you always wanted find the way why my suggestions are bad. Do it as you please...

cmjohnson · Joined: 03 Apr 2006 Posts: 5180 Location: Buried under G90s

No, I'm just trying to think it out. Don't be hurt that I'm not yet quite agreeing with what you are suggesting.

As for doing this mod in the HDM and VDM, which affects all geometry and size adjustments, all electrical signals ultimately can't exceed rail voltage limitations. While reducing the gain of VDM and HDM amplifiers would result in finer steps of adjustment, they'd have to have enough adjustment steps in reserve to be able to make the larger corrections that may be needed. This is all driven digitally and ultimately the control signals pass thru a DAC and end up being amplified and sent on to the yokes. But if you simply drop the amplifier gain, you are likely to find that you don't have enough bit depth in the DAC to make larger corrections in this lowe gain scheme.

Let's say that you have to make a 12" correction to geometry at the right edge of the screen, a correction that can easily be done in the unmodified projector, but you've reduced amplifier gain by half. The correction that would be a 12" move at the right edge of the screen may now be something like a 6" movement, and inadequate for the task. That's not a problem if you have more "room" in digital adjustments to get there, but if you run out of bits and the DAC is registering 11111111 and it's an 8 bit DAC at this point, then you have reached the limit of adjustment range. And it's not like you could just drop in a 12 bit DAC and get your full adjustment range back. Oh, I suppose that's possible, but I think you are now in the realm of a major re-engineering effort. Essentially you'd have to build a whole new CLM to support enhanced bit depth to give greater fineness of adjustment and retain your accustomed full adjustment range.

Having a new, advanced, super high resolution capable CLM would certainly be nice, but we're never going to see it.

Any tradeoffs of adjustment range in favor of precision of adjustments will have to be made with a critical eye cast toward the answer to the question, "How much range can we sacrifice to do this?".

Nashou66 · Joined: 12 Jan 2007 Posts: 16171 Location: West Seneca NY

the feedback resistors is the way to go as Gabor suggested. Also better op amps on the VDM help stabilize the control signals.

Aslo what Chris Stevens had done. All Wet tantalums on the CVA and VDM. With good quality Film by pass caps on CVA and VDM to replace all the ceramics. But the biggest stoping of drift( i now its not adding a finer control) is to get all the parts to the stable temperature where their values do not change or have the least amount of change between them all. and that is hard to do as each part
has a different temp where their values remain constant. Unless you can keep them at room temperature magically.

So this could mean matching all parts for all channels like in high end Audio equipment or precision measuring instruments.

HK Steve did this I believe. Not an easy thing to do with op amps tho.

Nashou

cmjohnson · Joined: 03 Apr 2006 Posts: 5180 Location: Buried under G90s

I've made arrangements to pick up a second complete and running 9500LC from a friend of mine who has gone digital, and I can understand why he did it. There really is no practical and good way to hang a 210 pound beast from the ceiling at his current house, not without spending big bucks to remodel a room and he's not doing that.

So, I'll take the second machine and am seriously considering working my way through the whole Chris Stephens mod playbook, plus incorporate the best of all the "modern era" Marquee mods as well.

It'd be kind of nice if we had a single thread here somewhere that defines the current state of the art for all the best mods available for the Marquee. (That's a hint, BTW, for anybody who can contribute to such a topic.)

Of course I'd test this convergence squeeze mod once it's figured out exactly how to best do it. If it takes a CLM mod, well, OK, but I am not too thrilled at the idea of doing mods on the most complex and finicky board in the whole machine.

The machine that I would (hopefully) build up would be based on the 9500HR special tubes that I've just acquired, which should outrun even LUG tubes for potential sharpness. That's what they were made to do, after all.

Nashou66 · Joined: 12 Jan 2007 Posts: 16171 Location: West Seneca NY

cmjohnson · Joined: 03 Apr 2006 Posts: 5180 Location: Buried under G90s

That's kind of useful but it'd be nice to have a list of the available mods all in one place, including, where known, which mods are or are not compatible with other mods.

For example, MP VNB mods are at...what version now? Same with the VIM.

Which are the best mods available today? And who's doing them?

Having all that info in one place would be supremely handy for someone who wants to improve his projector and isn't already well versed in what's current, best, and available.

cmjohnson · Joined: 03 Apr 2006 Posts: 5180 Location: Buried under G90s

Scott suggests that doubling the value of the current sense resistor is going to be the easiest way to do the convergence mod.

I believe that the stock value is 4.7K.

draganm · Joined: 08 Mar 2006 Posts: 8990 Location: Colorado

nash have you tried running a ground wire from the CVA to the heat sink (courtesy of MP). Supposed ot work wonder and remove a lot of heat too.
If you check the ground here with a meter it's not unusual to see half a volt DC or more.