CurtPalme.com Home Theater Forum

[cmjohnson]

Register to remove this ad. It's free!

Parts have arrived for me to rebuild (recap) three LVPS units.

That's up to 53 electrolytics to replace per LVPS unit.

It'll be a long afternoon's work.

I wish I had the schematics for the LVPS, then I'd have some idea as to where to place some
low value bypass caps across the electrolytics to improve impulse response.

And also figure out where to place some other caps to divert switch mode noise to ground before it ever leaves the LVPS.

[gregstv]

Maybe its time to get the CRO out and see what noise is on each rail. I have found quite a bit of ringing on the 808s power supply. I have reduced it quite a lot using snubbers.

[cmjohnson]

I need to do exactly that, but I'm really regretting having recently sold my Tektronix 2465B 400 MHz scope.

Fortunately, I do have other scopes but every one of them needs some form of repair. Even the best of them is overdue for recapping the power supply.

[cmjohnson]

I just replaced the glycol and cleaned the LC chambers in my main movie watching machine. Once it's back together it'll give me a good,
clean, sharp performance baseline to compare the Barquee's performance to.

[cmjohnson]

I put the crt asemblies back in, being very careful to do everything just right. And I have no picture, no emission from the CRT faces at all.

Everything works but no HV crackle on startup, and I've even tried powering it up with HV to one tube at a time.

I had an HDM failure light on the CLM once. I swapped HDMs. No more error lights.

I'm using exactly the same boards that were running in it before.

I've tried THREE known good HVPS units and two HV splitters, also known good.

Just in case, you know.


I've never had a Marquee that didn't display a picture manage to defeat my troubleshooting for so long before.


This is...kind of odd. I'm not seeing any reason for this machine to not be turning on the high voltage and giving me a picture.


Well, it looks like it'll give me something to do tomorrow. I'll pull the unit apart and check every assembly on the test unit, one at a time.

[Tim in Phoenix]

Say CM

Check for H sweep, V sweep and G2 as to clues why the high voltage is down. Perhaps the lvps has taken a dump.

[cmjohnson]

That would be an unfortunate problem if it turns out to be the LVPS as I'm currently using LVPS no. 3 after 1 and 2 both failed last week. I don't have a no. 4.

However, I do have a complete recap kit for all three of my LVPS units. Did you know that there are 53 electrolytic caps in one LVPS?

I've found, in each LVPS that has stopped working, that there are one or more puked caps. And they're always SMALL ones.

Come to think of it, when I first powered up the PJ this evening, I heard a faint squealing sound, dropping in pitch, like letting air out of a balloon by squeezing the nozzle end of it between your fingers. My first thought was HVPS, my second thought was HV splitter block, and now, come to think of it, I'm starting to suspect that I was hearing a cap letting out its one and only Final Faradic Fart.

And upon consideration, I THOUGHT I smelled something funny for just a moment. I dismissed it as possibly being my upper lip.

I'l pop the LVPS open and look for cap vomit.

Betcha I'll find some!


I estimate it's going to be a solid two hours labor to recap EACH LVPS by itself.


Hey, Tim, I've got a hypothetical question for you. How do you think the HDM will react if I connect a deflection yoke to it
that has HALF the inductance of the regular Thomson yoke? (Wired equivalently, but still half the inductance.)

Being able to run the Barco yokes is something that will really help to facilitate one of the goals of my HR project, which is to run
all Barco magnetics so I can get the full benefit of the presumed sharper focus that should come from a properly positioned K-D 3203 focus yoke as found in a Cine 9.

[cmjohnson]

It would appear that this LVPS did in fact fail, and where it failed was in the 390 volt circuit that feeds to the HVPS.

Every other function was working. The projector acted normally, relays clicking when appropriate, etc. except for no picture.

I'm halfway through recapping it.

[Tim in Phoenix]

Say CM

I am no engineer so clueless regards your yoke question.

[cmjohnson]

I recapped the first of the three LVPS units and this evening took pretty much everything to my workshop and tested everything.

The recapped LVPS works flawlessly, all three of my HVPS units are confirmed good, and I have not less than three good HDMs and three good focus boards, two of which are FGMs and one is of recent VDC manufacture. It turns out that I actually have everything needed to make one of my units into an Ultra. My main unit's motherboard is an Ultra capable type, with the traces to support the extended geometry functions. Firmware rev is 6.0 so this will be put back together as a true Ultra.

All the CRT assemblies work fine, and actually I found nothing out of the ordinary. For once everything really DID go my way, with no snags or show stoppers. Usually, for me, as Gilda Radner would have said, "It's always something. "


Now would be a good time to ask: Is ACON Ultra actually any different from regular ACON, and if so, how? Is the ACON control board different, is the camera different, or both?

[cmjohnson]

And...show stopper. My HDM and focus board and all three CRT assemblies and both power supplies all test good in my test unit.

But I'm still not getting any HV in my main projector. With the same set of known good parts.

Something's gone sideways in the sections that are very difficult to get to, given that the PJ is hung from the ceiling and backed
up against a wall, so close that the heatsink can't be swung open on its hinge, and there isn't even enough room in there to change
out the slot 2 option card. Just making connections is a challenge.

I'm going to try my replacement LVPS once it arrives and if that doesn't do it, it's got to come down for more diagnostics.

Not really looking forward to that. It requries an attic crawl so the unit can be lowered via the rope and pulley system.

[cmjohnson]

OK, this is getting silly.

I changed out the ENTIRE rear heat sink assembly, taken from my perfectly good, working test unit.

STILL no HV.

That leaves the boards in the card cage, the wiring harness, and, possibly, an LVPS that is not generating the 390 volt
drive signal to the HVPS EVEN THOUGH that same LVPS works perfectly in the test unit.

I've pretty much stripped the test unit. So it's not even usable as a test unit until I put it back together.

I have guns, lots of them. When I find the offending part that has put me through all this, it's going to be the guest of honor
at my next range day, which will be coming extremely soon!

[cmjohnson]

Heaven forbid, I may actually have to dive into the service manual, track down the HV inhibit signals, and start probing for them
to see what's not kosher according to the CLM.

[cmjohnson]

I started to work on adapting a Barco deflection yoke to work with the Marquee HDM and VDM this evening.


Wow, that got complicated!

While the yokes have the same numbers of coils, the way they're connected up to the wiring harness is WAY more different
than I first imagined.

The yoke is getting a really non-trivial set of modifications to make this all work, including removing the Barco stock connection board on the yoke body and replacing it with two of the ladder style connector blocks, removed from a donor Thomson yoke. Fortunately the Barco yoke has provisions to take up to two mounted connector blocks.

Oddly enough, the set of Barco yokes I got, all pulled from the same projector, are slightly different. One has balancing resistors
mounted under the connection block, to balance two coils, and the other two don't have those resistors. Manufacturing oversight? Intentional? I don't know.

Hopefully tomorrow the modified yoke stack will get its first live test.

If it's successful, then I'm going to document the mods and post them to the frankenyokes thread, because it becomes very relevant if this is a successful way of being able to make best use of the Barco focus yokes.

[haireez]

👍👍👍 Very interesting and looking forward for this. 👌

[cmjohnson]

A general note of interest for Marquee modders.

What's the most commonly suggested mod?

Going to low ESR and higher value capacitors.

Well, maybe that's not always such a good idea. By doing that you reduce the impedance of the circuits
when they get turned on. This creates larger inrush currents and if the inrush current is too high, the LVPS may think
there's a fault, and shut down because of it.

Guess what the symptom is?

Multiple attempts for the LVPS to start, and finally starting after several attempts when the caps are adequately charged
so the inrush current is not excessive anymore.

You can generally expect that the engineers that designed the Marquee knew how much capacitance was needed in the
circuits they designed. Although there is certainly SOME room for improvement and in some places, higher value capacitors and/or low ESR caps may be beneficial, this is only true in a limited sense. Don't go overboard on cap mods.

Too many caps added will cause more problems than it solves.

[tschaeikaei]

Yes CM, let me add that it could be a good idea to add a little resistance sometimes.
Depends on the voltages and currents involved, but it can be helpful.
A good cap parallel to the supply voltage and a fraction of an ohm in series.

http://www.2pif.com/images/low-pass-filter.png

If the results are not good enough, you can simply use multiple stages.
It's not too good for a high current/ low voltage supply, but it could be useful at .. let's say the 390V input of the HVPS.
Since the guys who developed the projector knew what they were doing, they added some (usually) 1 Ohms resistors directly
to the voltage inputs, then the caps.
Their second purpose is to act as fusibles.

[cmjohnson]

Yes, I've seen those resistors and understand their purpose. I've seen much more extreme implementations of inrush current limiters in the past, for example, if I recall correctly, the old Pioneer SX-1280 stereo (185 watts per channel!) I resurrected from the dead used something absurd like a 3 ohm, 20 watt ceramic box resistor switched into the primary side of the main transformer via relay and it was kept switched in for about five seconds.

In that five seconds it would get roasting hot. Sometimes that resistor would fail. Its 20 watt rating was marginal.


And then there was the Krell KSA-150 I resurrected last year. Its inrush limter resistors (10 watt size) got so hot, so often, that even though they were spaced off the PC board by about an inch, they'd still done quite a job of browning the board from radiated heat.


I for one am just not a proponent of lowering ESR and adding capacitance "because I can". I need to see a performance justification for it. If I do attempt it, I'll try it on just one channel. Say, the red channel on a focus board. If the "better" capacitor makes no difference when compared side by side with the stock green channel, I'll revert to the original part. Or a new one of identical specifications.

Switch mode power supplies tend to be a little bit finicky when you start changing load circuit impedances very much. I'd rather give them the designed load they were built for, for maximum reliability.



Now, as for my projector upgrade program, I'm in no rush, really taking my time with the modifications to the first deflection yoke, as it's way more complex than the job first appeared to be.

I've ended up transferring all three of the terminal strips from the original deflection/convergence assembly over to the Barco yoke, and duplicating the Thomson wiring scheme with the Barco coils. But admittedly, it gets confusing at times because I am not
completely absolutely certain about how every coil is configured internally.

It also appears that I may have to (gasp!) modify the HDM resonant circuit values in order to make this work properly, because
I think there will end up being some substantial change in the electrical values of the adapted deflection yoke as compared to the stock yoke.

So it's going to be time for me to dig out the manual and a calculator and start figuring out the resonances and required component value changes.

[tschaeikaei]

Hm, those resistors are not for current limiting only. They act as a low pass filter together with the caps.
I've been playing around with a linear output stage for my fan controller/ regulator and since I'm using bipolar transistors now instead of mosfets,
i nee clean DC at the base of those transistors. Two resistors in series on + (440Ohm and 4k in my case) and a 2.2uF between them to ground serves well
if staing under or at 2mA. The trick here: By adding resistance, you'll agree that you can take the ESR of the cap out of the calculation.
It's effect will be neglectable small. You'll end up with a perfect flat DC like from a battery. Important is that the first resistance is bigger than the second.

If you're using a cap that has lets say 0.2 Ohms, you could use resistors with 0.5 and 1 ohms.
That transfers it into a more usable dimension for higher current applications.

Something else not anyone knows: It's always good to add a small ceramic or film caps, especially if you're dealing with higher frequency noise.
Those smaller caps have better high frequency filter performance, in particular the often overlooked ESL.
That's what they did in the above picture. And that may be why id didn't add any caps to my projectors yet.
I may be in the future, but it's not simply "the bigger the better", like we agreed above.

3 Ohms seems extremely low for a mains input current limiter.
Since calculating the current involves a lot of ugly integrals and factors i do not know, i don't do it.
But 3 Ohms, that's nothing for 115V AC. Is this a ring core transformer we're talking about?

[cmjohnson]

I have a "sacrificial" tube assembly specifically for use in testing ANYTHING relating to a change to a deflection system.

I run it by itself, with sense jumpers placed on the motherboard to fool it into thinking three good neck cards are connected.

If I don't burn up that "sacrificial" tube then when I'm done with my testing I could sell it. It's a pristine, brand new condition 8" red LC assembly from a Marquee 8520. Just one scratch on the face of the C element, which you would never see on the projected image.