CurtPalme.com Home Theater Forum

[PaulB]

Thanks for that guys. I'm just walking in the footsteps of giants and trying to give something back to the forum

Still need to brush up my soldering skills though, that de-coupling sandwich on the VNBs was a nightmare, an absolute pig of a job that was. I'll be very surprised if I didn't cook one of the caps!

Thats a good link Tom. The layout of the page makes it very easy to read through, nice.
Paul

[Nashou66]

well I Finally Finished the final CAD design board layout for the blue gamma boost and thank god
I sent jarmo the files to look at as he caught a mistake I had with one resistor( R199) being
parallel and not in series !! But APC in alberta are quick!
I sent them the files they needed and three days later the boards were delivered!!

here they are I made 6, i need to put the parts on and test it on my Modded 2005-03p Vim.




Looks really nice doesnt it!!

Now lets hope i got it right and it works ok !!!!

This will be part of the 2006-03p Vim upgrade once I get to that section.


Athanasios

[Nashou66]

STAGE I CLM Mod

Ok now on the the Control Module or the CLM. I did the mods McPherve talked about in the
JeaHong_lee thread. Also Mike parker posted some pics of his mod to these rails
as well(he used larger values) and this is when I first attempted to do any upgrades
to my marquee component level wise. Back then I was a complete newbie and forgot to put the
caps on with the right polarity and I fried something I could not fix back then.
On this next section i will not take pics of me doing the removal and of install as I think we all now
have a good feel of removing and installing parts.
I will show the locations on the non moded board and then the moded board. there are older versions
of CLM's(like the one I blew up) and I have not studied the schematics for them to name the parts
on those boards, so this is only for the newer boards as in the pics below. If i get time I may try to
mod an older CLM and post back here but some caps are the same and i'll mention those.

Mike parker did this mod to clean up noise the 5 and 15 Volt rails sent back into the rest of the projector
and he also found that after doing this focus sharpened up as well, I saw this also ,with just these mods.
These caps Decouple certain IC's on the board.
In the next section we will improve on this more by going after the caps right at the IC's.

Stock CLM with daughter boards removed:


Moded board


Ok first we will replace the four tants near the left connector pins, these are C1,C2,C6 and C113.

Four tantalums to be removed



We replace them with higher voltage value's and temp, Mike P also raised the cap values I have not on this
CLM but plan to on the stock one and compare the two. Since I do more mods that might not need the higher
values in these locations here.

EDIT: I originaly used a 47uf 40 V cap for C6 and C7 I have changed them
to 100uf 20v Oscons the previous cap was not an Oscon.So ignor the 47uf 40v caps

For C1 and C2 I used a 10uf 50v 105c low ESR cap, panasonic FC series. MP used 100uf.
For C6 I use an oscon 47uf 40v cap, sanyo made these.
For C113 I went with a 33uf 16v vishay tantalum, these are different than any tant I have seen before
smaller size but seem to work well, I ordered the smaller size by mistake ,the original
is a C size and I got the B size.

Now also here on the two pins that feed the + side of C2 and the - side of C1 we put 220uh inductors.
I use murata's as they were nice and small to fit in there. you need to lift or
cut(I lifted by desoldering from underneath) pins 27(-) and 28(+)
and solder the inductors from the lifted pin end to the hole of the removed
pin to complete the connection to the caps.

All caps and Inductors in place(one inductor is hiding under the other)
Also you can see the small orange 33uf Vishay tantalum




Next we will remove Tantalums C3,C7 and C43

Three tantalums to be romoved



C3 and C43 we replace with the same 33uf 16v tants
C7 decouples the same Ic's as C6(before) so we use the same 47uf 40v Oscon here.

Caps replaced, Oscon must lay down with longer leads to have daughter board fit over top



Next remove C12



Replace with another 33uf 16v tant



And last remove C145



Replace with same 33uf tant from before



Next time we'll replace a few IC's and more caps

Athanasios

[Nashou66]

CLM MOD STAGE II: IC's and Individual IC Decoupling caps.

EDIT: September 15th 2009: do not change the caps X31 and X32 to tants, these
are powered by the battery and using a polar cap here seams to drain the batteries if changed
already replace to the 100nf ceramic caps.

We will change out three IC's on this board and also a slew of caps, I have not gotten around to
all of them but have done at least half. there over over 100 ceramic decoupling caps at each
chip on the board. Most of these are digital circuits and you know what that does... yepp put
some digital noise into the power supply. Ken Hotte(KBK) changed out all these caps with no
issues, others have done them and have had start up problems. I have done 52 of the 122 I
counted in the schematic ( I will add this as a down load) so far, the rest will happen also.
The schematic helps with seeing which power rail goes to which caps. You will need to use a
continuity tester on your multimeter to make sure you put the right polarities of the caps to the
right side of the pad. From the schematic you find the large cap(ones we changed in previous stage)
that have polarities and locate all the caps that are fed from this cap or caps in a few cases.
You now need to take out your continuity tester put one end on the positive side of the power
rail cap and put the other end on the pads of the decoupling caps at each IC , when it
beeps(makes continuity) you found the positive side of the cap.
Just to make sure I also then test that cap i just found the positive side to and move the tester to the
negative side and test the other pad of the IC cap... it should beep also. I do about 5 caps at a time
then test the board.
I blew a few tants not getting this right so be thorough and don't rush!!!! You may blow more than
just the tants!!! So double check you have the right polarity,this is why i only do 5 at a time.

Here is what KBK said in his post on AVS Archives
KBK(Ken Hotte)

[Nashou66]

Just a reminder to everyone, DO NOT DO THESE MODS TO A NON RUNNIG PJ !!!!!

When doing thses mods in the thread I assume every one should be doing a few changes to a board and then testing
those changes by watching actual movies. I try to use a movie that has a scence that goes from black
to very bright and back again Like in Aeon Flux. This is an extream test for circuits in the projector i would think
so it would show up any problems quickly. At least 15 minutes of viewing is what i suggest.

I will re print what made me add this post from another memebers thread.

Paul B had a problem with his CLM and admitted to not even having tested any of his mods at all!!!!

[PaulB]

Conversation from another thread moved here for continuity and relevance. Paul

[Nashou66]

I was adding some more 47uf caps to my CLM and realized sometimes its difficult to find the correct polarity of the pad as some pads beep for both of the power cap polarites, so a pad on the 47uf tant will beep continuity to the +and- side of say test cap C2. so I added this to the original post for stage II.

EDIT:Alternate method for cap polarity: Some times its difficult to find which side of the pad
is positive and negative as the next cap switches which main power cap you test from, I found this
out today (5/23/2008) when adding another 10 caps. So you may need to look up the data sheet for the chip and find the power input pins and put the appropriate polarity of the decoupling cap to go to that pin on the chip.You
follow the trace from the pad to the pin on the chip, so if your following the +5v pin the + side of the tant
goes to it. the same for the -5v pin - side of tant to that.


Athanasios

[Nashou66]

Well I added about 30 more 47uf 6.3volt caps to the IC's on the +/- 5 volt rails . remember we are only doing the +/- 5 volt IC's not the ones on the +/- 15 volt rails.

On those we add the 22uf 16v dipped or A case tantalums, its your preference. I used the same Blue caps as on the three Ic's I changed out but I did not put in the .1uf tants as well. You could keep the 100nf ceramic there if you want and just add the 22uf 16v tant also to save time. Remember only the caps in the schematic which are mostly numbered with an X.
KBK also suggested to try 33uf 16v tants also and see which you like. I went smaller to try an avoid any power supply start up problem that may arise...None so far.

EDIT: 05/25/2008 After changing out more caps the stability of the image has greatly improved.
I watched live free or die Harder and was very impressed with what I saw.
The slight shimmer that is in my unmodded 8000 that my 8500 I am working on now also had is diminishing
more with each upgrade to my CLM. You wont notice the shimmy from your seat you need to get up to the
screen or look in the tubes. KBK was right that it was ironic the last mod he did was probably the biggest
improvement to the stability of the image. What I am saying is, I really like these changes cant wait to
finish with the 15 volt rails as well.

Athanasios

[Nashou66]

I Finished all the cap mods to my CLM and watched about 45 minutes for different scene's I am familiar with and every think looked great. I was also using my Moded 2005-03p vim to see how the power supply would handle all the extra capacitance on both boards. While watching the wavy line came back so I pulled the clm to give it a good cleaning of all the chip sockets
and go over the board visually one more time when i noticed that 12 smt caps I put in were the wrong ones. I had 47uf 6.3 volts where I should have had the 22uf 16 volt caps. even though it was more capacitance everything was ok power up wise.
I was going to leave them then thought the lower Voltage would not be good for those 6 op amps . So I depopulated those 12
47uf smd caps and put in the 22uf dipped tantalums that were the right voltage. So i decided to watch some video to see how the lower value tants would compare. All looked good but then the power supply was acting up, half the image would be drawn and black horizontal lines would be in the image and at one point it lost HVPS and refocuses again. However when the image was up it was really nice. It only happened on certain scenes that obviously put more demand on the power supply. I thought what if I use the Unmodded VIM, I put that in and no problems for an hours worth of viewing, I put my Modded VIM back In and back to power supply trouble. So I removed 4 22uf dipped tants and replaced them with some physically
smaller size B 6.8uf tants, and tried it this way still the problem but not as frequent.

So now My question for the guys out there more knowledgeable than me on this stuff, Can caps of smaller value
but larger physical size draw more or need more power than higher value but physically smaller caps?

With the 12 47uf 6.3volt case A sized caps I watched two movies no problems at all and it looked great!
When i lowered those 12 cap Values to 22uf but to a physically larger size cap I encounterd the power supply problems.
is this Normal?

EDIT: problem was with the VIM, So those changes worked fine.


Athanasios

[tse]

Good caps use no power. Larger value ones take more electrons to "fill up" or charge but once charged take nothing from the power supply. The Marquee power supply monitors current and if too much is being drawn at turn on it will turn itself off. This will happen in the first second of power on if it is going to happen.

Bad caps might use power if leaky. If leaky enough to cause problems they will usually get hot and burn up. Electrolytic and tantalum caps can go bad if stored for several years. They might "reform" when used or might not.

Scott

[Nashou66]

[WTS]

I don't know if I'd be taking all the .1uF caps off in exchange for the 4.7uF ones. The .1 caps do a far better job of getting rid of the high freq. trash/noise on the supply lines than the 4.7uF do.

[Nashou66]

[Nashou66]

Well I removed all the 22uf caps and went with 100nf(.1uf) tants, this is the same value of capacitance of the original just a different type of cap, and I am still getting the problem with my modded VIM, I do not think there is anything wrong with that Vim either . I should test it in my other marquee(i'll do this later today). I also removed Caps from the VIM to lower the overall capacitance, I had caps in places I think really didn't need it . When using the unmodded Vim I have no problems...Hmmm maybe it is my modded Vim causing the issue? One thing I noticed is for the first 6-10 minutes it doesn't happen , only when it warms up the Black horizontal streaking occurs . Unless its the DPB acting up from all the constant removing and installing it. that will be the first thing I'll swap out as i have a few of those here.

Athanasios

[1031]

Hello. I have looked that cap chaging stuff..I think there is alternative way to reduce noise on those control lines. When i tested better op-amp on focusmodule, i noticet that output on that "better op-amp" was much noisier than originals..Why? those better ones are lower noise than originals. That better op-amp has much wider bandwithd so it amplifies frequencues much higher that actuall control signal is. Ok that better op-amp can be much stable, and noise that amp generates itself is lower than originals etc.. How about using EMI-filters on those control lines. Those are available at different freq.ranges so those can selected for proper freq.range so that actuall control signall is passed but noise is bloked?

https://www1.elfa.se/data1/wwwroot/webroot/Z_DATA/45c48160-7921-11dc-bb36-0019bbdf5d02.pdf

I remember that when i had 1272 sony, those was used in certain lines.

[AnalogRocks]

A different aproach. Great!

[AnalogRocks]

A different aproach. Great!

[Nashou66]

[1031]

[Nashou66]

[1031]