| Author |
Message |
dvh99
Joined: 25 Dec 2009 Posts: 2158 Location: nederland
|
| Posted: Sun Oct 17, 2010 8:25 pm Post subject: |
|
|
well i now have the el5166 in the 2nd stage and i do not see odd behaviour even when cranking contrast to 100.
_________________ 1 answer always poses multiple questions.
marquee 9500ultra HD10L moome hdmi1.3 v3+ some mods.
|
|
| Back to top |
|
 |
dvh99
Joined: 25 Dec 2009 Posts: 2158 Location: nederland
|
| Posted: Mon Oct 18, 2010 12:33 pm Post subject: |
|
|
picture of the project.
|
|
| Back to top |
|
 |
dvh99
Joined: 25 Dec 2009 Posts: 2158 Location: nederland
|
| Posted: Tue Oct 26, 2010 12:13 pm Post subject: |
|
|
anyone who knows how to attach the thingy in the black square onto the plug?
it has no thread on it.
|
|
| Back to top |
|
 |
dvh99
Joined: 25 Dec 2009 Posts: 2158 Location: nederland
|
| Posted: Thu Dec 23, 2010 9:01 pm Post subject: |
|
|
ok here is something funny i just noticed.
i have some problems with the addon board that i thought were because of an improper grounding scheme on my board but i have the vim now running as i am typing this on a addon board so the vim is outside the projector and the picture has never been so clean.
when the vim is inside the pj i see a lot of lines (grounding issue or emi?) going all over the place visible on the tubewith a dark background though not visible on screen but very apparant on the tubeface and now all those lines are gone like i said.
does this mean i have to shield the part where the vim goes into the slot because of emi?
_________________ 1 answer always poses multiple questions.
marquee 9500ultra HD10L moome hdmi1.3 v3+ some mods.
|
|
| Back to top |
|
 |
mp20748
Joined: 12 Sep 2006 Posts: 5689 Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
| Posted: Fri Dec 24, 2010 3:55 pm Post subject: |
|
|
adding boards or retro boards that deal with high bandwidth are not so plug and play.
You could spend the next couple of months trying to get to the bottom of this. There's is no simple fix or solution to making the noise go away once the board is installed inside the projector. You'll have to roll your sleeves up and try numerous tweaks and changes in order to either isolate what's causing it, it to make it go away from trial and error without knowing the cause.
I have hundreds or thousands of hours invested in running down noise. So I'm speaking from first hand experience.
Nonetheless, I'll toss out a few things that's critical for high bandwidth retro board design:
- The board should be as small as possible.
- It should also have as much ground planing as possible, meaning there's very little wire trace runs or components that's not close to a board ground.
- The opposite side of the retro board would need grounding in every area that does not have wire traces or components.
- the grounding of the retro board to the main board, could require better attention to where the grounding is connected between the two, and the amount of grounds to the main board may have to be doubled or tripled.
- Decoupling... is an art within itself, and in most cases could require that the power (main +/- feeds for the board) are sent through a filter network. Not just a cap or two.
There's a lot of discussion about caps and which are better and which perform better in these circuits. And what electrolytic should I use, when it really doesn't matter that much, because high bandwidth video circuits operate from fast switching supplies that require either inductors or smaller value non electrolytics caps to suppress, block or isolate the noise from being a problem in the video chain. Sometimes the solution requires multiple 0.01 to 0.1 caps in parallel. But in most cases, inductors are needed. This part is the most difficult, because the wrong inductor value can also increase noise.
So in short, it would be best to redesign your board as small as possible. And then make sure the components are reasonably closer to each other. Put grounding every where there's not a part or wire trace.
And before you start on any of this, i would suggest that you first run a ground strap from the convergence board to the screw terminal that I show in the link. That simple fix solves a huge problem in the marquee original design, especially if you're going after better bandwidth performance. That simple fix is also the best thing I've seen to help solve convergence drifting.
https://www.curtpalme.com/forum_archived/viewtopic.php@t=19648.html
|
|
| Back to top |
|
 |
dvh99
Joined: 25 Dec 2009 Posts: 2158 Location: nederland
|
| Posted: Fri Dec 24, 2010 4:21 pm Post subject: |
|
|
oh i meant the vim connected to a extension board with the addon board on it of course.
mike there is groundplane where there are no traces everywhere but around the opamps and i dont think i can make the board smaller then it is right now. i could squeeze 1mm here and there but all components are pretty tight
its funny but the vim picks up lots of noise when inside the projector and is absolutely free from any of that outside the pj.
underneath there are some traces (9v and -5v +5v) but the rest is all groundplane.
the 4r7 to the opamp is replaced by an inductor btw and decoupling is 1nf-10-100nf-1uf-6.8uftant.
as you see i used thick traces for the supply voltages and short paths to the decoupling caps.
the datasheets say not to have a groundplane under and surrounding in proximity of the opamp.
there are no floating groundplanes and i kept the supply lines away from the signal lines.
i just got a message btw from nashou and he says it might be coming from the hdm.
edit: i followed your advice some time ago by adding the extra ground to the conv board.
|
|
| Back to top |
|
 |
dturco
Joined: 06 Feb 2009 Posts: 3778 Location: Eastern Shore Maryland
TV/Projector: Runco DLP VX-3000i Marquee 9500 parts doner
|
| Posted: Fri Dec 24, 2010 4:30 pm Post subject: |
|
|
I'm no electrical engineer, but I would think all of that copper or your add on board is acting as an antenna and picking up the machines internal noise and amplifying it.
Does your Cell Phone work while your P/J is On?
_________________ Firefly rules. Can't stop the signal.
http://www.hulu.com/firefly
|
|
| Back to top |
|
 |
dvh99
Joined: 25 Dec 2009 Posts: 2158 Location: nederland
|
| Posted: Fri Dec 24, 2010 4:34 pm Post subject: |
|
|
my only phone i have is wired .
i must be a dying breed.
there is lots of copper on the original vim too but the addon board is now in closer proximity to the hdm, according to nashou this might be the cause.
_________________ 1 answer always poses multiple questions.
marquee 9500ultra HD10L moome hdmi1.3 v3+ some mods.
|
|
| Back to top |
|
 |
mp20748
Joined: 12 Sep 2006 Posts: 5689 Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
| Posted: Fri Dec 24, 2010 4:47 pm Post subject: |
|
|
| dvh99 wrote: | | oh i meant the vim connected to a extension board with the addon board on it of course |
I got that..
| Quote: | | mike there is groundplane where there are no traces everywhere but around the opamps and i dont think i can make the board smaller then it is right now. i could squeeze 1mm here and there but all components are pretty tight |
Your board design appears too large for the components you're using. Compare that board space to the same circuit of the original design on an 03 VIM.
| Quote: | | its funny but the vim picks up lots of noise when inside the projector and is absolutely free from any of that outside the pj. underneath there are some traces (9v and -5v +5v) but the rest is all groundplane. |
Yes, meaning your board design is some how capturing noise once it's inside the projector.
| Quote: | | the 4r7 to the opamp is replaced by an inductor btw and decoupling is 1nf-10-100nf-1uf-6.8uftant |
Replacing the 4.7 with an inductor is tricky. I would not go that way until I solve the noise problem first
| Quote: | | as you see i used thick traces for the supply voltages and short paths to the decoupling caps. |
Yes, but again, that's where you may need to work in a filter network. That is where and when you get into the parallel caps I mentioned earlier.
| Quote: | the datasheets say not to have a groundplane under and surrounding in proximity of the opamp.
there are no floating groundplanes and i kept the supply lines away from the signal lines |
A stock 50-2035-03P VIM HAS ground plane under the three opamps. It does not have ground plane under the AD834's only.
And since you're duplicating the original design, it would make sense to follow some or most of what was already working on the original VIM design.
And regardless of where the noise is coming from inside of the projector, the goal is the get the circuit to behave itself inside that same noise environment. That is what the original board design accomplished. And that should be your goal, because it has already happened in the original. And that is the case in every CRT projector design. The sweep, power, and other circuits generate and radiate serious noise through out the confines of the projector. The designers goal was to make the high bandwidth circuits function in those environments... unless you'll want to run the retro board outside the projector.
|
|
| Back to top |
|
 |
dvh99
Joined: 25 Dec 2009 Posts: 2158 Location: nederland
|
| Posted: Sun Dec 26, 2010 3:43 pm Post subject: |
|
|
mike i placed some components closer at the opamps some traces are thicker now and there is added groundplane.
|
|
| Back to top |
|
 |
mp20748
Joined: 12 Sep 2006 Posts: 5689 Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
| Posted: Sun Dec 26, 2010 7:26 pm Post subject: |
|
|
| dvh99 wrote: | mike i placed some components closer at the opamps some traces are thicker now and there is added groundplane.
 |
That looks really good there..
If I could suggest anything to change, ti would to make sure the grounding goes all around the board. meaning to make sure no runs or traces are running independent of ground on the outer edges of the board.
|
|
| Back to top |
|
 |
dvh99
Joined: 25 Dec 2009 Posts: 2158 Location: nederland
|
| Posted: Sun Dec 26, 2010 7:41 pm Post subject: |
|
|
no rest for the wicked
the best way to miniaturise is of course to use 0805 size components which i should have done in the 1st place, the board as shown here measures 7 cm by 9 1/2 cm instead of 8 1/2 by 12 1/2cm.
|
|
| Back to top |
|
 |
1031
Joined: 22 Mar 2006 Posts: 657 Location: Finland
|
| Posted: Tue Dec 28, 2010 7:33 am Post subject: |
|
|
Dvh99, you can get much "solid" groundplane if you route one of vcc trace bottom side of board. Now that trace "cuts" groundplane at many points and some parts of groundplane is connected via "narrow bribge"
Good idea to use 0805 size components! 1206 are quite easy to install, but board comes quite big..
Iīm not saying that your desing is bad (those seems now much better than my version ) but try to correct those narrow "bridges"
I know that those nearly isolated areas of grounplane are connected to botton groundplane, but "vias" through layers are not good for HF.
Try to think that top groundplane is for video and supply to op-ampīs and desing that as good/solid as you can, then if there is areas that are isolated, use vias to bottom groundplane.
_________________ Marquee 9500LC (Frankenyokes / Thomas electric tubes / HD-10L / +many modīs)
DVDO VP-50
New hobby, Rally
http://www.youtube.com/watch?v=vX2Rtpr1njs
http://www.youtube.com/watch?v=6ZP9FEFXV5c
http://www.youtube.com/watch?v=j065vei6j6s
http://www.facebook.com/pages/JTS-Racing-team/204443719572685
|
|
| Back to top |
|
 |
David_Web
Joined: 02 May 2007 Posts: 418 Location: Sweden
|
| Posted: Tue Dec 28, 2010 10:33 am Post subject: |
|
|
How bad would adding an RF cage be? Or three.
One might even add an extra plate above the vim where the quad decoder used to be.
_________________ SNR of people are ridiculously low.
|
|
| Back to top |
|
 |
1031
Joined: 22 Mar 2006 Posts: 657 Location: Finland
|
|
| Back to top |
|
 |
htguy1
Joined: 30 Jul 2008 Posts: 99
|
| Posted: Thu Sep 08, 2011 12:06 pm Post subject: Update??? |
|
|
Hi,
I think what you guys have done here so far is great! I just wondering how its come along since the last post?
One other comment I have is that the older 2005-03p does actually use the ad834 chip. (It also uses the H1100 opamps so thats not so good) Does anyone know if it has a higher bandwidth capability due to the 834 used on it?
I was playing with the ones I have and its seems to have a good pic but there is a lot of noise compared to my modded 02p vim. I noticed that it uses an LM317 v reg and was thinking there has got ot be a better option out that that could by a lower noise plug and play sort of replacement out there. Does anyone know of one?
|
|
| Back to top |
|
 |
dvh99
Joined: 25 Dec 2009 Posts: 2158 Location: nederland
|
|
| Back to top |
|
 |
htguy1
Joined: 30 Jul 2008 Posts: 99
|
| Posted: Thu Sep 08, 2011 9:39 pm Post subject: |
|
|
Hi,
Yikes! That better be one good v-reg for that kind of money!
|
|
| Back to top |
|
 |
htguy1
Joined: 30 Jul 2008 Posts: 99
|
| Posted: Sun Sep 18, 2011 2:45 am Post subject: more opamp options??? |
|
|
HI,
I see you guys are using the EL5166 as a drop in replaement for the cl449 I am wondering what your guy's thoughts are on:
THS3061 (TI)
LMH6703 (National Semiconductor)
LMH6702
FWIW I know the Nat. semi chips require a feedback resistor change, but I know that back in the day one of the mod companies used the LMH6703 on thier VIM mods .
God bless...
Mark
|
|
| Back to top |
|
 |
htguy1
Joined: 30 Jul 2008 Posts: 99
|
|
| Back to top |
|
 |
|
|