AVS Forum banner

Status
Not open for further replies.
1 - 20 of 25 Posts

·
Registered
Joined
·
101 Posts
Discussion Starter · #1 ·
I'll use this thread to document mods I'm making to my ECP 4100.


I like pictures, so anyone can readily see precisely what I've done. This might be more encouraging to someone else attempting a modification than descriptions like "All the caps in the video signal chain are pretty obvious".


My first and only mod so far was to the fans/grillework. (I've got to get the projector basically operational, aligned and tuned up before I start any electrical stuff) I'm pleased with the noise lessening result!


I got rid of the grillework entirely, opting to use nothing for a finger guard, as its ceiling mounted and dont really care about EMI issues. I freshened the fans by replacing the Delta Fans with 4 NOS NMB 3110KL-04W-B10 units.


I also used these rubber isolators, which lifts the face of the fan slightly up off the panel surface and supports the fan. These required removal of the studs and drilling 4 holes for each fan.


Photos show this mod and the look of it installed on the projector.
 

·
Registered
Joined
·
101 Posts
Discussion Starter · #2 ·
here's the other photo -
 

·
Registered
Joined
·
2,861 Posts
Yeah, looks good Joe. My 4000 only had 2 fans, so I am in process of adding two more. I built the filters last night, now I have to add them to the pj., How did you cut the grilles out on yours? relly a neat job.
 

·
Registered
Joined
·
1,390 Posts
Luckily the metal side panels are aluminum. Very easy to cut. For me, the grill removal was the single largest reduction of noise that could have been made. It was amazing the noise reduction after milling out the grills. It's almost an acceptable level with just that mod. I imagine that replacing the 4 standard fans with 4 lower noise fans will be all that's required for a nearly total silent PJ.
 

·
Registered
Joined
·
101 Posts
Discussion Starter · #6 ·
Actually I used a sawzall - kinda overkill, but it's what they had in the shop. I was lucky to find a narrow hack saw blade for it.


Of course to get it finished off, I worked with a half round file some and then a big stone on an air tool. Then some more with finer file. Took a couple hours to do all 4 positions with drilling the mount holes and all!


I'm surprised no one has asked about the rubber isolator mounts? I tried to look them up in Digikey and Mouser to no avail. Was hoping to provide a source for these, but I cant unfortunately. If I stumble into anything, I'll post it here. But it looks like a "please send me samples" situation for now.

These were originally used to isolate a fan on an ATX power supply. I'm sure they contribute some to the overall noise reduction.
 

·
Registered
Joined
·
2,861 Posts
I just use small rubber O-rings from the hardware store or small grommets from an auto store. They work great!
 

·
Registered
Joined
·
3,557 Posts
Here's a tip folks,


I have used standard foam weather seal (for doors, etc) available at your local hardware store for years to shut up fans. Simply by decoupling them from a resonating panel will drop the noise at least by half, sometimes more. Then you can start your other tweaks to continue reducing the noise.
 

·
Registered
Joined
·
2,861 Posts
that's a good idea too Mike. Never thought about the foam...I've always got tons of o-rings and grommets...now I'll have a use for all the excess westherstrip I have as well! Are you supposed to have this many parts left over from a project :D
 

·
Registered
Joined
·
101 Posts
Discussion Starter · #10 ·
I performed the power input cap mod to the RGBHV input board last night. Used two Nichicon 2200uf, 25V, LXF, 105C (brown case) and two 0.22uf 16V PEN Film ECW-UIC224JB9 SMD caps in parallel with each electrolytic, tack soldered to the back side of the PCB.


The Nichicons stand 40mm off the PCB, so now I have to take the side panel off the projector, in order to slide out the RGBHV input module.


Here's a picture showing the caps from the top side of the board, for those interested in their location. They're the big brown guys in the upper left hand corner!


If someone could point out the "12" signal path electrolytic caps (I can guess as well as anyone just looking at it, but, we like to be sure), to be upgraded, I'll do those and post another picture.
 

·
Registered
Joined
·
101 Posts
Discussion Starter · #11 ·
Here's a picture of the film caps, just to see what they look like -
 

·
Registered
Joined
·
101 Posts
Discussion Starter · #12 ·
Thinking about the 12 electrolytic signal caps, it'd be nice to have a schematic of the circuit. I'd like to see the signal loading - try to guess why they used 10uf 16V for this application.


As KBK once pointed out, whoever chose the components did so without regard to any impact to subjective aspects of the signal quality. Maybe they also chose the particular values used from a manufacturing standpoint (costs less to stock a bunch of 10uf 16V parts, which will do OK in this circuit, but you really only need 2.2uf and these parts will never see more than 4.5V across them)


Having a schematic of this circuit may allow one to choose a better, less costly upgrade candidate component - than just a brute force choice (10X the value and 3X the voltage).


Also, I have no idea what the low-end frequency response for a "R", "G" or "B" video signal is. You always hear numbers like "60mhz" and "100 mhz", but no one has much to say about the value at the other end of the spectrum. It's not "DC", because coupling capacitors are used -


One could figure out the low end bandwidth if you knew the loading Z for each cap in the chain. Again, maybe 10uf is a very rough cut at what it needs to be.


Then there's a consideration that if each 10uf coupling is identically loaded, you have all these "identicle" high pass filters in series, giving a very sharp rolloff. Maybe it's better to have the low end rolloff not so sharp, which may imply a staggered value selection of these caps.


I dont know video signals well enough to be able to say anything about what the low end response frequency or character does to an image. Maybe someone else here does?
 

·
Registered
Joined
·
2,861 Posts
Joe, I don't have the schematics with me here. But, I will give you the component listing tonight when I get home. As far as your other questions, I don't know wnough about the components and their selection to be of any help to you there. Based on KBK's lead, we simply replaced the signal caps with same value (or close) but a larger voltage. Cary also made the suggestion (which I did as well both to my board as well as the others I have modified for people) to add a tant in parallel with the lytics for the high frequencies (I think this was for the highs...going from memory here). The lytics handled the low end ok, but the tant (I used all 10mfd @ 25v) helped on the other end...Maybe I have it backwards, but I don't think so. I also replaced the decoupling caps with same [email protected] tants per Ken's suggestions. I assume you've found the couple of threads about the 400 mods? There was also one on the 3000 or 3100 that had some good info in it as well.
 

·
Registered
Joined
·
1,390 Posts
jjasniew -


Those are cool. I think I'll give 'em a try.
 

·
Registered
Joined
·
101 Posts
Discussion Starter · #17 ·
I replaced the 3 10uf 50V rubycons with 4uf 50V film caps. Yes, I went DOWN in capacity, in exchange for an increase in cap quality.


Since I have no idea of what the circuit loading is (I have no schematic) I cant say what the change in high pass cutoff is, for the values I used.


I reset the drive levels for a grey background and all looks well. I still need to pooge the input board RGB signal chain, to be able to say if the change has done some good, as that's upstream of these, signal wise.


There's 2 330 uf caps on this board. W/O a schematic, I have no idea what changing/replacing these two would do. The circuitry around them isnt obviously "power supply" as far as I can see. Maybe someone can tell me it is?


Anyway, here's a pix of the board with the 3 4uf caps installed.
 

·
Registered
Joined
·
2,861 Posts
For your input board, Here's a pic. This is a board I haven't modded yet so you'll have to excuse the poor diagramming I did. The 2 big ones are the PS caps. All the ones circled in blue are in the video chain with the exception of C2, C4, C6 which I noted as such. If I remember, these are not. The 330s are in the chain though. The small ceramics circled in red are the decoupling caps I replaced with the tants. I also paralled some of these tants to the backside of all teh RGB chain caps.All of the RGB caps we have doen were simply replaced with same value, higher voltage, better spec'd caps. I think Cary adn I both ordered exclusively from DigiKey. Hope this helps.
 

·
Registered
Joined
·
101 Posts
Discussion Starter · #19 ·
Here's the mods I did to the input board;


1. Replaced the 3 100uf 25V power decoupling caps with 330 uf 25V Nichicon (brown sleeve). These are the second row from the left in pix.


2. Replaced the 7 0.1uf ceramic power bypass caps with 0.1 uf film caps (WIMA polystyrene, red case in pix)


3. Replaced the 9 100 uf 25V signal coupling caps with 150uf 10V tantalum. (Orange drops in pix)


4. Replaced the 3 330 uf 16V main output (?) coupling caps with 330 uf 25V Nichicon (brown sleeve). These are the first row from the left in pix. Underneath, I bypassed these with 0.22 uf 16V PEM film surface mount caps.


My reasoning;


1. Had the 330 uf on hand with the 105C rating, better cap than the old 100uf 85C parts.


2. The film caps are well known to provide better power bypass performance than the ceramic parts.


3. I dont believe the interstage caps see 10V or more across them; the tantalum part is a better cap than the 100 uf aluminum electrolytic.


4. I couldnt see where the three RGB signals exited the module, except through these 330 uf parts and the relays next to them. So they got replaced with 105C parts and bypassed with a film cap.


I also replaced the two main decoupling caps, such that they arent standing straight up off the board, so I can slide it in w/o taking off the side panel. These are also bypassed underneath with 0.22uf 16V PEM film caps.


The image seems overall crisper, for both the VB50HRTV and the PC. I am still lacking optical setup perfection, so, the effects of these mods, while perceptable, are being masked by the other problem.
 

·
Registered
Joined
·
101 Posts
Discussion Starter · #20 ·
Here's the film caps bypass location on the bottom of the PCB.
 
1 - 20 of 25 Posts
Status
Not open for further replies.
Top