AVS Forum banner
Status
Not open for further replies.
1 - 20 of 77 Posts

·
Registered
Joined
·
3,902 Posts
Discussion Starter · #1 ·
I don't know why I didn't think of this before...to many other things on my mind I suppose.

I few nights ago I blew away my XG's setup and started over from scratch, while I had the Hbox off, I keep thinking I had to get started on some hot air evac redesign I'd been thinking about over the last 2 or 3 months...basically it involved connecting my Evac ducting right up to the XG's side fan holes. I was thinking this would cause all the cooling air to be pulled through the pj, not around it.

Then last night while I had the Hbox down it suddenly hit me what was really going on with the XG and it's cooling system... MOST of the heat is created up in the top of the pj (as it sits on the ceiling)...on the neck boards and STK chips, all the pj's heat sinks sit at the top and convection cool out the screens.

THEN IT HIT ME! The Marquee I had here did basically the same thing, it huge rear heatsink convection cooled and I was able to take it from something that was so hot you almost couldn't touch it, to something that stayed at room temp no matter how long it ran...hummmm

Then I got it!

I"m building a cardboard plenum that will fit over all the convection holes in the top of the chassis...up were all the STK chips and neck boards are.

At this time I have 4 temp probes in place, one each in the front cooling fins of the F & C drive's STK heatsink, one in the blue gun's neck board, and one in the green gun's neck board (red should be same as blue)

Later today I will get baseline temps by running unit 2 hours in it's standard config.

Then I'll open up the Hbox, install the plenum, and turn the XG's side exhaust fans around to blow into the chassis. (everything must now work in the same direction)

My bet is the heatsink temps drop 30 to 40 degrees.
 

·
Registered
Joined
·
2,124 Posts
Kenny

Interesting.

I am @ work now so the PJ is not in front of me but if I remember correctly those convection holes are quite small.

Do you think you will get enough airflow through the holes.

Can't wait to see you numbers.

Steve
 

·
Registered
Joined
·
4,916 Posts
Kenny,


You do keep me thinking ALL the time.

I have been working on a very small hush box design that is just a bit larger than the projector and attaches to the projector. This is for a ceiling mount application.

I have taken a close look (somewhat....) at the airflow design of the XG and concluded this (correct me if you see something else as I am just thinking out loud)

On a table mount cool air is being pulled into the PJ from the front in the openings by the CRT's, in front of the power supplies, and from the grated opening/holes and other openings on the bottom of the chassis. The power supplies have their own fans which are exacuating their hot air to the chassis. Just below the CRT necks are three fans which blow air from the SDK area to the CRT neck/focus coil/yoke. The air is then being sucked out of the chassis via the fans in the side of the unit. So we essentially have airflow being controlled by these side fans. On the grated opening the bottom of the XG, they are located directly below the neck boards. As you know there are metal shrouds that cover the neck boards and the side fans are in line with this opening. So this opening is functioning as a cool air input for the neck boards even if you ceiling mount the PJ. The holes on the bottom of the chassis also appear to be air inlets for the SDK heat sinks, but I have a feeling that they also allow convection when ceiling mounted by virtue of the SDK heat sinks location. I also see that there are a series of holes in the chassis bottom below (or above) the SDK heat sinks.

What dies this all mean??

My gut feeling is that turning the fans on the side around will disturb the intended airflow from the three fans below the CRT's as well as the cooling of the neck boards. That is unless your plenum has some neg. pressure (which I assume you will be developing fo your exhause fan), In which case, what I might suggest is that the plenum only cover the holes and not the grated opening. This should suck air in the grated opening, directly over the SDK heat sinks and out your plenum. That way you can leave the side fans the way they are thereby not disturbing the airflow pattern they set up and not trying to create a positive pressure and turbulating the airflow pattern they have.

In fact, what might be the ticket is to cover all the round holes on the bottom of the chassis and put your plenum over the square openings that expose the ends of the SDK heat sinks, create a neg pressure there which will force cool air in the grated opening and over the heat sinks exiting your plenum.

Just a thought.


Terry
 

·
Registered
Joined
·
3,902 Posts
Discussion Starter · #4 ·
Hey Terry, looks like we both see the XG's cooling system the same way.

The way I see it though, if the pj is on the ceiling, then all the STK/neck brd cooling is done by convection...I get the feeling that the side exhaust fans would be moving mostly air from areas close to the red & blue neck brds, and air coming up from the access door area...and heatsinks at the chassis top (if on the ceiling) are somewhat enclosed by their own extruded sides!


Here is what I did and the temp changes...not quite the 30 or 40 degrees I had hoped for, but substantial anyway.

First I took baseline temps at the center of the tube cavity, inbetween the extruded fins of both the C & F drive heatsinks, and the green and blue neck boards. (figuring the red will mirror the blue)


Here's what I recorded before and after mod's (I will explain mod's afterward)

....................stock, evac @ back of Hbox............plenum installed

Tube cavity........................104........................... .............91

STK heatsink "C" dr...........127......................................102

STK heatsink "F" dr...........131.......................................105

Neck brd green.................131................................... ......92

Neck brd blue....................134................................. .......91


As you can see temp were dropped quite a bit...in fact, I think you could build a evac system and remove seven of the fans, and STILL cool better than the stock system. I believe the tube, side exhaust, green neck board fans could be removed. (I've already removed the fan between the red & green lenses.)


To get this drop in temps, I built a plenum that covered ALL of the openings in the top of the chassis (as it sits on the ceiling) The factory mount was used as the front part of the plenum...the cardboard section simply butts against the back of the mount, and is duct taped to it, then the front open area of the mount is taped off. This plenum comes back to the 4" evac duct, everything is duck taped to the chassis, and evac duct.

I then turned all three side exhaust fans around to blow inside the pj, this keeps everything flowing in the same direction...air blows inside the chassis, where it is sucked out the top.

I will know over the next couple days, but with operating temp this low, it may have the same "instant convergence" as the Marquee did once I got it's heatsink temp down close to room temp.
 

·
Registered
Joined
·
4,916 Posts
Kenny,

WOW, that is fantastic!! what a temp drop.

After I wrote the above I did some looking and wondering about the removal of all the fans except for the P.S. ones with a good suction on the plemum. Could this mean no hush box required??

So what we are going to have to do to insure that a system like this works, is to "fake" out the fan feedback. Or maybe better yet provide fan feedback based on some thermo couples placed at stratigic locations to monitor internal temps.

I like where this is headed.


Terry
 

·
Registered
Joined
·
2,124 Posts
Kenny and Terry

You guys are obviously far ahead of me on this. I just do not have the time with my wife and kids to do all that much experimenting.

I do have an idea for you to consider however.

If you want to confirm the airflow of a stock XG I think you could use some kind of colored air (dry ice maybe) so you could see the air movement. I know any kind of vapor is a no no with these units but I figure you would only need to apply it for a very short time.

Just a thought

Kenny

I am curious as to why you are doing this. I thought your foam hushbox was already real effective and the temps are already low. Not that I am complaining. After all we all benefit in the end if you find a better way and share your findings. I remember the filters when I had my Barco.

Thanks

Steve
 

·
Registered
Joined
·
4,916 Posts
Steve,

I'll tell you what I find fasinating about this, the idea of a relatively simple way to quiet down the XG without the use of a hush box. The XG is a relatively nice looking CRT that has a consumer look to it. It is also very small, something you loose with the hush box. Painted black it lookes very nice on the ceiling.

Kennt I would probibly also leave the fan in the heat sink if the HDef board as it doesn't look to be in an area of good airflow.

Another possible idea if you don't want to route the plenum chamber out via a tube (which really does get rid of all the noise). Two or three nice big quiet fans on top of it at slow speed similiar to the Barco 1208 fans which are very quiet. Heck the entire cooling system of a 1208 is just 4 fans.


Steve,

I just used my cigar to confirm airflow.....lol


Terry
 

·
Registered
Joined
·
3,902 Posts
Discussion Starter · #8 ·
Exactly, think about this for a minute, I did this exact thing with the Ampro 3600 that was hanging on my ceiling, I was able to take all the fans out except the small power supply one. A 70 cfm bathroom fan pulling through 20 of 4". Temp drop were just about exactly like this (focus brd from 145 to 115 etc.) I need to measure the signal voltage on the fan's yellow wire, We could then start cheating the fan logic a couple at a time.

It might be possible to even remove the fan on the Hdef board, what about a 1" hose mounted to it. THis hose could run up to a coupling in the chassis, inside the plenum. IF more suction was needed, the inch hose could run through the plenum into the 4" duct...that would create a decent air flow in that hose.

So what I'm thinking is, you drill a 5" hole in the ceiling in a predetermined spot, in relationship to pj's ceiling mount. Install a tin plenum to the chassis top/ bottom, install the pj, hook up 4" evac tubing to flange on plenum, chaulk ceiling around evac tubing/ plenum. ( find rope dipped in joint compound works good)


Steve, the only reason I do these things is because I like to see if I can improve this display system in some way, keeping the electronics 15 to 20 degrees cooler makes every heat producing item last that much longer, making the whole system more reliable.


One bummer, with this overall temp drop, something has happen to my red gun's flair. The astig is still good, a nice round dot, but when the red and green dots are overlayed and refocused, I notice the flair has changed. It's not a major misadjustment, but for reason I realize something was wrong watching a movie earlier.

At first I thought it was focus, I untaped the red lenses, reset the focus pot, but something just didn't look as sharp as it had the two days before this cooling system rework.

It's a job for anouther day, I going to watch Lemans now!


...Does it seem reasonable that the red flair could make this minor change because of the temp drop? :confused:
 

·
Registered
Joined
·
2,124 Posts
Terry

I understand completely and yes I agree it is a beautiful machine.

Since you are American I hope it was not a Cuban. LOL

Steve
 

·
Registered
Joined
·
796 Posts
All that work on finishing a hush box and you guys find a way to do it better without one!


Any chance you could post a picture or two?
 

·
Registered
Joined
·
4,916 Posts
It will be a couple of weeks before I do mine Mark.


Kenny, I have been looking for a solution to the G70 which has three large loud fans in it just after the neck boards that blow the air out a confined area in the back. No cooling slots on the bottom or sides. This vacuum method might be just the ticket for the G70 noise problem.


What type fan are you using. Is it one of the bathroom exhaust fans??


Terry
 

·
Registered
Joined
·
2,124 Posts
Guys

I was inspired by this last night so after the wife and kids went to bed I had a closer look.

Note: I will always refer to top and bottom when viewing PJ in floor mount .

I REALLY like Kennys idea but I do have a concern similar to Terrys.

The 3 fans taking air off the STK chips and blowing it over the tube necks (not neck boards) will be working against the airflow if you extract from the all the holes in the bottom of the PJ.

It looks to me like (in the stock config) the PJ is blowing air out the sides and the top as it sits on the floor. The air comiing out the top is greatly impeded by the H-Def board and I think this is a shortcoming in the stock config. If you reverse the SIDE fans it will blow air out the bottom through the grill under the neck boards and the top over the CRT necks unless the direction of the 3 fans over the CRTs are reversed as well. I worry that if these three fans are reversed there may be less airflow since much of the convection holes on the bottom of the PJ are covered by heatsinks.

Time to go get some more probes and experiment I guess.
 

·
Registered
Joined
·
3,902 Posts
Discussion Starter · #13 ·
Looking at those three fans over the tubes, I don't think they actually pull air from the STK chassis openings, they are simply not positioned right for that...all they do is recycle tube cavity air...I believe they do nothing more than move air over the tube's def coils. I agree it would be better to reverse their direction of flow...however I'm not in a mood to remove my tubes to do it right now.


Terry, sounds like the G70 would be an easy mod...if all it's fans move air to the rear, then a plenum could be built that mounts to the rear and simply pulls air from the front to the back.

Yes I use a bathroom fan. At this time I'm up to the 110 cfm model. The 4" tubing I use is that "expanding aluminum stuff, it starts out about 2 ft long, then you pull it out to 8 ft sections. It's alot smoother and stronger than the vinyl stuff.
 

·
Registered
Joined
·
2,124 Posts
Firstly I would like to thank Kenny for giving me the ideas. I am essentially following his procedure but just taking it a little furthur. The beauty of it for me is the 500cfm fans I am using are cheap and supply more airflow than the fans most of the guys are using. The only drawback is they are noisy. I have solved that by putting them in another room with my HTPC and other noisy components.

Here is what I'm gonna do.

1) Remove current hushbox

2) Extensive temp monitoring in stock config to get base temps

3) Build 2 plenums. 1 like Kennys for the top of the PJ (asssuming ceiling mount) covering all grills and the ambient "holes." Another plenum to cover the 3 side mounted fans currently pulling air from the neck boards and out of the PJ. This plenum will be a split deal and join above the PJ near my joists.

4) Attach my current 500cfm EXTRACT fan and ducting to plenum #1 over the grills @ the top of the PJ. Run another length of ducting parrallel to my current ducting to another room and attach another 500cfm fan. Attach this new fan in a blow config to the new ducting which will attach to plenum #2 and force air into the PJ where the current side fans are.

4) Remove all 3 side fans (to allow forced air in from duct fan #2) and disconnect the other 3 def coil fans.

5) Extensive temp monitoring on this almost silent hushbox less design and deciede if any more fans can be turned off/removed.


My current fan (as stated in an earlier post) is at the end of 20 feet of 4"dia rigid venting in an adjacent room. The new fan and ducting will run parrallel to it and push air into the PJ. I had thought of making a closed circuit design using one fan but am afraid it may heat the air too much and lose too much pressure due to the lengths involved.

I think the real challenge will be the bulding of plenum #2 as it needs to attach to the side of the PJ (@ the extract fan locations) and join above the PJ @ the joists. Since all the covers will be on the PJ it will be a challenge to seal these plenums to the PJ and stil have it look decent. Good thing I've got some real talented sheet metal pals @ work. The rest should be easy.

Once I get this working and assuming the temps are acceptable I will try to install some smaller INTERNAL ducting with the goal of removing the H-Def fan and the green neck board fan. That would leave only the power supply fans and should be real quiet. I figure more silent than a Barco 800.

If anyone has any other suggestions/ideas on how to improve this let me know.


Now to find the time.

steve
 

·
Registered
Joined
·
3,902 Posts
Discussion Starter · #15 ·
YEAH!!! Now there's a great idea!

With that much air flow you should be able to take an 1 1/2" tube out of the "pressurized" delivery air, and route it down to the H-Def fan, thereby eliminating it.

Also, You'd have to put a temp gauge on it, but I think you'll also be able to completely remove the green neck board's fan without a worry...as I recall, there is a good sized opening in the chassis directly in line with that neck board's heatsink. Your vacuum plenum will cover this opening and cool air will be pulled from below heatsink, straight past it.
 

·
Registered
Joined
·
4,916 Posts
GO for it Steve, and let us know what the results are.

I think that with a 500cfm fac on the suction side there won't be that great a drop over a 20' run, so I am not convinced that you need to put any kind of a positive pressure circuit on the input sides.

I won't be able to work on mine for a few weeks but will be very curious to see what the outcome of this is.

I suspect the reason we haven't heard from Kenny today is he is working on his.


Terry
 

·
Registered
Joined
·
3,902 Posts
Discussion Starter · #17 ·
No, actually, I hurt my back yesterday, so I was in watching a movie!

One thing you will have to do, get a reading of the signal voltage on one of the fan's yellow wires. Then we'll need to find a bunch of those female fan connectors and install the correct resistor from the fan's pin to the signal pin. Save all the original fan's intact, that way if you ever sell it, you can put it back to factory config.
 

·
Registered
Joined
·
4,916 Posts
GO for it Steve, and let us know what the results are.

I think that with a 500cfm fac on the suction side there won't be that great a drop over a 20' run, so I am not convinced that you need to put any kind of a positive pressure circuit on the input sides.

I won't be able to work on mine for a few weeks but will be very curious to see what the outcome of this is.

I suspect the reason we haven't heard from Kenny today is he is working on his.


Terry
 

·
Registered
Joined
·
4,916 Posts
Man I don't get it. It's like I'm stuttering....


Kenny, could this possibly me the solution to the noise of the 1292. I have a few and when I get back into town I think I will take a look at one.


Terry
 
1 - 20 of 77 Posts
Status
Not open for further replies.
Top