This is one case where you don't want an "innie" or an "outie" . You want it to point straight up.

Or in the case of a verticle build, to the side.

Thanks guys. I thought side wuz best but figured I might as well axe ya. Once again too lazy to search.

Just made up a CAD drawing for my OSD Panel cover.

Click to view attachment

Bites tongue and refuses to go there! Yeah, a bit OT, but when are such things completely OT?!?!?!

and....THAT'S what she said!

Ummm . . . SIM & NN . . . maybe others?: What's the best guess at how far I'll want my arc/plano distance with my 4.5x6.5 precon? I think I've got a handle on it but never hurts to ask. I'm modifying my lightbox now for the 575HMI.

Here's where I currently stand:

Using a couple of ceramic electric fence insulators for holders. Turned some brass pieces last night to mate the bulb nicely to the insulators. You notice the crimp connectors I intend to use for the lead wires already clamped within the lamp ends.

Click to view attachment

Click to view attachment

I'll J-B Weld some inserts and springs into the arse of the insulators (electrically isolated from the "hot" lamp ends, of course) to make them index and slide nicely.

Click to view attachment

@AV: I've obtained a current clamp. Do I want A = P/V or 575/95 = 6.05A? Dat right?

The J-B Weld sets up much faster in a warm oven.

The gizmos ride nicely on the threaded rod.

Click to view attachment

When it rains, it pours!!!

The caps are here!!! The caps are here!!!

Click to view attachment

Whoa! They must still be charged up a lot. That's why the pic is so funky.

Well this place had some specks.

HMI 575
HMI575
Average Life:
750 Hours
Color Temperature:
6000 Kelvin
Luminous Flux (Brightness):
49,000 Lumens
Base Type:
SFc 10-4
Watts:
575 Watts
Rated Voltage:
95 Volts
Amps:
7 Amps
Max Overall Length (MOL):
136 mm
Mounting Width:
115 mm
Diameter:
19 mm
Arc Length:
7 mm
Hot Restart:
Yes
Luminous Effeciency:
85%
Color Rendering Index:
90
Cooling:
Convection
Type of Current:
AC

So one mans 6A is another mans 7A....

So lets check the math.... 95V X 7A = 665 watts? Well it should be bright!

The current and voltage reading they give are adjusted for what a non true RMS meter might read. It would be nice if they would state that somewhere.

In the fine print on one of my lamps, I found a clue. They say that they "adjusted" the power reading by .86% for the power factor. (I have had a few conversations on the power factor question but I will leave it be for now)

so let try that.

7A X .86 = 6.02 95V X .86= 81.7 so the power would be 491.8 ....... too low.

Perhaps we should take the stock wattage #.

665W X .86 =571.9 watts.

Now that better. Just make up numbers until it feels right!

It also works if you only adjust for the current reading.

7A X .86 = 6.02 x 95V = 571.9 watts.

The problem here is real power V what the meter reads. Power is the amount of energy under the curve and the shape of the curve is not a perfect sine wave for a ballast. This non sine wave makes the meter read high in most cases. Without a true RMS meter it is hard to know the exact voltage or current.

Current in the meter is just a voltage reading taken across a known resistance. So if the wave form has a peak in it and the meter is reading the peak the all the readings will be off.

Reading the input current will give a slightly more accurate reading but the current wave form on the inpit will look a lot like the output.

I'm glad I cleared this up.

Great picture BTW. I feel the need to steel it.

Yeah, me too.


Steal away. I can do the same for any pic you may have.

Well, I'll get some readings although I won't have any idea what to make of them now.

Edit: The gear I'm borrowing is good stuff and should read true RMS. I'll get current readings plus potential across the lamp for a few cap combos as we discussed. Switch out the secondary ballast and add/remove caps. Sound good? I'll be testing within my actual lightbox with fans running. You indicated cooling could have some effect but I don't recall your exact points on that matter. I suppose I'll also allow the lamp to undergo some burn-in prior to initiating my tests.

<Yeah, right. As if I ever do anything the right way!>

Tell you what, I heard you can boost the kick out of one those by placing your tongue across it's points. Oh, and take some pics while your at it

On a serious note D, looking forward to seeing you break the ANSI barrier with the new light engine....well, hopefully within my lifetime

Slow work slotting out some aluminum angle for lamp adjustment. Boy, wouldn't a DIY CNC be nice?

Some of the odd slots and holes are due to recycling this stuff. Used a piece for the retro ceramic boondoggle (spelling?).

Adhered the homemade brass lamp end dongles to the insulators so as not to drop the dang things onto my pro reflector.

Used Blue RTV BTW. Handles 600°F. Guess that'll be cool. (Is that a pun?) I wonder what temperature "vanilla" RTV can handle???

Click to view attachment

Click to view attachment

Gonna try real hard to fire the 575 before the end of this coming weekend.

Good night Gracie.

All the meters I have seen that are true RMS will say so on them. It's a big selling point. If they don't say True RMS they aren't.

The lamp temp thing was about guessing the power level with only a voltage reading. To do that you want to have the lamp reading be as close to real life as you can. If you take current readings then the current should be fairly constant even if the lamp voltage changes a little.

Do you have a real amp probe or inductive amp meter or just the in meter amp probe? The in the meter amp probe is limited as to the amount of time you can take a reading for a given current for higher current values.

I would guess the even a Fluke meter's amp function will be around 60 to 120 seconds for a 6A load.

Stealing picture now to hang in the overclockers thread.

I have a Fieldpiece HS26 with an ACH Current Clamp head. 300 Amp A/C rating!

Manual here: http://www.fieldpiece.com/Manuals/Opman%20HS26.pdf

The manual states "Conversion type: Average measuring, rms indicating (sine wave).


Working piecemeal on getting the 575 rigged. Very busy weekend as all of them seem to turn out to be.

At least I managed to get the 575 in my lightbox. Let's see what else the weekend provides.

Click to view attachment

Dang. Installing second ballast and rewiring stuff. The ignitor is labelled as follows:

HPS400-3A, 90-100V lamps, HPS or MH 70W thru 250W or ANSI 100V HPS lamps

So, is this gonna be robust enough or would I best simply go with a new ignitor from the start?

I guess if I get far enough I'll have a shot with it but would like to source a "proper" ignitor.

Hold up, your installing a second ignitor? I thought your first was already rated 400w and all your adding is just a second ballast.

Ach! Lemme clarify. I bought an old S51 from AV awhile back. The ignitor on that one isn't clearly labelled and is kinda ratty lookin' and "shaky" - the core rattles a bit. It's an L1501 H4 from Advance. I'll halfta look up the specs on it. The one with my other S51 is brand spankin' new and I just figured I'd use it instead of the snotty lookin' one.

I'll only be using one ignitor as shown in the diagram up there a ways.

Sorry for the confusion. I do remember Natural Newbie experiencing an ignitor problem. Gotta check dat too.

I just noticed the way this "new" ignitor was labelled when starting to install it and thought "Whoa! Maybe these stock S51-packaged ignitors really ain't the deal for a 575HMI."

Going to bed. I'm beat.

I like the lamp mount, make sure that you leave some end play for lamp expansion, they grow more than you think.

The meter you have is a "normal" meter which takes .707 of the peak voltage and calls that RMS power. It is a nice meter but not true RMS. I would expect the current readings to be about 5% to 10% high. A long time ago I compared two meters one non RMS and my newer one with true RMS and measured this difference. From the overclocking thread.

" I ran the lamp for ten minuets and got the following readings.

Lamp voltage was 93.46v RMS voltage drop across a .4 ohm load was 4.546v RMS


The old reading with the non RMS meters was, lamp 103V "


This was the difference of the two meters I had, one RMS and the 103V reading was the non RMS reading.

I didn't test the current function because the current I was reading was out of range , 13A for a 10A meter.

I would expect your meter to read a little high too.

I feel kinda bad for selling a "ratty" starter (try it I bet it works), I do have a extra if you need one.

I don't know why you think the S51 starter's are not up to the job. They can either make enough voltage to jump the gap or they can't. My 1200 lamp has always started in about two seconds. Re-strike time can be a little long of about 5 min or so.

A 100,000 volt igniter will re-strike instantly but the ballast need to have thicker insulation to block the voltage or a large blocking diode.

The starter should work fine. I think NN had a bad igniter and all was well with the second one.

Just add sleep

@AV: Ain't griping about the ballast and stuff you sold me - it's just that at "face value" the ignitor isn't as awe-inspiring as the brand new one. It could very well be it's ten times the ignitor the "nice" looking one is. That's frequently the case anyway. Something looks like crap but was built a few years ago and built way better than some of the junk being manufactured these days. Damn, I'm sounding more and more like an old man each day.

Okay, so the meter isn't true RMS. Oh well.

The reason I was concerned about the ignitor is it is clearly NOT rated for the type of lamp I'm running but I absolutely take your word for it that all should be well. Thanks AV!

And no worries on having sufficient thermal expansion allowance on the lamp mounts. There's like a centimeter of travel on the springs holding it in place.

The wiring just keeps getting messier and messier.

Click to view attachment

But, it lives.

Click to view attachment

Click to view attachment

Click to view attachment

Obviously I've got some adjusting to do. Comments after this brief test:

1. The 575 is HOT!!!!! Can't believe what that extra 125W does to my cooling!
2. Lots of stuff to "clean up". Didn't soft mount the extra ballast and I can hear the buzz.
3. The clarity improvement is immediately apparent. Image is SHARP.
4. Damn nVidia Dualview is burning my arse. Won't behave at all tonight.
5. My fan controller causes wa-was with my fans when in less than full speed mode. They speed up and slow down instead of just running slow.
6. Still seeing that pulsing of my display. Now I want to see if it tracks with the fan wa-was.
7. The 575's are like little suns. Amazing little devices.

@AV: I just managed to fire this before I had to take my neighbors meter back. Got 5.7 and 6.6 amps (crap - that's if I recall correctly) between low and hi modes. That's with an exra 6uF on the hi ballast. Lamp didn't look pink but I was a tad skitterish when I saw 6.6 as I was expecting just around 6 or slightly over. Didn't get any voltage readings.

So, it works and I think it's gonna turn out great but I've still got a hell of a haul ahead of me.

I'll borrow the meter again when I'm able.

Later and thanks to all for assistance . . .

Your using a 4.5"dia x 6.5"fl precon with a 220mm fresz, aren't you?

Nope, a 330mm from 3DLens. Be aware I made no adjustments as I had no time. This was just a test of whether the bulb would fire and burn with my dual S51's and also I wanted to get a handle on how much current I was pumping through the lamp. I just barely had enough time to do the test before I had to return the current clamp.

I used DAZZ's pre-con calc and the one from RymGB and came up with distances for arc to precon to 330 but obviously I'm W-A-Y off as the light cone isn't even close to covering the fresnel.

I'll have to revisit all that stuff and see how good I can do.

Guess what? IT'LL BE AWHILE!!!!

Well!! .....*interlocks fingers and cracks his knuckles*....this might be fun. Speculating appropriate distances from my experience and others - for a 15.4", 330mm fresz & 4.5" precon - I'll say 280mm from rear fresz to plano of precon...and about 35-40mm from edge of your HMI lamp to plano. Just to be sure though, is your precon crown glass? If so, hope it's loosely fitted and that your cooling is up to par.

The lamp calls for 7A (6A real current) so 6.6 should be close enough for a start. We don't know exactly how accurate the meter is with your setup so I thing 6.6 is a OK value for a non RMS meter. If it is reading 10% high that would be 5.94A which is good for a long distance relationship .

Yeah but long distance relationships just never seem to work out

I thought I heard something.


I was using 34mm and 277mm using DAZZ's sheet but wound up projecting a light cone so small that I got that screenie up there.

The cone wasn't even close to covering the rear fresnel.

Now, to be fair, I was once again rushing through the test like a chicken with its head cut off and maybe I really mussed up. Like off by an inch or something stupid like that. Maybe an inch/mm conversion error. Who knows? I'll revisit this soon.

Thanks!

Edit: The precon is borosilicate and I cool my lightbox with two ~4" 12VDC computer case fans at half voltage (in series). The heat output is so much higher than the 400W though I may bump the fan speed depending on how things work out. The box gets too hot to touch. A quick shot with my non-contact thermometer gave me >140°F and it was - IIRC - already off and cooling.


Okay, so I'll leave things where they are. At least that's the message I'm getting from your post.

Hard to believe I was lucky enough to get something right the first time though!!!!

Just to offer up evidence of how sad a state of affairs my life is in here is my detailed notepad for how to position the arc/precon/rear frezzie. Fresh out of my pocket.

Click to view attachment

And I have the audacity to wonder why I can never seem to reach the finish line.

This is bad. I'm sitting here converting mm to inches and holding my fingers apart while envisioning how far apart my components are and squinting my eyes staring up at the ceiling. People are starting to talk - even more than they usually do!

Funny, you just described how I normally look while taking a 'dump'

One word - FIBER!!!

Back to the grind.

Plugged holes in lightbox made from all my damn changes. Just used the Blue RTV. Mostly just 1/8" holes from drilled out rivets.

Readjusted the lamp and reflector to get closer to where they should have been. I was quite off in my rushed measurements. Arc >1 centimeter farther from lamp than it should have been. Hard to measure the way I was doing it. Can't wait to see if I'm covering the whole panel now or at least very close.

Cleaned up my wiring (and some other stuff) a bit with re-routing and zip ties.

Gotta let the RTV set up. Should fire it up again in the next few days.

Either can I.
If it doesn’t then I’ll have to look back through all the equations in the spread sheet, not a fun thing to do.

DJ

You are infallible Oh Great One. Besides, no one else has shown your formulae to be bogus, have they???

I have had two things tickling my noggin:

1. Is my 4.5x6.5 really 4.5x6.5? Probably is. Not sure anyone else actually bought the same one I did.
2. SIM said "about 35-40mm from edge of your HMI lamp to plano". I interpret your stuff to be arc to plano not edge of lamp to plano. The RymGB technique of getting distances is less refined than yours (although a hell of a lot easier to plug #'s into! ) and was more like to the geometric center of the lens or some such. Anyway, both your and RymGB's formulae gave reasonably similar results taking into account their differences.

Anyway, I'll see how things are looking when next I fire the beast and post the most precise measurements between all my components that I can muster . . . eventually.

I honestly don't know why I ever started this damn project.

Here we goooooooooooooooooooooooooooooooooooo:

Readjusted lamp and reflector. Like 35 or 36mm from arc to plano (not counting rim or true cylindrical 4mm or so of the lens "bottom", this is right to the arse plane of the physical lens/glass). Like 11" (sorry for mixing units but that's what it's sposedta be - honest) from plano to rear frezzie-wezzie.

Light cone looks good - maybe even too big. The frame wuz cut for a 17". I'm now using a 15.4". So, those black lines I drew designate approximately where the 15.4" really is above that fresnel sandwich. (Sandwich? Yum!)

Click to view attachment

Hadda just plug the damn 15.4" alone since I couldn't for the life of me get my nVidia card to drive both panels. Damn goofy-ass software crap!!!

When I look at the panel I see this odd border - wot dat is??? Any speculations?

Click to view attachment

Sharp eyes bass-turds will have noticed that screen of death. YYYYAAAAAGGGGHHHHHH!!! I tried the repair thing it suggests but have no idea how to or even if I can fix that problem. Again, any help out there? I can just do a whole reinstall I guess. The fun NEVER ends.

Click to view attachment

Okay, how about one more. Here's the arc image at the triplet and four inches closer to the FS mirror. Guess which is which. Yep, the sharp one is closer to the mirror. Why? What am I missing here?

Click to view attachment

Forgot to add that I have that bloody "hot spot" right in the middle like I think sensibull battled. Gotta go see what he did. Or was it Hirudin? A search will tell me but not tonight since I'm tired and a bit pissy over this. And I also see a distinct ring - almost like fresnel rings around the hot spot. Sorry some of those pics were with my crappy-cam but I did break out the good cam for the triplet arc images. They were shot with the same manual settings BTW. Don't make much of the color though as I made no attempt to get that right and I did a PhotoShop adjustment of the pics before posting.

Later all . . .

Edit: Some glowies too but that's no worry AFAIC. The edge light you'll notice in some of the panel pics IS blocked with a flat black border just not right at the screen surface. Also still having that odd bleeding effect you may have noticed me posting about in another topic. Oh, it'll be awhile until I resolve all this stuff.

Whoops! Forgot to post this one and wouldn't have been able to sleep. Tomorrow's a busy day.

Click to view attachment

The hot spot will get less and less the better you get the lamp adjusted.

I would move the light box just slightly further away from the fresnel and then move the lamp closer to the condenser.

Just think back to the Windows 98 days when you had to re-install every six months just for fun.

The strange purple framing on the LCD I have no idea.

I did this to get the last of the hot spot gone, the final black spot was larger.

http://www.lumenlab.com/forums/uploads/pos...-1160431689.jpg

If that's Windows XP, you can use this guide for a Repair Installation:

http://pcsupport.about.com/od/operatingsys...stxprepair1.htm

A Repair installation should only repair system files and your data should still be intact. Some third-party programs may have to be re-installed as the Registry will likely be overwritten.

That bottom pic, with the very dark/brown corners looks just like how my projection was. I never tried to fix it and I pretty much just lived with it so I can't offer much of any advice on how to fix it.

As for that center hot spot, that musta been sensibull as I didn't have that.

BTW your spring loaded lamp holder is amazing!

It looks like your arc image is forming to short. A quick fix would be to move the pre-con closer to the fresnel and then move the arc closer to the pre-con.

Now let’s work out why:

Is the triplet modified?
If it is then that is your problem. The triplet could have a 50mm longer FL.

If you haven’t modified it then:
What are all your current distances and FLs?

Triplet FL =500,
Rear fresnel FL=220
Front fresnel FL = 650 (600)
LCD= 15.4
Pre-con=4.5” *6.5”

Arc to plano=?
Plano to rear fresnel=?
Rear fresnel to front fresnel=?
Front fresnel to LCD=?
LCD to triplet centre=?
Triplet centre to screen=?

bleh


Amazing? I'll tell you what's amazing. The fact I haven't blown this stuff all up with a keg of black powder in frustration!!!

Joking!!! I'll get it right someday. Thanks for the kind words on the lamp holder. I think it's <said like a haughty Brit> Rahthah eleghant.

@Quasi: Thanks for that info!

@AV: I'm thinking today maybe the border is nothing to worry over. Maybe just some crazy artifact of my particular arrangements. The hot spot is low on my priority list as I figured it'll get better with other adjustments. But I'll keep your trick in mind and I did remember it. Nice job BTW. Nice touch with the paint.

I'll getcha bettah numbahs and/or verify these above sometime soon Mum but thought I'd toss this stuff out to yah now fer you to mull over.

Thanks DAZZ.

Whaddaya make of Hirudin saying he had the brown corners like I do at present??? Ain't dat cure-ee-us?

My opinion

First of all - your pro triplet, if I understand correctly after modifying and then remodifying it, you used a 17" panel and a 400w lamp previously and produced some good images with no problems whatsoever. So, I doubt you misplaced the trip's lens. What I am concerned about is if the longer fl optics could be intensifying the HMI's arc somehow. I say this cause sensibull used a 22 Beseler (560mm) w/600-650mm fres, your using the pro (500mm) w/600-650mm fres and I'm using the 18" Beseler (457mm) w/550mm fres. My 'spot' isn't as distinct as compared to both you and sensi's.

The glow spot..........this may be a quick fix for you and I don't know how it will appear on your screen but if you turn the HMI lamp wherein the 'nipple' on the arc globe faces the precon directly you might see it deminish (unless your lamp is different from mine). In my trial with an 18" Beseller, again, the spot isn't as sharp and defined as yours or sensibull's but wider and faded. It disappeared when I turned the nipple around but it left a slight visual tainted halo on center...and....the lux numbers dropped . But give her a try if you want to see what I mean. Placing the 'nipple' behind towards the reflector created a more wider spot with another ring of brighter light and lux was boosted. Just so there is no misunderstanding, I'm not using the pro reflector but the Proxima's (@ 50mm dia) if that has anything to do with the use of the HMI's smaller arc.

The purple/blueish frame.........I'm thinking if it ain't showing up on screen, I wouldn't worry about it. It might just be how your 15.4" appears when the crystals grid array is powered on. I noticed on my 10.6" that one side of the LCD didn't gleam as much as compared to the opposite side (left to right) no matter how much I polished the AG with with 'Mother's Mag Polish' but on screen it was fine (that is when the LCD is on. When off I saw a shade difference and lux readings) AND I seriously doubt your fresnel frames were made smaller than your LCD by mistake..............................right?

Dark corners.........not to step on DAZZ's suggestion but I'd try both ways. First - I'd just move the entire light engine back about 10 to 20mm (thats 290 to 300mm for you) to see if it got any better. If not, then move the arc closer to the precon and try again. If it does get better, then move the engine closer for more gain, hopefully.

@SIM: Yeah, I agree that I've just got to knuckle down and do a bunch of fiddling all at once to see what works and what don't.

I like the nipple thing. Great idea. Real easy - well, relatively easy - for me to do as well. I ain't reaching into that lightbox while she's a-burnin' though!!! Just cooling down it looks scary enough. And even AV related to me that once you see one of these 575's energized it makes one wonder why they don't just explode after reaching full intensity and I wholeheartedly agree. Like I said up-ploD aways - they're amazing little devices. Truly an engineering marvel even though they pale by comparison to say . . . oh . . . the Space Shuttle!!!!!

My frezzie frames should be oversized enough. But I'm always open to ideas. I can see into the sandwich between the field fres and LCD and it looks like the lighting is fine but I don't think that's really a good way to do it - by eye I mean. I have given some thought to opening up my frames a bit larger - but unless I can convince myself it's necessary I doubt I will. I can increase the field to LCD distance and watch the shadow and that should tell me what I need to know regarding whether the frame windows are large enough.

Maybe I've not seen others say this or maybe I'm just not remembering seeing it - old brain and all that - but I'm still amazed at the qualitative differences in sharpness of the projected images produced by a "small" arc lamp versus a "large" arc lamp. You've done extensive experimentation with "large" arc lamps - how do you feel about there being a clear difference?

I know Supra says that even with a "large" arc lamp that one can achieve great results and I tend to fully agree based on my limited experience and what I've seen here but I still maintain superior results may be realized with a "small" arc lamp no matter how "tweaked" a "large" arc setup is - the only difference between two setups being the lamp itself.

I have a 2.5" sony xga lcd from a camera with the same blue border around it - that screen has taken quite the pounding though...

The blue border has to be a shadow of sorts. Look how diffuse the edge is. Just seemed odd. Maybe it is telling me something though.

@DAZZ: The LCD to trippie distance is around 23 incheroonies. Ain't tried adjustments yet. Perhaps this weekend.

Did a fresh windoze install last night. Sigh . . .

Made my computer pull an all-nighter to download Service Pack 2. Even decent DSL speeds can be painful when you've got Bill Gates in the equation.

Ouch. SP2 by download... Then all of the Windoze updates that come on AFTER SP2... I feel for ya, man. (I have SP2 on CD, though I just updated my distro CD, and slipstreamed SP2 and MOST of the updates since then onto a CD-R. Naw, I never have to reinstall my PCs... never.)

[MS rant]
Personal pet peeve with Windows... I have a lot of portable hard drives, and every time I update the actual drive in the portable case, Windows registers it as a hardware change. So when I started getting rid of 200G drives for 320G, then 250G for 320G, then going to 500G... Anyway, after a few of these changes, Windows insists on needing to be reactivated, and the registry gets all f'ed up, so I have to reinstall periodically. Then of course, I have to register manually, since my install key has been used before.

I'm just happy that I have only had one WGA event so far.
[/MS rant]

Blue/purple border...

I wonder if this is an artefact from your fresnels, but then I'd expect a red/orange effect at the outside. Light is refracted from a given lens according to it's energy level. Thus blue light is refracted more than red light by the fresnel lens and precondensers. This SHOULD result in a greater concentration of blue light to the center of the LCD than to the edges.

Unless... Somehow the blue light is being refracted ENOUGH more that it's actually at a greater concentration towards the edges of the LCD, but that would take some real funky alignments.

I'm curious though if the blue/purple shadow is consistently through the viewing angle. Basically, does it project, or is the decreased green only an off-axis phenomenon?

Remember that the light that's off angle, like toward the camera position in that pic, isn't what's important, it's what gets directed towards the triplet that matters.

Interesting comment about the border. Off the top of my ignorant head I'd come back at you with this - the geometry of the whole situation doesn't settle well in my gut when I think about what you're saying. That is, the light cone is a circle and the "open" area of the fresnel/LCD sandwich is rectangular. Wouldn't that pretty much dictate I'd NOT see such an "even" border around the LCD edges if what you're getting at is actually the case? I hope that makes sense.

I think I’ll wait, gone got self confused. What was the problem again?

Here’s what I get:
Click to view attachment

Click to view attachment