In most cases, I had to remove the links from the solar-cell preamp in order to acheive Dolby level and leave any headroom.

Current was checked and was found to be at (or slightly below) the manufacturer's recommendation of 450 mA for the optical l.e.d.

As I mentioned, switching the pre-amp gains to HIGH cured the problem for now, but my concern is whether they will continue to dim.
 

We have been using/testing the visible LEDs since prior to their inception. The first ones are still going at about 9000 hours of multiplex use.

What I have found is that the biggest change occurs over the first 3000 hours and then the dimming gets more gradual.

When you put them in new, start off with the current set to minium on the LED (may sound obvious but many people did not). Go ahead and start off with the links removed on the reader board (more gain). On Component Engineering power supplies, they just started sending out the supplies so they will put out very low current at their minimum position (as in you may not be able to achieve level without raising it a bit). I only give the LEDs enough current to barely go over Dolby-Level (top red LEDs lit on a CAT. 222 for instance).

During that first 3000 hours check you Dolby Levels about every month or so...then step it down to about every 6-months.

At this point, the 15,000 hour claim looks doable for analog.

The jury is still out on switching the LEDs on and off when the projector isn't running. The is a toss up of shocking the LED over saving the intermission time of just buring the LED. If you choose to switch the LED on and off, put a follower relay on your motor circuit (it must be on BEFORE you need sound, particularly on digital) and switch the AC input to the LED power supply. This will give the LED a soft start so as not to shock it. By having the power supply come on with the motor, the electronics will have stabilized prior to the changeover opening.

Presently, the LED current needed to achieve good Dolby Digital video is higher than that for analog so I feel that the digital LEDs will not make 15,000 hours. At about 2 volts p-p on the perforation burst, the Dolby digital reader will be failure prone. At 4-4.25 volts p-p you stop benefitting from increased illumination (on the LED system). If you don't mind a little higher error number and turing the LEDs up more often, you can start them in the 3-volt range.

Steve
 

I do not like the basement reader idea at all. It takes away the possibility of moving that DA20 to another house and they just aren't very reliable. I ALWAYS recommend going with a penthouse reader and have never had one of those fail.
 

And I agree, Kelmar basement readers *suck*

Aaron

 

Why are there not LED sockets for easier replacement after the failure? Seems like this would make your lives easier.
 

It is almost always the LED itself that gradually loses brightnes ( and the current drops proportionally as well) The current can be brought up to increase the brightness, but only to a certain point before the power supply maxes out and can;t produce enough current to keep the LED at the proper brightness, or the LED itself fails.

Why not a 'socketed LED' that can be plugged right in? I cannot really say, except maybe for the fact that alignment of the LED is very critical, especially for digital, and i figure a socket will get 'sloppy' with time and become impossible to align maybe? Who knows?

The LED assemblies for Kelmars can be replaced by replaceing the arm that has the LED assemblies in them and rewiring the LED arm to the power supply, If set up properly in the first place, it should'nt bee too difficult to do, however of course you must completely realign both digital and analog for optimum performance.

SDDS LED arrays fail in a similar fashiion as well, usually after 10k hours of constant use.

Aaron

Dolby currently has a long running test going in which large quantities of the LEDs are operating continuously and having their light output measured regularly. What they are learning is that for many of the LEDs there is an initial instability which ususally means an early drop in output, but in some cases there has been an increase. This seems to settle out after about 30 hours. Following this, for most units there is a long and gradual decline in light output, but for a very few rogue pieces there is a much more rapid fall-off. So far, they have plotted each of their test LEDs for 500 hours and they are well on their way to 1000 hours. While this is a long way from the 10 to 15 thousand hours we are all hoping to enjoy, the trends are becoming clear. It seems to be Dolby's intention to try to establish criteria which will make it possible for them to cull the potential problems out before they ship.

So, what does this mean to you today? It means that you should go back and review Steve Guttag's response. The big enemy is heat, so run those LEDs at the lowest current you can and still achieve Dolby level for your analog, and 2.5 to 3 volts of video for your SRD.

I am sorry to be so long winded, but there are some other issues in this discussion that I might be able to shed some light (at 660 nm) on. First, however, PLEASE understand that my comments apply only to Component Engineering's readers. I have no right or authority to speak for anyone else's.

Our power supplies are Current regulated, therefore, the voltage has no meaning. It is whatever the LED wants. Current regulation helps to assure that the light output will stay stable longer and will compensate for the pop corn oil that gets into connections. We recommend an initial setting of about 150 mA for analog, and 400 to 450 for SRD.

We also recommend that the power supply be switched on and off with the motor. You must be muted, though, because while the power supplies do soft start, there will be a thump because the LED responds so quickly. Also, the bipolar pre-amp supply needs a second or so to stabilize. For SRD there is also the matter of establishing signal presence which Steve mentions.

The LEDs in our mounts are plugged in and can actually be removed while in place. It takes a short #1 Phillips screw driver and care, but it can be done and the alignment should be good enough for you to finish the shift with OK sound. This brings up the matter of LED alignment. The light from these LEDs is a very thin band which is quite wide, but not at all high. It is very important that you optimize the vertical adjustment because it will make a major difference in how much current you need to run.

I have rambled on long enough so I won't go into the digital stuff right now, but you are invited to check out <www.componentengineering.com> where you can find a good description of how to set these up. You can also find more on how to change the gain setting on the analog pre-amps.

------------------
Bill Purdy
Comonent Engineering
 

Is that more specifically what you meant? I'd just like to get a bit more clarification.

Thanks.
 

Since the Dolby will still be in non/sync mode when the motor is activated, I assume the "thump" caused by the l,e,d, turn-on will not be a performance problem (likewise for the point at which the motor turns off).
I assume also, that the power supply will have plenty of time to stabize during the seven or so seconds between motor start and dowser open.

Brad brings up a very good question about the habit some operators have of "inching" down the film to a cue point. Although we discourage this, old habits are hard to break.
However, since you advocate the switching of the unit off during intermission anyways, may we assume that this will not be a problem?

Once again, thak you for your input.
 

While I too deplore the practice of "bumping" the film down, most of those who do it do so with the manual motor switch whereas the LED supply can be switched by the automation's motor relay. I recognize that this isn't all neat and tidy (how do you turn the LEDs on for servicing?), but you can definitely improve the life of those rascals.

The other way to look at it is that if you take the cost of your entire sound system and derive the percentage for the cost of a replacement LED, it will probably turn out to be a most infinitesimal number. But, remembering that it is the source of all that sonic magnificance without which the rest of the system is meaningless, just replace the thing every couple of years....but watch it regularly.

------------------
Bill Purdy
Comonent Engineering
 

Of course, we could use one of the unused switches on the rear of the automation apron for this, but it means one more switch for the new trainees to remember.

The biggest problem with inching the motor on and off with the motor switch is that motors draw much more current to get from a stopped position to a running position than it does to keep it running. This practice is particularly hard on the motor and the switch (not to mention the projector's gear train which gets to "slap" against both surfaces of every gear tooth).

In most automations, the motor manual switch is rated for very occassional use (technicians screwing around, or emergency non-automation purposes). They are not designed to carry the high motor-starting current day, after day.

 

To Bill, I have a Kelmar analog reverse scan installed in my private screening room. (I was told that the CE unit wouldn't fit properly in a Century JJ with the 70mm bypass path.) Anyway...to the point. I have my power supply relayed to power up on a separate relay in sync with the douser. Of course I can switch it on for manual operation, but I almost always use the automation. Anyway, I've never had a thump or anything in the sound. It is perfectly smooth, as if the led was on all the time. Is this a CE thing or something in my particular setup?
 

The answer to your JJ question is in two parts. If one is thinking in terms of analog conversion only, there is no problem. You remove the original exciter lamp bracket, the slit lens and the photo cell bracket. The new lens/pre-amp assembly slips into the original slit lens mount and the new LED assembly fits right where the old cell bracket was. If you are doing the complete conversion including SRD, however, then things change. It has always been a basic concept that our readers can be installed without drilling or tapping any holes or otherwise cutting any metal. In order to avoid having to cut into the bulkhead isolating the exciter lamp housing, we had to drop the digital reader down farther and so were trying to read the digital track too close to where the film was leaving the sound drum. More wrap was needed around the drum. Well, as long as we are going to make this change lets go ahead a make a few more improvements to what has always been the weak part of the Century head. This meant that we had to sacrifice the 70mm bypass skid and roller. We felt the trade-off was worth it as Ballantyne was already cutting a hole in the front of the picture heads and bypassing the whole sound head, and while we didn't like it, this could easily be done in the field. Most people, however, have decided to just abandon 70mm. In any event, our new pieces sure improve on Century's original design. These pieces plus a good VKF sprocket and a change in bearing loading render this a really good sound head.

------------------
Bill Purdy
Comonent Engineering