If I only had a Brain!

[Luca Brazzi]

Well... I have a nice working DIY projector... (actually its more of a first time through testbed/prototype) It uses the Lumen Lab standard lens set, a 250 watt DE HQI bulb with an IKEA Napkin Ring Reflector, and a BenQ 567 LCD monitor...

It works great! Im happy with its performance (although I scratched my LCD/Fresnels building/rebuilding it about 50 times to get it right... oh well)

Now for the next thing...

One thing missing from our DIY projectors that Ive noticed is... for the most part... they lack some basic intelligence that commercial projectors have.

What do I mean?

Well... a commercial projector has some basic built in intelligence that allows it to "understand" a few things about its operation. First off... if you shut off a commercial projectors power... it knows to let the fans run for a while to help cool down the bulb... it also knows that if you try to power back on too soon not to turn on the bulb, etc....

Ive been considering starting a project to develop similar features for DIY projectors, but would like some input as to what other functionality would be desired?

Incorporate temp sensing maybe?
Dont allow the light to come on unless the fans are running?
Automatic fanspeed control?

Heres one... how about an auto off after X hours if there is no intervention (for those of you who fall asleep with the projector running)

[Luca Brazzi]

How about a Bulb Life counter that tells how many minutes/hours the bulb has been lit?

[Stocker]

A standard hour-meter is simple enough for the bulb life.

The fans running with the bulb should also be very simple.

As for the fans running after shutoff, that could be a simple R/C time constant and a little logic. Nothing magic here, but perhaps a cut above the average DIY job.

Go for it. I suggest starting a new thread in the appropriate forums for each question so some of your questions don't get lost in the discussion of the others.

[Luca Brazzi]

What Im trying to do is to get a holistic view of the entire system. That, I think would best occur from projector building folks...

As for the fans running after shutoff, that could be a simple R/C time constant and a little logic

True... when you think of individual functions they are each relatively simple, until they start adding up. Then trying to get a completely integrated solution becomes more difficult, and the circuitry far more complex if individual control circuits/discreet logic circuits are involved. Like the fans running after power shutoff as you said is a simple RC time constant, but what happens if you want to also ensure that your lamp doesnt come on if your fans arent running? Then add fan speed adjustment based on temp, etc.

Im thinking along the lines of a microcontroller based design with all the logic embedded. Take a microcontroller, stick on some sensors/mosfets/etc. and write the code to monitor/control everything...

[mbates14]

suggestion.

go with a pic. like a PIC16CF84 or something like that.

get some books, or online manuals and learn how to program one.

That will solve all your little needs. hour counter LCD on the side maybe, use portB to drive the relays to control the fans, ballast, yadda yadda. and have port A to drive an LCD, or temp sensors.

and use the LCD as a status display. Hell, you can find premade code to read write ports, use LCDs and all that stuff.

Im using a PIC a friend programmed for me to control my ice machine.

The thermostat and timer went bad, so we programmed a PIC to run the cycles and temperature sensors (reverse-biased diodes).

Sorry, kinda off topic.

[Luca Brazzi]

go with a pic. like a PIC16CF84 or something like that.

Yes...

Ive been programming PICS for many years...thats why Ive been thinking this could probably work... Im trying to gather suggestions on possible features based on different projector designs/experiences/etc....

I think something like this could be especially helpful in those designs (most of the ones Ive seen) that use Computer fans for cooling.... computer fans are NOTORIOUSLY unreliable if they get too much dust in them... if you embed a computer fan inside your projector where you cant see it spinning... how do you know if its really running? Maybe the bearings are shot/etc and its drawing current but not spinning (just jittering) I guess an indicator light could give you a clue but then thats yet ANOTHER dedicated circuit if the individual control circuit choice is used over a controller.

What I would like to get to is a total control system. When I push the on/off button, the controller takes over and turns on the appropriate projector subsystems if the correct parameters are met, if something go wacky... It sets off an alarm/shuts down/etc

Question: If you had a choice for a Bulb Usage timer would you rather

A) Have a timer that measured the time the projector was on
B) Have a timer that measured the time that Electricity was applied to the Ballast
C) Have a timer that measured the time that the bulb was actually producing light

[cbm5]

If someone came up with a preprogrammed microcontroller, then hey, some of the more obsessive among DIYers might invest the extra expense and effort and debugging time. But most of us realize that we're not going to get a projector with all the bells and whistles, we'll just get a cheap projector. It's easy enough to remember to turn off the fan sometime after you turn off the light, and it's no major loss other than a few cents of power to leave the light on a couple extra hours. The only thing that could possibly damage an expensive component is the fan failure that results in a damaged LCD, and there are readily available products to monitor temperature and control fans. http://www.crazypc.com/Merchant2/mer...uct_Code=5220B

If you want to go all out, then no one's going to stop you...but DIY projectors have been around for years, microcontrollers have been around longer, and we haven't really seen many people make a controller. I'd say that if it was necessary, it would already be common.

[Luca Brazzi]

Originally posted by cbm5
If someone came up with a preprogrammed microcontroller, then hey, some of the more obsessive among DIYers might invest the extra expense and effort and debugging time. But most of us realize that we're not going to get a projector with all the bells and whistles, we'll just get a cheap projector. It's easy enough to remember to turn off the fan sometime after you turn off the light, and it's no major loss other than a few cents of power to leave the light on a couple extra hours. The only thing that could possibly damage an expensive component is the fan failure that results in a damaged LCD, and there are readily available products to monitor temperature and control fans. http://www.crazypc.com/Merchant2/mer...uct_Code=5220B

If you want to go all out, then no one's going to stop you...but DIY projectors have been around for years, microcontrollers have been around longer, and we haven't really seen many people make a controller. I'd say that if it was necessary, it would already be common.

Very true...

The more I talk to folks... the more I realize.

What about this...

What if you could embed a controller into your projector that solved one simple problem?

For example... I was sitting in front of my PJ typing my last post, and listening to music on the PC I have hooked to it... Then... I realized I needed to go to the store to get cigarettes, and I said to myself... I WILL NOT leave the house if even for a minute with that projector running because who knows what will happen between the time I leave and the time I get back to turn it off? Its kinda like the IRON.... nowadays you dont have to worry about leaving your IRON on because someone embedded a timer in it so if it sits unused for X amount of time... it turns off by itself (unless you have a REALLY old IRON). Now... someone may say, "Well then just put a Timer on the Projector" and again Id say... "True.... that will work...very well... but when you start to try to add something that is more than simple on/off it gets complicated"

[Luca Brazzi]

Ive GOT IT!!!!

Thanks folks!!!

What Im going to do is design the controller to use a motion detector as the "Reset" for the projector staying on! If there is no motion in the room for 2 hours... the projector will turn off.

Of course Ill use a Pet Sensitive Alarm system motion detector... I wouldnt want the pooch walking around when Im not here to keep the projector running...

[CCMCornell]

Why not just get that Honda robot to tend to the projector while you're at it? Or, you could try B-4, Data's (from Star Trek) surviving brother. He's kinda stupid, but I think he could handle it!

But seriously, your ideas are great and it's fortunate that you have PIC programming experience. PIC programming and even embedded Linux devices are things I've always wanted to start learning to add to my DIY capabilities. Maybe you'll share some tips for us inexperienced peeps when the time comes?

[Mikey p]

Do you mean the projector light will turn off but the fans will stay running to cool it down?

[tjh]

Originally posted by Luca Brazzi
Ive GOT IT!!!!

Thanks folks!!!

What Im going to do is design the controller to use a motion detector as the "Reset" for the projector staying on! If there is no motion in the room for 2 hours... the projector will turn off.

Of course Ill use a Pet Sensitive Alarm system motion detector... I wouldnt want the pooch walking around when Im not here to keep the projector running...

Sounds like a great idea, and I like all the other ones, but what happens when your watching a movie or something? wouldn't that be detected as motion?

[Luca Brazzi]

Sounds like a great idea, and I like all the other ones, but what happens when your watching a movie or something? wouldn't that be detected as motion?

I dont know... I know that motion sensors are triggered by infrared I dont know if the PJ would produce enough infrared light to trigger one though... Ill have to do a test.

Why not just get that Honda robot to tend to the projector while you're at it? Or, you could try B-4, Data's (from Star Trek) surviving brother. He's kinda stupid, but I think he could handle it!

Dont you think that would eat into the money I save by building a DIY projector?

[Luca Brazzi]

Sounds like a great idea, and I like all the other ones, but what happens when your watching a movie or something? wouldn't that be detected as motion?

Ok... did a test...

I ripped the Motion Detector off of my Alarm System, and powered it using a 12v wall adapter. Even putting the detector right in the light path in front of the projector 2" from the projection lens didnt trigger it. Keep in mind though... I have an IR mirror inside the PJ, and Im only using a 250 watt bulb.

Maybe someone else with a different configuration has had an experience where their projector set off a motion detector?

I doubt that there would be enough concentrated IR in a projected image to look like a human beings IR body signature though...

[mbates14]

comment about the fans:

Most high power output fans, like in commercial projectors have a third yellow connection.
the Sense wire.

So if the fan isnt running, the CPU will see no RPM reading from the sensor, so guess what, you can have an errorhandler built in the pic that says, Hey, im not reporting anything from the fan. So, go into shut down, and print an error code on the LCD.

Its not hard.

When i say LCD, I mean one of those little 2x20 or 2x16 8-bit joggers.

[Luca Brazzi]

Most high power output fans, like in commercial projectors have a third yellow connection.
the Sense wire.

Good thought...

For those of us... like me... though who are using plain old computer fans... there is no sense wire, just the black and red, so either I:

A) Sense airflow in front/behind the fan

B) Sense Votage/Current running through the line... ex: Fan off=0 volts, Fan running=12 volts, Fan powered but not running=? (I guess the answer would be something other than 0 or a steady 12 volts)

I think the most reliable way is to detect whether there is airflow through the fan (fan is actually spinning), but that would probably be the most hardest to sense.

Another thing that could be done would be to place an IR detector/emitter pair on either side of the fan... then the controller could detect whether the fan was spinning/not spinning easily by detecting whether the IR signal being picked up by the detector was fluctuating. A rotating fan with this setup should output a fluctuating voltage from the detector as the fan blades interrupt/allow the IR light to pass through to the detector. A NON rotating fan would output a steady voltage from the detector.... either always ON... or always OFF depending upon whether the fan is stuck with a blade blocking, or allowing the IR light from the emitter to pass.

The question is...

Can an IR detector distiguish between the IR emitted by the emitter on the other side of the fan, and the IR generated from the lamp?

Time for another Test!

[mbates14]

well, it would be a rather complicated circuit, but with those type of fans, youll most likely pull a tad more if the rotor is locked.

aka "locked rotor current", even though it is NOT AS MUCH as a standard motor, but a brushless motor drivin by a steper ocillator, it will "still" draw more current, as since the magnet rotor is not turning, its a heavier load on the drivers.

so, lets say hypothetically, if the fan is running at 12ma. well, you can make a circuit that knows the fan runs around 11 to 13 ma.

well, when the rotor is locked "ammeter to find this". it might pull a couple more miliamps.

lets say if it runs regularly at 12ma, if you stop the blades, it might start pulling 15ma (to get it started again).

if one of the drivers shorts, or opens, itll start bobbin back and forth, and it will draw less current (one of the drivers isnt there to pull the current).

Fans fail in one of the two conditions. If the oscillator dies, and it might not pull hardly anything. you can definatly see that.

You would have to build a current sensing circuit somehow.

[mbates14]

well, if you use IR, remember now, you still need a monostable circuit, or some kind of uPC. because when the fan runs, if you think about it, its a modulated light. (like a roller ball mouse, or optical mouse).

and the frequency of the light going on and off as the blades pass by can be monitored with a controller. if its steady, or dead, itll kill the relay, and disable the ballast, and maybe a warning LED might go on, hell, you can make whatever you want happen. I love that!!

but hey, you may be on to something there. the faster the fan runs, the faster the light will pulse on and off, and the higher the frequency. You can program a microcontroller, or you may be able to get away with using standard logic gates, and monostable "one shot" oscillators. and when the frequency gets past a threshold, (slowing down fan), you can have a gate toggle a relay to kill the bulb, and light a warning light.

[CCMCornell]

Lucca,

There are definitely 12Vdc computer fans with RPM signal. For computers, I have been a fan...err...of the quiet, inexpensive and reliably undervoltable Panasonic Panaflo L1A's which *don't* have RPM sensor, but I know a lot of Quiet PC enthusiasts who spend more on fans that do have them. I've also seen some lesser expensive fans with RPM monitoring but of unknown quiet-potential. I'll scrounge around to find some recommendations, but try out http://www.silentpcreview.com

Also, there are computer fans with thermistor control, so they can run after a certain temperature is reached. That may be an option.

[Dazzzla]

Just monitor the heat comming from the fan with a thermister.

DJ

[superdaveumo]

Another option might be to use a photoresistor. Place the photoresistor on one side of the fan and a focused light source on the other (a laser pointer perhaps?). The resistor would be MUCH easier to work with than the IR. My only concern would be if the photoresistor could change resistances that quickly... Hmmmmm...

Perhaps a prewired IR switch. Can't be that expensive. IR is a pain in the butt.

[cbm5]

It would be a hell of a lot cheaper to JUST USE FANS WITH SENSE WIRES.

Or go to the root of the problem: you don't want overheating? Use temperature and forget about the fans. I should already have a few Dallas or Maxim samples lying around, 1-wire interface thermometers etc.

[pinkmouse]

AD do a fan fail chip, can't remember the P/N, but it just monitors the negative/earth lead for fault conditions. IIRC, quite cheap as well.

[Guy Grotke]

I agree with cbm5: It doesn't much help you to know your fan is spinning, if your LCD is still running too hot. Rather than bother with fan RPM sensing, just implement 3 or 4 temperature sensors. The 1-wire sensors are very cheap & easy to use with a microcontroller. Using a bus interface would mean that people could add the right number of sensors for their projector design.

Then give it some PWM outputs for driving MOSFETS that can control 2 or 3 12 VDC fans. If a temp gets too high, it would be able to increase the fan speed. If that fails to cut the temperature, then you get a shutdown & alarm.

There are some nice optoisolated solid state "relays" (really triacs) that you would use to control power to the lamp circuit.

If you do want to use PWM to control a DC fan, then it is a lot easier with a microcontroller that has hardware PWM generation. Doing it well in software can be a pain, if you are trying to do other stuff as well.

[Luca Brazzi]

Just monitor the heat comming from the fan with a thermister.

The thing I like about measuring a light source (IR/etc) is that it can be used for 2 purposes... First it can be used to determine if the fan is moving/stuck... and it can be used as the input to the fan speed control logic (frequency of the pulses from the detector=fan speed).

From what Ive read around here... and what Ive noticed on my own 4 fan projector, it would be beneficial to lower the speed of the fans as low as possible to reduce noise.

I agree with cbm5: It doesn't much help you to know your fan is spinning, if your LCD is still running too hot. Rather than bother with fan RPM sensing, just implement 3 or 4 temperature sensors

True... the ultimate goal is to protect the LCD from overheating, so yes temp sensing is definitely important. The fan speed sensing thing would be mainly for alarming you that there is a problem with your projector... Ive seen many PJ designs where folks completely embed fans inside the PJ where they cant be seen. Even on my PJ... where the fans are visible from the outside... Id like to be alerted in some way if one of them stopped spinning/never started spinning/etc even if it doesnt result in my PJ going into a meltdown situation.

IMO a fan problem could be treated 1 of 2 ways by the controller... A) Cause the PJ to shut down... or B) Cause a warning alarm to go off to alert you that there is a problem but not cause the PJ to shut down, and.... it could be different from fan to fan in the same PJ design... So what Im thinking is that how each fan is handled by the controller should be selectable for each fan. For example... in my PJ I have 3 fans blowing in, and 1 fan sucking air out... I would want to be able to tell the controller (if one of the 3 input fans stops working... no problem just warn me... but if the exhaust fan stops working, shut down the PJ)

If you do want to use PWM to control a DC fan, then it is a lot easier with a microcontroller that has hardware PWM generation. Doing it well in software can be a pain, if you are trying to do other stuff as well.

Either that, or there would need to be external circuitry to do the PWM.... Basically I would just want the Main controller to tell the external PWM circuitry/controller what the current setting should be for that device, then leave the actual PWM signal generation to the external controller.