Running the XJ-A140 With Missing Diodes

[OzOnE]

Hi,

Actually, I'm not sure if the pot would be necessary, or worth the effort - I'm hoping the A140's "current sense" voltage will work the same as the A130. The pot would only be for finding the "voltage sense" value in case it differs a fair bit from the A130. Once the best value is found (so the PJ stays on reliably in all modes), the pot could be replaced with resistors.

The mod will probably work with the resistor values I used for the A130, or something close. Please don't quote me on that though. lol

If you did use a pot, you would still need to solder very small wires to where the resistors normally go, and short across all four current sense signals (then do the same for the voltage sense signals).

I found the resistor divider value for the A130 by looking at the o'scope shots and using an online calc to get close to the voltage needed.

I think with the A140 driver board, we only need to know if it's a hugely different design or PCB code to the A130. I haven't seen a photo of the A140 driver, but I'd put money on it being the same as the A130.

(The laser driver and Phlatlight driver chips probably have plenty of headroom, so all they would need to do for the A140 is to change the current setpoints and the processor firmware).

btw, I tried finding the photos on the Hungarian site using archive.org , but I couldn't find them? Maybe he was asked by Ca$io to take them off?
No problem though, I just need a quick photo snap of the A140 board if anyone can find one?

I searched the forum and didn't find much. "Things", and a few other people showed some photos, but it wasn't confirmed if they were A140's or not?...

http://www.photonlexicon.com/forums/...336#post202336
http://www.photonlexicon.com/forums/...0+driver+board

@Keith - It will be quite difficult to find or build a driver for the hard drive motor unless you can mod the original PCB for speed control? I think those brushless motors need some fairly accurate three-phase signals to drive them?

A brushless motor chip that I'm quite familiar with is the Allego A8902, which was used in practically all DLP projectors for a long time...

http://www.datasheetcatalog.com/data.../0/A8902.shtml

EDIT: btw, the Allegro chip needs a serial data input to control the speed etc. So, we'd be back to needing an AVR or PIC chip again.

tbh, It's going to be a real nightmare trying to sync your motor with the Ca$io without any feedback though - it would be like the DJ mixing challenge from Hell!

What I was thinking of was sacrificing the original phosphor wheel, or making a lightweight colour filter (which is slightly bigger than the original wheel) and sticking it to the original wheel.
That way, you'd have the correct speed / feedback, and you'd only need to move the wheel slightly to shine the beam through your new filter(s) instead.

@NITAGRO - Not sure about the different projector models? I'm assuming they all use a similar laser diode setup, but I don't know what the differences would be without seeing one dismantled.

Split into two posts for clarity (hopefully)...

OzOnE.

[OzOnE]

OK, @Kobra - you've got me thinking about whether the four resistors need to be removed or not... It might be possible to leave them in place, but it means that the output from the DAC will be "fighting" with the output from the feedback opamps. I don't know if this is healthy for the board, the resistor divider would probably need changing, and it might pull down the feedback voltage too much and keep the PJ from staying on?

tbh, there should be plenty more of these board available in the future, so I wasn't too worried about removing the resistors.
Note: The original resistors (R366 to R369) are 1K on my A130 boards.

btw, There are other points on the board which may be easier to solder to (for modding), but it might look quite messy with more wires everywhere...

For instance, if you look at the underside of the board (the "normal" way up, with the processors at the top) - the right-hand sides of R366-R369 are connected directly to the processor pins (ie. the same sides which point towards the black diodes D317-D320.)

You could just remove the resistors, then bridge across these four points instead. This combined signal will be your "current sense" INPUT to the processor.
You then just need to solder a wire link from Pin 2 of the DAC IC110 (or the right-hand pad of the missing R187) and this new "current sense" signal.

The voltage sense signals are taken from the laser bank outputs (ribbon connector CN301), but they then go through resistor dividers near to the processor pins.
I can't see an easier place to bypass these because the relevant resistors are just as close together as the small caps are. However...

The laser bank anode outputs go into R109 (underside) and R110, R111, R112 (top side).

The "lower end" of each resistor divider is then formed by R126, R129, R130, R133 (all top side). The opposite end of these resistors are simply Grounded.

I normally leave R109-R112 in place for the mod because they are fairly high value (around 20K I think, I haven't tested them off-board yet).

The only way around this fine soldering is to start cutting tracks: not exactly ideal!

OzOnE.

[OzOnE]

Right, I finally made time to finish building the sequencer yesterday...

I had the usual few hours of frustration at first - I couldn't get the chip to program for a few hours (needed extra caps), I think I fried the first AVR chip, and then I couldn't get the sequencer to work properly (it wouldn't enable the Red LED because it wasn't seeing the falling edge of the Laser Bank ENable).

I now realize that when you set the standard interrupt pins on these AVR chips to look for "Any logical change", it means exactly that! ie. It will generate an interrupt every time the pin changes state, but it seems you can NOT then test to see if the pin is actually high or low! (at least I don't think you can). This might be great for frequency counting, but not so great for this project.

Anyway, the solution was simple. You need to use the PCINT type interrupt instead which just means changing a few lines of code. This then allows you to check whether the pin was high or low.

So, the board is basically working! The connection to the projector couldn't really be easier. It's just two wires for CIDX and Laser Bank ENable, then the Ground and Power (12V) wires to the PSU.

I currently have it running some LEDEngin LEDs and it is driving all three colours (RGB) in perfect sync with the phosphor wheel / Red Phlatlight.
The 12V PSU voltage from the projector actually isn't high enough for the voltage drop of these LEDs (around 16V), so they're nowhere near their full brightness atm.

The Phlatlights have a much lower voltage drop (they're more like a giant LED made from many standard LEDs in parallel), so it should be possible to drive them from this board using 12V from the PJ.
(It can currently only manage 3 Amps per colour, but I'll give it a try later anyway).

Note: This is all on an A130, and it's still necessary to carry out the "Laser free" mod first. The original phosphor wheel isn't bypassed yet either (that's the next thing to do).
I'm still confident (/hopeful) that these mods will also work on an A140.

I haven't tried shining the RGB array into the PJ yet because I'm using my old DIY module which is too big to fit in there.
I'll see if I can sneak a peek at the full colour image by routing the RGB array via some mirrors or something, then figure out how I'm gonna fit my Blue and Green Phlatlights!

btw, It's completely possible to upgrade the sequencer for driving much higher current LEDs. It's a simple enough circuit.
You can always use an external PSU of course.

I'll try to get my crappy camera to hold a charge and get a video uploaded too.

OzOnE.

[kobra000]

dang Ozone..your really shucking and jiveing on that thing....I salute you!

you said the pj is capable of doing 3 amps per color...now thats the projector power supply or something different? cause im wondering if you can combine the power for all three colors to get a total of 9 amps (im itching to try my SST-90 and color wheel,just got it built the other day,had to use a speed control from Hobbyking for RC stuff to control the HD motor..simple enough only 8 bucks, and smaaaaaaaaaaaaaaalll! but takes forever to get here from china!
i really like your idea,but im just not that hip on programming and circuit building..most of what i do electrically is just repairing stuff...but im still learning and open to new ideas..have you got a circuit diagram that i could study for this sequencer? with symbols and whatnot?

love your work man..and i cant wait for the pics...im still having trouble scanning the parts you needed pics of my scanner's drivers are completely out dated and wont work with XP for some reason ( service pack 3 messes everything up man games everything!) and my darn CD key for VISTA just crapped out on me..(vista sucks so much i wont even spend the 10 bucks for a new key!) so i have to wait till my windows 7 for builders comes here from newegg,hopefully i can get it going that way.

sally forth young gents!

Keith.

[OzOnE]

Yep, I'm running it directly from the Ca$io PSU. Only four wires needed (if you don't use a separate Ground for the control signals).

The Red Phlatlight is probably run at around 8.5 Amps or more, so there should be enough power in reserve. Also, with all three (RGB) colours, only one LED is ever on at a time. There might be a very tiny overlap, but I'm going on the assumption that the PSU will handle it OK.

It should be upgradable to however many amps you need. I'm looking at replacing the LM350's with MOSFETs instead. It should run a bit cooler then too.

(I did try some MOSFETs in a similar circuit a few years ago, but I don't think it was able to regulate the current reliably?)

I'd like to send you a board as promised, but it's a bit messy as it is. It ideally needs a proper PCB to be made for it...

Could anyone possibly make a few simple single-sided PCBs?

I have some etchant and a PCB here to do it, but I normally use the toner-transfer method and my laser printer exploded (It decided to FART it's toner around my room one day).
The PCBs never turn out too great either.

I might put the diagram and code up soon. It's not exactly "trade secret" stuff since I've already described how it works in previous posts (pretty much).
The circuit is fairly simple. It's just the AVR chip, three LM350 regulators, 10 resistors, a few caps and a push button.

I added a nice feature last night where you push the button and it cycles between some test modes (Ca$io mode, Red, Green, Blue, White).
I needed this because I realized I couldn't easily run the projector AND see underneath the mainboard at the same time. I can now force each LED to stay on, it just needs a temporary external PSU.

I now have the Green Phlatlight running inside the PJ alongside the original Red one...
I had to fit a mirror for the Green Phlatlight, remove the first chunky lens, rotate the first dichro by 90 degrees, then fit a small mirror in place of the second dichro.
It could probably do with another relay lens because the Green is slightly further away than the Red.

@Keith - Don't worry about the A140 photos. All I really need to know is whether you think it looks the same as (or similar to) the A130 driver board?

OzOnE.

[kobra000]

Hey guys..any progress lately?..ive been shut down due to a seized truck engine....you know what they say life sucks without a car!

[kobra000]

Ozone how is everything coming along?

[kobra000]

hmmm gonna have to breathe some life back into this thread...well got my little hard drive motor color wheel thinking made up...made from a stripped Mini dvd that i kinda sorta spun down to jus under 1 1/4 inches..I ordered my film from a dude on ebay red,green,and blue.and gave each one an equal 3rd of the wheel (figuring this out was not easy,im not that great wih fractions!)..Hobby king had a great speed controller that works like a charm for the engine for only 8 bucks! (cell phone charger works great too as a DC power source though the speed controller though the speed controller is for R/c helicopters is can take up to upwards of 50amps! amazing little thing)..its got a little pot on it that allows me to max out he engine at 7200 rpm though i think it could go as high as possiably 10,000 rpm..and its still quiet! just a steady hum...but i dont want to burn either up so i wont push it...now i just gotta figure out if i can tak on to power it from he Projo PSU.i doesnt take that much at all...ill take pics soon,i just became and Ubuntu user and its a learning curve,im still very curious if color wheel Rpm with the right color sequence will match up to he pulses of he DLP..if anything its a cheap experiement to proove i was wrong lol.

hope everyone is awesome and doing great

Keith.

[bysu]

could some guys tell me how to short the contact pads , my XJ-M140 contact pads is different from A140.

thanks

[kobra000]

hmmm seeing as i only have xp with he a140 i couldnt really tell you..but earlier in thread here are a load of pcitures that will show you how...some i have posted myself and others have posted.is basically pad to pad wih a small blob of solder..flux works well too and keeps everything nice and neat


Keith