Video Projector Ballast Bypass Help

[elektroshok]

Hello,

I am new on the forum and new with electronics.

Recently I started a project in modding a video projector, and did so successfully, and now got another that is giving me a hard time.

It's a NEC VP400 Video Projector that I got without a lamp and I want to put in it a Xenon Light since I have a 55W Ballast and Bulb, the problem is this.

The VP has a 5pin connector on the motherboard that goes to the lamp ballast, pin 1 is the +3V Lamp OK, pin 2 is GND, pin 3 is the signal to fire the lamp, pin 4 seems to be GND to and pin 5 I believe is ECO Mode. In the previous VP when I shorted the pins for the lamp OK signal it powered on without any problem, on the NEC it still gives a Lamp Error.

From what I found on the internet some VP's have a Serial Communication between the motherboard and ballast, and for this reason I'm stuck. And I can't find the service manual for the VP, because in the manual I'm certain I could find the traces for the Lamp OK signal.

Here are some pictures with the VP.

Motherboard with shorted Pin 1 and 2 on the ballast connector

Ballast with opto isolator legs shorted

Ballast with a chip for whom I can't find any data

Another with the Motherboard and two GND's

Pictures with the chips on the motherboard








And a clip with the motherboard and the lamp Error code:



Can anyone help me in bypassing the ballast? Thank you!

[elektroshok]

Hello again,

I've hooked up an LED to the Pin 1 and 2 of the connector that is supposed to be the input from the ballast and the LED lights up, did the same for pin 3 and 4 and nothing, measured the voltage and I get ~2,5v and if I measure pin 3 and Screw Ground I get ~5v and Pin 3 connects to two opto isolators on the ballast, is 2,5v enough to trigger the opto isolator?

I did the LED part because I was inspecting to see if the LED will flash due to serial communication.

Please don't laugh

Maby the lamp error is due to other problems? I would really like to make this projector start because it has a better optics block and better LCD's than the other projector I've modded.

Any idea is welcomed. Thank you!

EDIT:

I have another Idea, instead of fouling the projector that the lamp is firing I could make the ballast think the lamp is OK and then it would send the right signal to the projector. Any ideas on how does the ballast know when the lamp is running and send the signal to the optocoupler??

Here are some pics with the ballast:

[elektroshok]

I've tried to ground the pins that go from the small board on the ballast to the larger board and still no luck, and as I am new to electronics I can't think of any way the ballast talks to the small board an send a signal that the lamp is working.

It could be as simple as putting a resistor in the right place, but what is that place .
If the ballast is using serial communication to send a signal to the motherboard something must be telling the serial IC to send that signal.


Please if you could give me some ideas.

Thank you!

[Towger]

Is there a microswitch on the lamp door?  On the Sanyo projectors this in is part if the Lamp/Ballast check circuit. With some models many a new lamp has been bought when the problem is the microswitch not getting pressed properly by the door.

[elektroshok]

Yes there is, and is soldered in closed position, and even if it was open, the error would be the lamp door not the lamp itself, the error I get is a Lamp one. every sensor has it's own error, lamp cover solid orange and the projector does not start at all, Temp 2 red Blinks, Fan Error 4 Red Blinks, Lamp Would not start 6 Red Blinks, and that is what I'm getting. The projector sends the signal to the ballast to start the lamp and the ballast tries to start it and if the ballast does not reply with the lamp OK signal in 10-15sec the projector gives a lamp error. I need either to find the lamp OK pin on projector chip and ground it, either to foul the ballast into thinking the lamp is lit.

[elektroshok]

Silly question, could I damage the motherboard if I take a battery and an LED and try to find the trace of the lamp OK signal by searching for continuity with the wire between the LED and the battery?

No other ideas left, tried with the multimeter and is hard because my crappy multimeter takes 2 seconds to display the resistance and 2 seconds multiplied by how many connections on the motherboard is forever, with the LED the response is instant.

Thank you!

[elektroshok]

Tried with the Battery and an LED and could not find a single spot on the motherboard that is directly connected to that PIN 1, perhaps its connected directly under the CPU, could this be impossible to hack?

No ideas from anyone?

[bowlofpudding]

i ran into a similar (possibly the same) problem as you, with an NEC VT48. it has the 5 pin connector. pins 1 and 2 send to the ballast. pin 2 seems to give a pulse while the lamp is on. pin 3 is 3.3v+ going to the + of the optocouplers. pin 4 is gnd and 5 is data going to the projector. i saw a data stream while the projector was on. it seemed to be a serial stream repeated every second. i programmed a pic micro to send this data and was able to get it to turn on, and stay on so far. keep in mind im using a pickit (that a built) to read the data stream. this is what the data stream looks like. i dont think its the osram usart, but it might be. i read somewhere it could be dali, but i dunno. and i realize this post is kind of old, but there seems to be so little information about this, so i thought i would try to help

[Kjelt]

i read somewhere it could be dali, but i dunno.

If it is Dali it should be bi-directional so the motherboard should issue a command and then the ballast would reply. The voltages of Dali are very specific and easily identifiable see picture below, the '1' should be between 10V and 22V so not easily to be mistaken.
The protocol of DALI is 1200baud serial.

If it is only a single stream from ballast to motherboard it looks like something proprietary, just a status signal of some sort.
If it is a one wire protocol so data and clock in one signal chances are high it is Manchester encoded so try to find the baudrate and decipher the stream with this encoding and see if it takes you a step further.

[bowlofpudding]

i didnt notice the voltage specifications before, only that it was a serial stream. the voltage on the projector side never goes above 3.3v or below 0. as far as i can tell the projector only sends a short pulse to the ballast, al though there is an "eco" mode while would lower the bulbs brightness, so there either must be serial stream going to the ballast to control the brightness, or some sort of pulse width modulation. i think manchester encoding is whats going on here, although the timings arent very tight, but again im only measuring it with the pickit2's logic analyzer. blindly sending the data its expecting seems to keep the projector alive. the last projector i modded was as simple as sending a low to the projector to tell it the bulb is ok, and i know the osram ballasts usually use a usart at 9600 baud (if i remember correctly) but this does seem to be entirely proprietary. im just hoping my led seems a more suitable light source in this projector than the last one, seems most of the light is being wasted. puts out a good image in a completely dark room, but just seems it should be brighter. well anyway heres the code i used (its for a pic micro) in case it can help someone else
<code>
void main()
{
    //Initialization
    ansela = 0x00;
    anselb = 0x00;
    anselc = 0x00;

    //Interrupt initialization code
    option_reg = 0xC0;
    osccon=0b11110010;

    porta.2=0; //
    porta.4=0; //
    porta.5=0; //
    portb.4=0; //
    portc.0=0; //
    portc.2=0; //
    portc.6=0; //
    portc.7=0; //

    trisa.0=0; //icsp_dat
    trisa.1=0; //icsp_clk
    trisa.3=1; //icsp_mclr
    trisb.5=1; //rx pin1 data
    trisb.6=0; //led
    trisb.7=0; //lamp_opto
    trisc.1=0; //icsp_aux
    trisc.3=0; //relay
    trisc.4=0; //tx pin5 data
    trisc.5=1; //   pin2 data

    porta.0=1; //icsp_dat
    porta.1=1; //icsp_clk
    portc.1=1; //icsp_aux
    portb.6=1; //led
    portb.7=1; //lamp_opto
    portc.3=0; //relay
    portc.4=1; //tx pin5 data

    //Delay: 420 us
    delay_us(255);
    delay_us(165);

    //Loop: While 1
    while (1)
    {

        portc.4=0;

        //Delay: 420 us
        delay_us(255);
        delay_us(165);

        portc.4=1;

        //Delay: 440 us
        delay_us(255);
        delay_us(185);

        portc.4=0;

        //Delay: 1 ms
        delay_ms(1);

        //Delay: 220 us
        delay_us(220);

        portc.4=1;

        //Delay: 440 us
        delay_us(255);
        delay_us(185);

        portc.4=0;

        //Delay: 380 us
        delay_us(255);
        delay_us(125);

        portc.4=1;

        //Delay: 460 us
        delay_us(255);
        delay_us(205);

        portc.4=0;

        //Delay: 380 us
        delay_us(255);
        delay_us(125);

        portc.4=1;

        //Delay: 860 us
        delay_us(255);
        delay_us(255);
        delay_us(255);
        delay_us(95);

        portc.4=0;

        //Delay: 800 us
        delay_us(255);
        delay_us(255);
        delay_us(255);
        delay_us(35);

        portc.4=1;

        //Delay: 460 us
        delay_us(255);
        delay_us(205);

        portc.4=0;

        //Delay: 800 us
        delay_us(255);
        delay_us(255);
        delay_us(255);
        delay_us(35);

        portc.4=1;

        //Delay: 860 us
        delay_us(255);
        delay_us(255);
        delay_us(255);
        delay_us(95);

        portc.4=0;

        //Delay: 800 us
        delay_us(255);
        delay_us(255);
        delay_us(255);
        delay_us(35);

        portc.4=1;

        //Delay: 860 us
        delay_us(255);
        delay_us(255);
        delay_us(255);
        delay_us(95);

        portc.4=0;

        //Delay: 380 us
        delay_us(255);
        delay_us(125);

        portc.4=1;

        //Delay: 460 us
        delay_us(255);
        delay_us(205);

        portc.4=0;

        //Delay: 380 us
        delay_us(255);
        delay_us(125);

        portc.4=1;

        //Delay: 860 us
        delay_us(255);
        delay_us(255);
        delay_us(255);
        delay_us(95);

        portc.4=0;

        //Delay: 1 ms
        delay_ms(1);

        //Delay: 640 us
        delay_us(255);
        delay_us(255);
        delay_us(130);

        portc.4=1;

        //Delay: 500 ms
        delay_ms(255);
        delay_ms(245);

        //Delay: 500 ms
        delay_ms(255);
        delay_ms(245);

    }
    mainendloop: goto mainendloop;
}

</code>

[bowlofpudding]

well this seems to have been a pointless exercise. i get about 1/5th the brightness of the original bulb using a 100 watt led module. and i get about the same amount of light output using a 10 watt led. the problem seems to be getting the light focused enough to get where it needs to be. i was considering trying a DLP projector, because i imagine they make a little better use of the light, only losses would be through the color wheel. i didnt bother taking out the filters to test, i dont think it would have made a worthwhile difference. and i still have a bulb with a lot of life left in it, i just want a solution that i wont have to replace for a long time. im considering trying the 3 LED approach having an led for each lcd. i think bypassing the bichromic  mirrors would cut down losses but i know so little about optics.

[Kjelt]

A HID bulb is pretty awsome if you look how tiny the arc is, it is really a few mm lightsource. If you go to a white led you should make optics to first concentrate all the light to a single point since there is no space in the projector that would be some additional box.
If you look at modern led sources for projectors they have a totally different approach, they remove the colorsplitter (being a colourwheel in a dlp or a prisma in an lcd projector) and use R, G and B leds directly.
Even then the led projectors still don't have the same lumen output as their hid competitors.

[bowlofpudding]

i was considering using the "phlatlight"s that are out there. seems they can put out a pretty tight beam and its a single led, so it should make a pretty even light. i would like to go the 3 led route using those leds, but finding the room to fit them behind the lcds, plus a fan for each. it would be a pretty involved project, and may not even be worth it. the leds state 2,225lm red 5,400lm green and 1,050lm blue but i dunno how many lm would be output considering all the losses.

[tchleb]

Thanks for your Solution, can you please post a schematic and some photos of your work?

[tchleb]

I tried your solution with an arduino, because the beamer doesn´t stay on. I connect pin 12 to pin1 (beamer pcb)

Code: [Select]

int txpin = 12;                 // tx_pin

void setup()
{
  pinMode(txpin, OUTPUT);      // sets the digital pin as output
}

void loop()
{
  digitalWrite(txpin, HIGH);   // sets the tx pin on
  delayMicroseconds(420);




   
                          //Loop: While 1
    while (1)
    {

        digitalWrite(txpin, LOW);

        //Delay: 420 us
        delayMicroseconds(420);


        digitalWrite(txpin, HIGH);

        //Delay: 440 us
         delayMicroseconds(440);


        digitalWrite(txpin, LOW);
       
        //Delay: 1 ms
        delay(1);

        //Delay: 220 us
        delayMicroseconds(220);

        digitalWrite(txpin, HIGH);

        //Delay: 440 us
        delayMicroseconds(440);
       

        digitalWrite(txpin, LOW);
       
        //Delay: 380 us
        delayMicroseconds(380);


        digitalWrite(txpin, HIGH);

        //Delay: 460 us
        delayMicroseconds(460);


        digitalWrite(txpin, LOW);
       
        //Delay: 380 us
        delayMicroseconds(380);


        digitalWrite(txpin, HIGH);

        //Delay: 860 us
        delayMicroseconds(860);


        digitalWrite(txpin, LOW);
       
        //Delay: 800 us
        delayMicroseconds(800);


        digitalWrite(txpin, HIGH);

        //Delay: 460 us
        delayMicroseconds(460);


        digitalWrite(txpin, LOW);
       
        //Delay: 800 us
        delayMicroseconds(800);


        digitalWrite(txpin, HIGH);

        //Delay: 860 us
        delayMicroseconds(860);


        digitalWrite(txpin, LOW);
       
        //Delay: 800 us
        delayMicroseconds(800);


        digitalWrite(txpin, HIGH);

        //Delay: 860 us
        delayMicroseconds(860);


        digitalWrite(txpin, LOW);
       
       
        //Delay: 380 us
        delayMicroseconds(380);


        digitalWrite(txpin, HIGH);

        //Delay: 460 us
        delayMicroseconds(460);

        digitalWrite(txpin, LOW);
       
        //Delay: 380 us
        delayMicroseconds(380);


        digitalWrite(txpin, HIGH);

        //Delay: 860 us
        delayMicroseconds(860);


        digitalWrite(txpin, LOW);
       
       
        //Delay: 1 ms
        delay(1);

        //Delay: 640 us
        delayMicroseconds(640);


        digitalWrite(txpin, HIGH);

        //Delay: 500 ms
        delay(500);

        //Delay: 500 ms
        delay(500);
      }
      }

[Brunch]

Hi guys

I found this topic searching for same answer about Osram ballasts. I don't know if it's still of interest but I've attached the protocol data sheet.

I tried an NEC LT-25 a few years back, but I didn't realise it was not the simple shorting trick and I am not into pic programming, so I put the project on the shelf.

Alan

[neckro]

Hi, i have a Sanyo PLC-XE31 and i`m trying to remove the balast.  Is any way to do that with arduino ? My ballast is a PT VIP 3AC/380 03 Rev. 2 Osram. Thank you.

[Robertg]

Hi there,
I just found this thread and wondered whether somebody is still interested in this issue...
Even though I'm nowhere close to actually understanding the protocol used by the Lamp Driver
(mine seems to be made by Ushio Japan / Phihong), I managed to crudely decode the packages
sent and received by either side. Punching these found bits into a very improvised program
left me with my projector up and running though(without ballast of course).

If anyone is interested in the data, I am willing to share/investigate further.

Projector: NEC VT48
Bought off Ebay for a couple of bucks, Lamp smashed to bits/exploded.
Lamp Ballast: PHG201G21 150W Ushio Inc. 250-420V= in, Four wire out(Lamp + separate Ignition), 5 Wire interface PWB002A

Hooked up a Halide lamp and took my scope out; results => see attachment.
Your findings might deviate, my lamp settings seem to be Eco => decreased Power consumption.

Obviously there's all sorts of issues to be resolved e.g. right now theres no acknowledged lamp shutoff
when i turn off the projector, causing it to beep and do a 1min cooldown run on next startup etc.
So much for now, i hope this helps

Regards,
Robert

EDIT: Two identical attachments, removed one

EDIT2: Micro used is Attiny 25 with Atmelstudio and AVR Dragon.
     Since Neckro replied via PM I'll look for compliance to the Osram UART...

[Robertg]

Quick update on my progress:

The Protocol does not seem to comply with the Osram Standard.
I didn't find evidence in the decoded Data transmission(didn't check for fancier things like Manchester encoding since that wouldn't allow the triple high or low we can clearly identify) and there is no lamp shutdown instruction either.
My Micro is now acquiring the data being sent by the logic board and responding with the
(sort of) appropriate answer by comparison of the first Byte of the instructions I recorded with the scope so far.

While the projector seems to be working normally, its actually sending out a variation of the second command after a certain amount if time. Only the second Byte seems to be altered so the µC throws the second package at it. Even though the projector's operation does not seem to be bothered by that, it assumes the transmission went wrong and repeats its query a lot faster than usual: Frame time drops down from ~3 seconds to several times a second...

This is likely to be the issue causing the Fan going nuts on the next power up.

I'm not sure whether I'll reconnect the Ballast and investigate further since requests for Lamp parameters etc.
can be numerous. I will however try to get the Arduino of a friend to communicate with my Tiny Micros via the
Interface so that we can have a Functional Code for further Debugging and tinkering.

Did any of you do some further analysis regarding the protocol?
My knowledge is not satisfactory for assertions about this(had to learn about interrupts and so on), so I'd like some opinions on what I concluded.

Second Command: 0001   1001   0110   0110   0010   1101   0110   0001
Variation:                0001   1001   0110   1010   0010   1101   0110   0000

Regards,
Robert

[Edivando]

Good afternoon I'm new here in the forum.

I'm having the following problem.

I have a benq projector mp772st with burnt lamp.
I'm trying to adapt an LED 100W. I've done the adaptation with the LED and the image was cool.
My problems are: the projector was not working giving lamp error to start. I put a remote control board tv to generate the PWM pulses at the entrance of a optocopler connected to the projector. It worked for a few seconds (90 seconds) after he fought with giving lamp error.
I'm buying an Arduino to control these signals and try to keep the projector running.
My ballast is a philips I do not know if it uses the same uart protocol osram. Could someone help me in this mission ??

Projector: http: //benq.co.in/product/projector/mp772st
ballast: http: //www.ebay.com/itm/OEM-INFOCUS-IN104-Ballast-4H-0GF37-A03-Board-Card-/231707691895 >>>> the same as the infocus IN104

Pardon my English. I well understand English, however error much to write.

[neckro]

Robertg, any news about Osram ? 

[ahad]

Hello
Very thanks for your reply. I have a(https://www.nanodlp Mitsubishi FD730U and want to replace its lamp with a UV lamp for 3d printer. I saw some comments in ( https://www.nanodlp.com/forum/viewtopic.php?pid=13206#p13206 ) and it refers to a Russian site.
(http://forum.arduino.ua/viewtopic.php?id=125&p=4)  But just now I don't know how and which command to send to mainboard instead of ballast.
Very Thanks