Spurred on by my success with the cBright XG2 lamp bypass surgery I decided to have a go at doing the same for the Optoma HD70 I own. This projector is about 5 years younger than the XG2 and is native 720p so from my point of view this is a much more appealing conversion project (I’d love it if it was 1080p native but we can’t have everything).
I suspected from the start that this would be a much harder projector to convert over to LED. The HD70 projector has an eco mode which extends the lamp life by about 500 hours by cutting the power output. From the hacking point of view though this is bad news as it means there is more communication between the motherboard and the lamp ballast.
In the best case scenario the communication from the motherboard to the ballast comprises ON and ON+ECO, this can be done with just two bits of information. The communication from the ballast to the motherboard would be WORKING and BROKEN, this can be achieved with one bit. The worst case scenario is that the motherboard and the ballast are chatty and swap extensive data over a serial link.
The worst case scenario is bad because unless you know what data is supposed to be transmitted it’s unlikely you’ll be able to figure out what what to send to the projector to tell it everything is ok.
30 November 2015
I’ve tried shorting across the ballast to motherboard optocoupler to trick the projector into thinking the bulb is good but without any luck. With the jumper / short in place the projector briefly boots and then after about a minute it cuts out and reboots. After three attempts it gives up and switches off. I think it’s unlikely that I’ll be able to determine the protocol this projector uses.
Notes
Optocouplers
NEC 2581AWJ625 – Stepped side indicates the input, pin 1 and 2 the output is on pins 3 and 4. For these optocouplers that means the output is on the side where NEC is printed. The ballast to motherboard communication is therefore the optocoupler on it’s own closest to the side of the board.
Ballast
Hitachi Media Electronics Co Ltd – Model: T1C82541-4
Seems to be quite a popular ballast (for some reason nearly all the pages are in German though) and a replacement could probably be sourced if you needed it. I couldn’t find a circuit diagram or communication protocol. The 1C82552-A on the ballast control board doesn’t turn anything up in Google.
- MB3759 – Fujitsu – Switching Regulator Controller
- 2901 6115B – JRC (Japan Radio Corp) – Probably a single supply quad comparator. Couldn’t find the exact part number but several similar components turned up.
- DG T5K – Unknown
- Z4 – Unknown, back of the board.
Final Update
I sold the projector. Further research showed that this upgrade probably would never work for a number of reasons. The light needs to come from a very small source if the picture is going to be well focused, that just doesn’t describe LED arrays at all. Additionally, it was going to be very difficult to trick the projector into starting up without the high voltage source present. The protocol used to communicate with the high voltage source would need to be decoded and spoofed, I just don’t have the equipment or time to do that.
References
- How to dissemble and clean an Optoma HD70
- Optoma HD70 Service Manual
- Tricking a Projector into Thinking the Lamp is Good – Not useful for the HD70
- Projector Bulb Conversion to LED – Not useful for the HD70
- LED Project Lamp V1.0 – See version 2.0 below.
- LED Projector Lamp V2.0 – Decent lamp build project.
- Video Projector Ballast Bypass Help – EEVblog – Good discussion of the problem and an example of a solution.
- Value for Dummy Load Resistor on Projector – Description of the protocol problem
- LED Multimedia Projector – Not useful for this projector but a good build
- Discharge Lamp Control Circuit – Probably completely irrelevant but looks like it does roughly the same job.
- Discharge Lamp Operating Circuit – Also probably completely irrelevant.
- Mega-thread on bypassing ballasts but very little useful information