Savel brain dump in English!

This is first version of laser diode power supply. I borrowed circuit from die4laser.com and changed it a bit. It is classic current source using n-channel power device. It is better to use p-channel devices here, but now I have only n-channel mosfets on my workbench. They are cheap and easy to get. Modern mosfets a very compact and powerful devices. Now they are very steady devices and even can work in very high temperature. During experiments I overloaded supply and transistor self-melted solder (I am using lead solder)- working temperature of particular device is 150oC!

Resistors R4, 5, 6- they are used to set max current. R8- is shunt. If we place here 1 ohm resistor, we get 1V reading for every ampere. R2, 9 -1k. R1 is about 600 ohms. R3, C1- is soft start circuit. Not implemented in current design. C2- only few hundreds of picofarads. Only to prevent from auto-generation. C3- is a bigger one. Transistor- any n-channel mosfet powerful enough to carry working currents and voltages.

This is PCB image. If this picture is printed with 200dpi resolution, it can be used to build PCB.

This WAS prototype of power supply. It killed few laser diodes. So I made another version and I place circuit diagram in next blog entry when I’ll translate it from Lithuanian version. I recommend to use regulated power supply and not more than 5V.

There are lots of rumours in the internet about high power DVD lasers and about hacking them. I decided to test some of the rumours. Theoretically, there are two lasers in modern DVD recording device. Power of the laser must be strong enough to change the physical properties of recordable media. In this post, I didn’t examine IR (infra red) laser used in same device for CD/CDRW media. I tested RED laser used in DVD recording and reading. The newer DVDRW device, the faster recording speed- the more power in laser.

Laser diodes is CD/DVD drive come with various optical lenses. These optics is not very useful for burning experiments- the laser beam is separated and then focused in very short beam. And for real “laser gun” we need long range optics. There two types of laser diode modules- stand alone lasers and lasers with multiple photo-diodes (detectors) in same package. Older laser diodes were built with photo diode build in package. This diode is used to detect laser generation and to measure optical power. In some new devices this diode is missing. Even if the package have three pins, I didn’t find photo diode inside, even when I disassembled diode.

Laser diodes are sensitive devices. Especially they are sensitive to static discharges and over voltage-over current modes. I damaged 4 diodes experimenting with my power supply- the transitive process during power-up and power-down killed them. On some laser diodes there is optical element glued to package. It is beam splitter- you can get remarkable interference lines when using these splitters, but for power experiments we need to remove it. After removing, the “can” is open to free air (and some diodes are made without protection), so keep “open” diodes dry.

Laser diodes, same as LED diode is current, not voltage user. But using simple ballast resistor is not good way to solve the problem. The PSU for laser diode is current regulator with dual feedback: one from current, another from photo diode. And PSU with over voltage protection. The PSU is very simple (I’ll post two version of PSU I build). The components are very cheap and everything can be found in old computer motherboard.

Here is the image of working DVD recording laser diode without collimator optics. Other, simple diode without heat sing is seen too. Never use high power laser diodes without heat-sink. The current in threw this diode is about 150mA and according to the datasheet, the optical power is about 100mW.

If I place diode to my digital camera lenses I see such nice image. The light is dissipated, so there is no possible damage to camera detector.

Here is first experiments trying to focus laser beam. I used optical collimator some some old IR laser. This collimator is not very suitable to high power RED lasers. There is big problem to get good collimator for such laser.

And now some words about safety. Lasers used in DVD-R device and working in full power are potentially dangerous devices. So: NEVER STARE TO LASER BEAM WITH REMAINING EYE! IR (infrared) laser are double dangerous- they are powerful and invisible.

I found datasheet for Rohm laser diode RLD65PZB5: max optical power (pulsed, 50% duty cycle) is 240mW! Current according datasheet is about 400mA.

I would like to introduce part of my workbench. It is computer side of my workbench. There is electronics workbench, but it is not described here. I’ve made some order here before making pictures- regularly there is a big mess here:

Small description of the stuff:

1. Very old, >20″ SGI (Silicon graphics) monitor. It is time to throw it away, but from time to time I need to connect so old unix box of some PC with nonstandard video timing. This monitor is capable to sync to very wide range of sync signals (or even without them- sync on green). The only problem is rare 3W10 connector. I’ve made some cables for it. Very heavy.
2. Some old unix boxes: SGI Indy, Sun Sparc Xterminal 1 and Sparcstation 5 (from recent blog entries you may understand were I found last one). And what is displayed on CRT screen.
3. My regular computer, every day it is getting older and weaker compared to new ones. It is P4, 3GHz, HT computer. (nor D, nor Duo). with some SATA drives and 2GB of RAM. I was sick of out-of-RAM messages and virtual memory performance. Now memory is cheap and I recommend to invest some money to RAM.
5. Two 14″ LCD monitors. I love Lot’s of desktop space.
6. some HDD.
7. Black keyboard. I love classic small keyboard. I hate “natural” keyboard and keyboard with “media keys”.
8. Classic Russian tea cup. Regularly I drink white coffee with lots of sugar.
9. Sun keyboard. It is only connected to test sparcstation.

And now backstage…
(more…)

From computer tomograph PSU, I removed battery of capacitors. It is 12 high power electrolytic capacitors. 6 x 12000μF and 6 x 8200μF. All caps working ant 400V (450V surge). Total 121200μF (0,12F). We can charge them to about 400V and store about 9696J of energy. Lots of energy. And lots of lethal energy. It can kill you.

More about capacitor calculations in older blog entry.

I achieved personal top record in braking electronics. Today (Nov 14th) I disassembled computer tomograph. As whole device was too big to fit in the van, I don’t have classic photos of the doughnut (donut). I have only partly disassembled device images.

Here is the images of “super computer” displaying human body slices:

In fact, it is very old, Sun Sparcstation 5 computer, with 256Mb RAM, some external CDROM and hard-disk. With SunOS and Philips software.

Here is the remains of donut. The device was ancient. It was made in 1995, but the design is much older. Main custom computer is made using DIP chips in 5V technology. The only identified CPU was MC68020 working at 16MHz, and several custom made or maybe MIPS CPU’s on several printed circuit boards.

It was X-ray tomograph. Images are produced using rotating x-ray tubes and 192 detectors. This whole bunch of heavy metal is rotating around human body. This system is powered from 380V 3 phase mains system. The power is converted to higher frequency, so all x-ray high voltage transformers are much smaller. But main PSU is very big, full of big capacitors and power IGBT transistors. Lots of power is used for such device- mains braker is 250A.

As device is very old, only few interesting components were found inside. It was great disappointment.

Maybe I have some superstitious beliefs. So the are no Atmega USB 13 blog entry and no project #13.

Next toys connected to my AVR-USB testing board are two Freescale semiconductors (ex Motorola) accelerometer chips: MMA3202D and MMA7260QT. First is “catastrophic” accelerometer chip with high-G (100g X-axis and 50g Y-axis) sensor. If you apply such acceleration to human body it will squash to yogurt. Bet in real world such forces are achieved in various accidents- just drop down your hard disk to stone floor and the force will be about this. Sensitivity is not very high, but it is possible to detect static gravity force or free fall.

Next chip is more sensitive, so called low-G sensor. (Selectable Sensitivity (1.5g/2g/4g/6g)). It has 3 axis. As this device required 3.3V power supply, I placed small linear regulator on board. All other stuff on PCB is to filter electric noise. The device is very noisy. First I thought that my PCB and AVR analog input is the source of the noise, but after visiting Freescale seminar and looking at original demo board of the same chip I found that the chip itself is very noisy.

There is small example in software to demonstrate the usage of the chip. When jumper “firmaware” in ON, the LCD display is showing the position of sensor PCB. It is useful to use such feature in video and photo cameras to display pictures in LCD display. Or to reduce handshaking. I didn’t implemented “free fall” feature in my source. But is very easy to do- just detect 0g on all three axis. It is useful to use 0g detection in some notebook computers or hard disks. Just to park shock sensitive devices and prevent them from shock damage. Another interesting device is small shock logger- just to collect data in some parcel. But there is bad side from this technology. In near future, all handheld devices will have such loggers. And if your device will fall down to ground it will log this accident. And in local repair shop, your device will betray you.

The software is very simple. All freescale chips are analog chips. I was asking about I2C or other chips (they are promoted in website), but freescale representative was very mysterious about the developing schedule. So now I just measure the output with ADC converter and interpret the answers.

Source code: 20071102.zip.

There is NO PCB for these devices. My laserjet printer is bad and PCB are very simple. So I hand cut the traced with small graver. Even for leadless QFN chip.