The Laser Galvanometer Thread!

I'm on the verge of ordering a 2D laser galvanometer from China: EBay. The seller says that the mirrors should be OK with 405 nm lasers, and it has a USB connection and an ILDA connection, i'm guessing the "show card" takes the USB and outputs ILDA. ILDA is an open standard so it should be no problem (more than the work involved) implementing an opensource driver card. Or one figures out how to communicate with the USB board provided and saves oneself some labour.

Good or bad idea? Anyone here worked with this kind of equipment? Is there a cheaper way to do this without sacrificing precision? Comments are very welcome!

... I've done this with similar cheap chinese galvo-sets and a Dynamics nano driver for 445nm-diodes.

It works, but the main problem is to get a precise spot in the desired working distance.

With my setup the fokus in 150mm working distance fron the mirror-head was not a spot or square, but more a line with 0.4x0.1mm.

To get the focus shape finer, I have to expand the beam to more than the 5x5mm mirror size ... so I've changed my setup to servo-motors, that can rotate really big (and heavy) mirrors with 40x40x6mm size.

You can read something about this here - [forums.reprap.org]

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Thanks for the link, i'm reading through that thread right now. Exactly the kind of info i wanted! But, the problem you mention is caused by the laser, not the galvo? I have a lot of reading and questioning to do before i can even start considering what laser to use - i've come to understand that it is quite a topic...

Maybe i should state my question more clearly. Assuming a perfect laser, what properties of the galvanometer scanner will affect the final laser spot quality and its movement? Things i can think of:

* Quality & suitability of mirrors: wavelength compatibility, adequate intensity tolerance, what more?
* Coils & magnets: haven't studied this much but i read about the guy who built his own. What weak points could one expect in a cheap bought piece of equipment, and what would probably limit the precision of a homemade one?
* Control circuitry: resolution, interference, what more?

I'll soon be starting another thread to discuss the laser part of the problem. Anyway, what do you think will be the limiting factor of the cheap galvos bought off ebay? This is a long-term project for me, i don't want to just throw something together and have it work "more or less" - ideally i'd like to approach the theoretical limits of resolution.

... if you drive the galvos with max. 50% of the stated speeds, they are accurate enough for laser-curing or engraving.

The limiting factor is the beam-quality of the laser and the energy densities you'll need.

With a good gas-UV-laser (e.g. HeCd-lasers used for UV-curing) you can use the small mirrors with around 8x8mm areal size on a typical good scanner-head and will receive a spot diameter of around 30 microns in 300mm distance from the galvos and focussing lens.

But with diode-lasers the beam quality is so poor, that you have to expand the beam diameter to something like 30mm to get a fine enough spot in 150mm distance from the galvo-head.

So the limiting factors in the DIY-range are mainly the beam quality and mirror sizes ... and with high power lasers for material processing (like engraving or cutting) - again the mirror size to expand the beam enough, so it won't atomize the mirror coatings ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Today i got the galvo to move!

The "show card" did not have a USB connection, only ILDA, which means a digital to analog converter is required between the computer and the show card. I had no idea of this when i placed the order...

The ILDA standard is eh...not very modern, it seems to me. Instead of talking serial or something you're expected to feed the controller bipolar analog signals (-5 to +5 V) for the X and Y positions. I was a little confused, the standard document specifies two pins for each coordinate, one positive signal line and one negative. My interpretation, after some serious confusion, is that a positive voltage between 0 and 5 volts should be applied to the positive signal line, and the inverse to the negative line. Anyone know if this is correct?

Anyway, i tried using a slowly oscillating PWM signal (arduino:s "fade" example) smoothed with a resistor/capacitor, and inverted it with an opamp (lm324) for the negative, and it seemed to work! A very good sign indeed, but it doesen't necessarily mean that i got it all right. I should add that the + and - signals did not look very similiar on my ol' scope, the inverted one was much smoother (might be the scopes fault though - it's probably a decade older than i am!). I do not even begin to understand why one would go through the trouble of sending the same signal on two different pins...what would happen if i just tied the negative signal to gnd instead? Or +5 V, for that matter? As you might understand i'm experimenting wildly now, breaking new ground with a high risk of catastrophic failure, probably.

Next step must be getting a decent DAC. 14 bits of resolution would give me 2^14 = 16 384 steps, with an axis length of ½ m that's about 30 microns per step - i'm leaning towards a 16 bit version, just to be on the safe side and not introduce unnecessary limitations. Single-supply, low-voltage is fine so it'll be cheap anyway - looks like they're popular in audio applications but 16 bits must be seriously outdated... It should have an internal voltage reference and output 0 - 5 V, anything more that i should think of? The output will be connected to the positive signal line and inverted with an opamp, the output of which will be used as the negative signal.

Last week i bought a diode laser controller very cheap (i think) at an auction, hopefully i'll have it delivered tomorrow. Right now i don't have a working laser so all i can see is that the galvo seems to work more or less as expected. Which is not so bad, considering my n00bness in the field of electrical engineering.

... the ILDA or analog +/- signal is more the 'hardware-side' of galvo-driving ... I'm using an USB-driver, that converts the serial sent values into the analog +/- signals.

With unipolar signals the possible driving speeds are much slower and hard to get perfect linear -- so it's made with +/- for better symmetry.

The 16k range is not so 'outdated' as you think -- in the digital world something like 64k or higher is no problem ... but creating precise analog voltages in this accuracies (and fast setting or switching them) is pretty hard work

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

I looked at the Lasershark project and its opensource but is only 12 bits, has more features than i need, and since i already have a show card i think i can get what i need (a simple DAC) for a fraction of the cost. It just means i have to learn how to build one, which, after all, is a good thing!

Disregarding the reasons i can take count of the fact that a bipolar analog signal (one signal in combination with its inverse) is expedient for driving a galvo, but i fail to see the point of using such a scheme for transmission. Inverting an analog signal seems to be quite trivial (i am a noob, this is the first time that i've used an opamp for something that i thought out myself without stepwise instructions so please correct me if i'm wrong, i will be grateful!) so why not do it as close to the object as possible? Now i need a bipolar power source.

Anyway, am i right in the assumption that the negative signal is supposed to be simply the inverse of the positive? What could happen if they don't match too good? Anyone see a risk of damaging the drivers/galvos? I turned it on for just a few seconds for the test...

Yesterday i got my Arduino-DAC-galvoscanner experimental setup working. Wow!

I used an Analog Devices AD5732R DAC on a breadboard, no other components needed (though one should definitely add some decoupling capacitors). In hindsight i should've choosen the AD5734R version with 4 output channels instead of 2, the bipolar signal requires 2 DACs per channel, two channels equals four DACs, obviously. Now i'll use an opamp to generate the inverse signals, but i'm not sure how that will affect accuracy. I use the PSU i got with the galvoscanner (+15/0/-15V) to supply the DAC chip with analog voltages, and power the digital logic via the Arduino.

Three things i've noticed so far:
* I connected all the GND pins of the DAC to the breadboard GND, and from there to PSU GND and Arduino 0V. When i put the PSU/Arduino connections close to the analog GND i got huge periodic drops (perhaps 20% of signal strength) in the output signals, probably coinciding with activity on the data bus though i haven't verified that. Just by moving the other connections away from the immediate vicinity of the analog GND pins made the visible problem go away, adding some capacitors to the supply pins should also make things better. Might be good to remember when optimizing resolution.
* I had the Arduino output a steadily increasing binary value (for loop with i++) to the DAC, and i thought it was kinda slow - it took seconds to reach top. Then i removed everything that had with serial comm with my computer to do, and the speed increased maybe hundredfold. Just forget about handling both galvoscanner and contact with outside world with an Arduino at the same time. Maybe a RasPi would be better? I know i will run into interesting problems when trying to draw something more complicated than a straight line.
* I'm a little surprised, the galvo scanner thing isn't as silent as i'd imagined - it makes a definitely noticeable, buzzing sound, even when its not moving. Is this normal, and what's causing it, control loops? They shouldn't be oscillating?

I have my galvos 20kpps ordered from aliexpress.

I have removed the large mcu from the show card and get the pins of ad5734 and driven thru arduino.

It worked like a charm (ofcource powered an laser first)

I like to draw patterns in cloths for sewing. But the 8bit dac should be replaced with 24 or 32bit so that using from far, it should point accurately to points. Btw i will also use it to etch my pcb (photoresist uv laser?)

Can anyone have any suggestions?

Quote
ajaybnl
I like to draw patterns in cloths for sewing. But the 8bit dac should be replaced with 24 or 32bit so that using from far, it should point accurately to points.

... if it's not for burning, but only visual, then better use a video-beamer.

My sister is weaving and sewing medieval cloths and pennants - to project the images on the fabric, I've modified one of the cheap, randomly available Casio XJ-A130 DLP-beamers (which were salvaged for the 445nm-diodes), so it's working with the remaining red LED only

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]