Repairing and Upgrading Laser Spirograph

Posted in Education Outreach, Projects by Bill
18 Feb 2012
Repairing and Upgrading Laser Spirograph

My favorite local museum the Science and Discovery Center of Northwest Florida had broken exhibit that was collecting dust. They asked if I could fix it, and never one to back down from a challenge I set out to bring the Laser Spirograph back to life. What follows is how I fixed it and added options for new features!

The Laser Spirograph was an exhibit that had been broken since my good friend Tish became the director of the museum. Never seeing it in a functional state I assumed the exhibit consisted of a laser shot through a set of scanning mirrors and finally onto a screen. The instructions on the front asked guests to adjust two knobs to change the pattern on the screen and talked about the relationship between two frequencies.

The broken exhibit. Click for full size

Any exhibit with that sign is begging for attention from an engineer.

The two pictures were taken after I removed the electronics from the exhibit for a closer examination. What I found inside was electronics that were nearly 20 years old. The laser emitter was an old fashion gas laser, which looks pretty cool but it’s brightness doesn’t compare to modern diode lasers. It still seemed to be functioning.

Day 1 on my work bench.

The scanning system on the other hand seemed to be completely dead; they moved, but only ever so slightly. I dissected the system a bit to find how it functioned. At the heart was the board below.

Function generator board.

This board, aptly labeled “Dual Tone Generator” did just that; it produced two clean-looking sine waves.  The two signals were fed into an amplifier for speakers and into the laser scanner driver board.

The frequency of the two tones was governed by the position of two 25 turn potentiometer, one for each generator. The pots were attached to large wheels for museum patrons to turn on the front of the exhibit. The mechanical engineering on this part was impressive, the pots were connected to the wheels via a friction clutch so that when the pot hit an end-stop, the wheel could still be spun by eager hands without causing damage.

Frequency adjustment pots with clutch.

I didn’t dive much further or try to reverse engineer the function generator because it was working, scoping its output gave a clean looking sine wave. The problem was lying down stream of it.

Next I removed the plate the laser was mounted to and flipped it over to reveal the driver board for the laser scanners.

Scanning driver board

The driver board seemed to draw power from the laser board on the top side. It had two pairs of transistors in push-pull configuration, a handful of op-amps and a ton of trim potentiometers; none of which were labeled of course. The connections to the board were questionable. One galvo was connectorized, the other was soldered to the connector pins. I disconnected and removed the galvos for closer inspection.

Galvos

This is when I found the problem. From what I could gather Googling the part number, these were suppose to be 7 Ohm galvos. But when I measured I would get 1.5-2 Ohms from either of them. When I did the math for how long this exhibit has been operational and compared to the expected lifetime listed in the spec sheet, these galvos should have been long dead. It was time to get new ones.

Although I am an Electrical Engineer, I had no experience with laser systems. I wanted to confirm my findings and see what was out there as replacement options. I found and joined the LaserPointerForums community and asked around there. I tried contacting the manufacture of the galvos but got nowhere.

I finally decided (with the help of the members of LaserPointerForums) the best course of action was to replace the old open loop glavos with new and modern closed loop ones. I found a whole laser scanning kit on eBay that fit the bill and placed an order. The drivers for the new glavos stated they required a +/- 5V input of signal. This meant I’d have to add an amplifier to increase the audio level signal from the function generator to the level the drivers wanted.

This also gave me an idea for a new feature. I found a lot of DIY laser projectors during my research to repair this exhibit. Many people were taking cheap sound cards and feeding them through ‘correction amps’. Sound cards are 1bit DACs. They switch on and off at a very high frequency (in the Mhz) to create a pseudo analog signal. This creates a nominal voltage level (offset) a few volts from ground and varies high to low depending how long the 1 bit is switched on versus off. A capacitor removes the DC offset before getting to the audio jacks. If you remove the capacitor and amplify the signal you could use a sound card as a laser DAC. But you would also have to cancel out the offset without using a capacitor. The easy way to do that is to use a summing amp.

Source: Wikipedia

An op-amp in summing configuration sums signals and amplifies them. If you summed the high offset signal coming from the DAC with a negative voltage reference you could remove the DC offset without a capacitor.  Then you could use the sound card / DIY laser DAC to scan images or logos with the laser.

An example of an DIY laser projector. Source: HackaDay.com

I talked to the museum to see if they wanted that option. I figured an old computer (or better yet, an Arduino and MP3 Shield could ‘playback’ scanning signals for logos) could be used to generate signals for logos of museum sponsors  when no one was using the exhibit. They were interested, but didn’t want to invest in the idea just yet. So I set out to design and build my amplifier with hooks to enable the laser projecting option at a latter date. This is what I came up with.

Schematic. Click to enlarge.

Board.

The board consists of two dual channel Op-amp circuits. One is a basic amplifier for the function generator. The other is in summing amp configuration with an adjustable negative voltage reference. A telecom relay decides which signal gets through to the output. The design makes the upgrade easy by just requiring the addition of the signal source for the logos and control circuitry to determine what to put on the screen when the museum wants to add the ability later on. For now it will always pass through the amplified signal from the function generator.

You can download my design here, but I warn you the summing amp side has not been tested. However, it is a common design among DIY laser projector hackers.

Fast forward a very busy month and I finally got the time to build the new board and test out the new glavos. I stayed after hours at work one day with the best technician ever and we machined a mounting plate for the new galvo assembly. After years in disrepair, the Laser Spirograph finally projected it’s first pattern on my ceiling.

Houston, we have Laser!

The new electronics.

Closeup.

The first slow day at the museum I brought over the repaired guts for installation.

The empty insides of the exhibit.

Now with more Laser.

Closeup.

The front-side.

Through the looking glass.

Here’s a short video showing the operation of the now repaired Laser Spirograph exhibit.

The museum was quite happy to get one of their favorite exhibits back on the floor. All my work was documented and will be added to the ‘big book’ of exhibit information and will hopefully make it easier for the next person to fix it in 20 years.

Update

Well, as some of the comments below pointed out, there are risks involved with going the cheap route. After a week of continuous duty, the galvos failed from overheating. They were replaced, and the exhibit was modified to become ‘on-demand’ as opposed to continuous duty. A timed delay relay was added to the exhibit.

New timer spliced into power cord

Zoomed out

An LED backlit button was added to the front. The relay was two switches. On switch only closes when time has run out. I wired the LED’s power through that switch and to the 12V out of the relay so that the LED was only on while the exhibit was off and ‘sleeping’. This attracts attention and begs to be pushed, turning the exhibit on for 60 seconds.

New button from behind

Light on, exhibit off

Light off, exhibit on

This should greatly reduce the wear on the parts and keep it running much longer. I felt this was a better option then just adding fans. There’s no reason why everything in this should be running all the time.

Share