Topic: Problems calibrating a new laserboy DAC

[c60]

Hi James, and other laserians.

I'm having some issues calibrating this dac.  After your help figuring it was one of the LM072 chips I replaced those with an LM358 (another dual op-amp) and tried to do calibration on it.  At least until I have time to go pick up the new chips.

I've managed to zero all the colour channels but when I check the X or Y I get +8.5V and -8.5V and nothing I do with the voltage offset seems to do anything.
This is still without a test signal from the soundcard.  I'm using the pins right on the IC's to check, so pin 1 for the xy and intensity and pin 7 for the color channels. Although the color and intensity channels testing on the jumpers yields the same numbers.

Any idea what's going on or where I should probe to see what's happening? It's not the actual chips as the color channels are behaving normally and swapping them around doesn't seem to change this annomaly. 

Nils

[James]

Take a look at the photos on this post:

http://laserboy.org/forum/index.php?topic=561.0

Notice the DB25F ILDA port. See that it has some black magic marker around some of the holes. Those are to mark...

1 = X
2 = Y
5 = red
6 = green
7 = blue
8 = intensity
13 = +5V shutter

... all in the top row.

Use a paper clip or piece of wire that fits nicely into these holes and measure your tuning voltages there with respect to GND.

If you measure about +9 and -9 (or a bit less) on the chips' pins 4 and 8 that is normal. That's the power supply to the chip.

Also make sure your sound device is in fact in 8 chanel mode and make sure the Windows sound system is set up for 7.1 (8 channel) speakers.

James. 

[c60]

I found the problem.  Turns out some of the pads either wern't there or came off during soldering.  I did notice those boards didn't take up the solder as good as other ones I've soldered.  So after careful circuit tracing it was the x + y where the little traces come from the soundcard input to the first resistor's.  They were really small or perhaps some of the coating got over too much of the copper or I soldered too hot.

With a little scraping of the trace and an extra wire for good measure I managed to get around the problem.  You might want to inspect those little solder pads.

What voltage do you set the XY at once you're playing the circle test?  5V AC? 

The next part seems fine although I'm getting nothing on my blue channel out and intermittent  on the Intensity Channel on my multimeter.  What exactly does intensity do with an RGB diode projector?

 Is there a setting in laserboy where you tell it what type of projector you're exporting to?

Nils

[James]

I use a 50W iron with an adjustable power stand. It's pretty much a light dimmer. I have it set to just a hair above 3 on a scale of 0 to 10.

So far, I've built 4 of these kits and I have had ZERO issues with the boards. I think they're really nice!

I use an AC volt meter to measure X and Y when playing the circle wave and I set it to about 1.5VAC. Then I know it's safe to plug in the projector and set the gains for X and Y visually to get a good circle.

That last channel is something you can set up in LaserBoy. You can make waves with a mix of RGB for single color lasers. There are a few ways to mix the RGB. Look in the [Tab] menu for options.

James. 

[James]

Just so everyone is on the same page...

Here are some instructions for tuning the correction amp....

Use a wave file that you can make in LaserBoy.

Go to frame 10 the big white circle.

Make a wave of that frame.

Use that in Spider Player.

Make sure your sound card driver is set to 6 or 8 channels.

Hook the black lead of your meter to the GND test pin.

Pull off the jumpers.

Turn all the offset pots (500 ohm) down all the way counter clockwise (up to 25 turns).

Turn all the gain pots (100K) up all the way clockwise (up to 25 turns).

With no signal, turn the voltage regulator pot (also 500 ohm) so that you are close to zero volts on all outputs. Look for the lowest voltage of all channels and make sure you are just a bit on the positive side.

If you measure the DC offsets directly from each channel of the DAC, you can measure the negative offset voltage from the test pin on the correction amp (V_off) and adjust it to be -1.5 times the offset you measured from the DAC channels. That will get you pretty close.

Let the thing sit for a while and settle. Re-adjust as needed.

Now with all the channels measuring some little positive voltage (from about 30 to 120 millivolts) you can trim out the offsets. Turning up each 500 ohm pot per channel adds a bit of negative voltage.

Once you have all the channels trimmed out to near zero you can play the circle wave. Since it is white, it will drive all the color channels to max. Turn down the gains on the colors to +5.00V. Use the AC setting on the meter to measure X and Y to something small like about 1.5V AC.

Now you can put on the +5V limiter jumpers. This adds the zener diodes to the outputs of the color channels to bleed off anything over 5V to ground.

Finally, you can hook up your projector and adjust the X and Y gains visually to get a near perfect circle.

Please ask me if you have any questions.

Other than that, please let me know how it goes!   

James.

[c60]

I managed to get it mostly configured and driving the laser.  My green is still not working and I'm not sure if I have the colors gained properly, but the circle is a nice big circle and I did a little color test and it seemed to have gradients, just no green.

[James]

Wow. That sucks. I guess you could do a little wire re-routing on the back of the ILDA port pins and use the I channel to drive green. It's easy to change the input side.

I don't know.... It's probably better to try to fix it than mess it up more.

[James]

Use your meter and measure resistance across each resister 10K and 15K. Make sure they are all the same. The green channel will probably measure different. Measure back from the pads where the resisters are soldered and make sure the solder joints are good.

Pull the chips and measure to the socket holes. Also measure from the socket holes to the output holes on the DB25 ribbon. Look for continuity.

All 6 channels are exactly the same except for the 4 with the zener diodes. But with the jumpers pulled, that is out of the circuit.

[c60]

It's all good, I managed to get it all sorted.  I think it was a calibration issue.  Calibration feels like that circus act where you are spinning all those plates, and have to keep them all spinning.  It's very analogy  

I have full color and have my levels nicely calibrated.  Now onto the software side.

It occured to me though that I could use the program puredata and have some realtime feed control.  That might be a neat way to hack together a player program.  As a proof of concept I just hooked up a bunch of oscilators to the various outputs and sure enough it makes all sorts of nice lissajous patterns.  My next thing will be to hack together a little sample player so I can trigger the little sequences I make up with laserboy. Or hook in my midi controller so have some realtime control.   You can pretty much do anything with puredata.

The fun starts now.

[James]

That's awesome news!

Take some pics or even some videos!

We'd love to see what you come up with.

Please feel free to ask a lot of questions about using LaserBoy. Somehow it always ends up making LaserBoy a little better!

I really appreciate the fact that you are looking through all the negative bullshit about sound card DACs and LaserBoy.

Seriously! I know it's out there! I see it! Thank you for seeing past all of that and giving it a try. I will support you as well as I can.

Get back with me on Skype and I can show you how to do some really cool stuff in LaserBoy!

Also be sure to check out some of the stuff that Dean Hammonds wrote and ask him about it (also on this forum).

James.

[c60]

Here's a simple example using pure data, I just hooked a separate oscillator object to each channel and played with the values.
http://youtu.be/UaZxmkm3sWo
 this is going to be real fun to experiment with, especially with some midi knobs and faders, or a wii chuck with some accelerometer data
And a multi sampler playback so I can cue patterns easier.
So many ideas.
  An arduino patch with some graphic feedback could be a neat aid in tuning, you could watch all levels at once.
 
 What about balanced sound cards? Is there some way to get around hacking soundcards?  This would be a "pro" option, ¼ patch cords and some big tuning knobs, or maybe even digital pots? Lots of places for "improvement" but I also find the analog/digital crossover of the existing tech quite amusing :p
I do love some of the patterns I was able to get with laserboy. For making clips it seems very powerful.  I definitely want to learn more.  What I'm really after is the ability to tweak in realtime which pure data should work great for.  I wish laserboy had a preview and realtime playback, or perhaps I'm missing something. 


As for any negative of sound cards DAC's it's greatly overridden by the freedoms of not being locked into softwares and ecosystems.  I don't know how else I could use laserboy/pure data/VLC/abelton live/openlase, a wii controller... The list is infinite


Oh and I'm still using those lm358, they might not fare as well when I start pushing it but so far so good.
That's another question I had about galvo's what's good for them? How hard can you push them? And what does that actually mean? Noise is generally a bad thing? Is there some way to limit the signals to them or should the drivers handle that?


Did I take the red pill or the blue pill, hey wait, there's a green pill in here too?


Sorry bad humour


Lots to learn in the realm of LASERS!


Thanks for my next great time waster =)

[James]

Wow! I'm glad you are getting into this so well. It sounds like you have quite a new and complex perspective to add to all of this.

The thing with scanners is maximum velocity. That translates to the slope of the curve or the change in voltage over time. LaserBoy deals with this by limiting the graphical distance between samples (or points in space).

So many questions and so many different ways to look at this!

I think you will provide at least as much insight as you do questions. Please keep on doing what you are doing!

Yes. You are correct. The echo Layla 24/96 sound device is balanced at zero volts and has enough voltage to drive the laser projector with no external electronics. But it's not USB and for Windows XP it is not a standard 7.1 device. I saw at SELEM 2013, that Windows 7 makes it work very well and I never saw it like that before. (Thanks DZ!).

Look for the "blank pill". Keep your mind open and your own! Be original.   

James.

[c60]

So are you saying with a balanced output soundcard no external, or limited, circuits are required?

I have an older prosonus firebox http://www.presonus.com/products/firebox
It's got 6 balanced outputs.  Runs over firewire and linux/windows/mac all are good.  Also it does 24bit/96khz.

I'd love to get that as a "backup"

As to limiting the slope of the curve, how would that represent in audio? an ADSR filter? with puredata I can easily do simple maths on the audio data too. seems like a good thing to put into puredata as the last chain.

It looks like puredata has no problems with multirack audio in it's sampler tables as long as it's less than 64 channels. I'm still working through the specifics.  You have to specify how many header bytes and data bytes and big or small indian.   Hopefully I can cue up predone sample loops from virtualboy and massage the data with some live midi controls..

[James]

I'm not sure about that sound device. If it really only has 6 outputs, then you won't be able to put stereo audio into the last two tracks. That's kinda' nice if you think about it. One wave file... double click... Spider Player opens up and you have a full color laser show with stereo sound!

It's also nice to know that if you have your waves made properly with the right signal polarity, you never need to worry about driving the color channels over +5V. That's what the zener diodes do for you.

One way to limit velocity of the scanners is to roll off the high frequencies. That's pretty easy. But it's not the whole story. If you consider a simple triangle wave, you could have a slow rise and fall time in voltage, but since the wave has an instant change in direction at the top and bottom of the wave, represented by the sharp angle, it actually has an infinite series of harmonics that go all the way to the top of the audio spectrum.

If you roll off the high end of a triangle wave, you end up with a sine wave which is all round.

Any time you go from motion to an instant stop or from a stop to motion, you have discontinuity in the function. That gives you all kinds of high frequency harmonics. That's OK!

In LaserBoy, you set a maximum distance between the points (samples). So it's really a visual or graphics process that ends up becoming a kind of audio or wave filter.

If you make your wave snippets in LaserBoy and you optimize them there, you should already know they are safe for scanners.

You can also open waves made in other ways in LaserBoy (as visual vector art) and analyze them for the maximum distance between any of the consecutive samples.