Sound card DAC tutorial

[arcaneinsomniac]

If you look back in this thread, I made a list a while ago about all the parts you will need with links. When I tried to order the board from Dr. Lava with all the parts, he refunded my monday because he does not offer that anymore. You can try though...

Okay, I'm gonna look through. However I'm noticing a lot of posts deleted by a moderator. I fear it might be in one of these posts...

Edit: found it! sweet.

Also saw a video of your scanner on youtube. that is so freakin' sweet man.

Sounds like I am going down the same path you once did. You even had the same scanner unit I do! I've got no laser mounts or analog laser drivers yet but that's not gonna be difficult to figure out compared to this DAC stuff. Except maybe if I decide to try to focus a mitsubishi 635 for red. Assuming there's any success to be had with that, 1.5 watt of 445, 200mW of 532, and 500mW of 635 would make for one kick ass projector on the cheap (by my estimates it may be possible to keep that just over $500)

 

[paulzimm]

On page 4 of this thread post #90 has my parts list and all the links where I purchased the components from. Also, if you look at the top of page 5 you will see the front/back of the bare PCB that someone super-imposed if you want to make you own. Best of luck to you friend!

 

[ElektroFreak]

When I tried to order the board from Dr. Lava with all the parts, he refunded my monday...

Man.. if someone tried to refund my monday I'd tell 'em to keep it..

 

[arcaneinsomniac]

So I finally finished reading this entire thread and the whole Spaghetti/Gary debacle is rather depressing.

What I do find encouraging, however, is that at its basic level the soundcard DAC requires nothing more than what amounts to audio samples. Being so simple in its essence I wonder why there aren't more open-sourced softwares for laser scanners.

If I ever write any laser show software (which I probably will, now that I think about it. I'm always looking for new ideas for cool software) I will probably release it for free. I don't know about anyone else, but the chances that I'll bother checking out any software that costs money (let alone $600 worth of money) is slim to none. The easiest possible way to get people to try out your software is to make it free. The only software I pay for these days is video games. And what I'm paying for isn't the software itself so much as the servers and services that are provided to make it all work. I find the concept of paying money for a large integer to be somewhat absurd. People are gonna reverse engineer it. If you spend all this effort to prevent people from picking your awesome creation apart, it's more likely to be wasted effort as a function of how useful the software is. Spend that time doing something productive...

And thanks for your help, paul, I appreciate it! I'll post back with results when I complete the parts acquisition phase.

 

[arcaneinsomniac]

These trimmer pots are incredibly expensive! What's up with that? At at least $2.50 each, for 12 it's $30 right there. And that DC/DC converter is $7!! No wonder the Dr stopped selling it as a kit for $40.

Can somebody help go over the functional differences between the (D106E as labeled on laserboy PCB) +/- 9V DC/DC converter vs paulzimm's recommended component which provides +/- 12V? I could tap +/-12 from PCI since I am using a PCI card for now but later on I'll want a more flexible USB interface so I will want it.

Also, the PCB also labels 20K ohm resistors. What's up with that? Should I pick up some 20K's as well?

Here's my order on digikey (I prefer to order from one place. And digikey's shipments get here within 2 business days usually. Though I'm thinking about giving mouser a try one of these days):

Index	Quantity	Part Number	Description	Customer Reference	Backorder Quantity	Unit Price	Extended Price
1	10		296-1775-5-ND	IC OPAMP JFET 3MHZ DUAL 8DIP		0	0.49000	$4.90
2	10		A100204-ND	CONN IC SOCKET VERT 8POS TIN		0	0.10900	$1.09
3	10		478-1837-ND	CAP TANT 10UF 6.3V 10% RADIAL		0	0.57800	$5.78
4	40		10.0KXBK-ND	RES 10.0K OHM 1/4W 1% METAL FILM	0	0.09100	$3.64
5	7		3296W-503LF-ND	TRIMMER 50K OHM 0.5W TH			0	2.56000	$17.92
6	7		3296W-103LF-ND	TRIMMER 10K OHM 0.5W TH			0	2.56000	$17.92
7	10		338-2367-ND	CAP ALUM 10UF 63V 20% RADIAL		0	0.15000	$1.50
8	2		LM7905CTFS-ND	IC REG 1A -5V 4% 3 TERM TO-220		0	0.60000	$1.20
9	1		811-1421-5-ND	CONV DC/DC 1W 5VIN 12VOUT SIP DL	0	7.61000	$7.61
Subtotal	$61.56

 

[Bionic-Badger]

These trimmer pots are incredibly expensive! What's up with that? At at least $2.50 each, for 12 it's $30 right there. And that DC/DC converter is $7!! No wonder the Dr stopped selling it as a kit for $40.

I was saddened that I couldn't get in on buying one. You can get comparable 10k trim resistors for less than $2.50 though, for example this one. Ebay is a good place too. Just look for the number of turns and the form factor. Cheaper trim pots with lower values can be used if you have some static resistors to offset it too. Might not be worth the time though.

Can somebody help go over the functional differences between the (D106E as labeled on laserboy PCB) +/- 9V DC/DC converter vs paulzimm's recommended component which provides +/- 12V? I could tap +/-12 from PCI since I am using a PCI card for now but later on I'll want a more flexible USB interface so I will want it.

The range determines the maximum voltage swing the correction amp can produce at its output. The ILDA standard uses +/- 10V for its X and Y coordinates, so having only a +/- 9V will artificially limit the maximum angle your galvos can utilize. Depending on the op-amp you use, the actual output voltage swing may be even lower than that. Also the voltage offset will affect the range you can use too. It may not be a big deal if you're only using smaller angles.

Oh, and guess who burned out his +/- 9V regulator with a wrongly polarized tantalum capacitor?

If you have access to a cheap 24V power supply, you can use that and a rail splitter like this.

Also, the PCB also labels 20K ohm resistors. What's up with that? Should I pick up some 20K's as well?

Not sure. Maybe they're for ESD protection, or the safety interlock? You'll need to look at the diagram.

Never hurts to have spare resistors on hand. Get them in quantity so you get the discount.

Here's my order on digikey (I prefer to order from one place. And digikey's shipments get here within 2 business days usually. Though I'm thinking about giving mouser a try one of these days).

I'm a Mouser man myself. Digikey's sorting and searching is like an anachronism from the 1990s. I only use them when I can't get what I need at Mouser, or it's cheaper.

 

[arcaneinsomniac]

Thanks Bionic-Badger for helping me out.

Just look for the number of turns and the form factor. Cheaper trim pots with lower values can be used if you have some static resistors to offset it too.

The trimmers I chose seem to be 25 turn units so it does look like I'm getting my money's worth.

If you have access to a cheap 24V power supply, you can use that and a rail splitter like this.

The power supply part of my scanner (spacelas pt-30) seems to have +24V and -24V. Is there a chance I can use that? Although that would mean the correction amp will be permanently stuck to the scanner hardware which I would want to prevent.

As for the 20k resistors I guess I'll pick up a couple of those too.

The thing is I can't really find a diagram of the layout for the laserboy pcb. All the different correction amp designs have subtle differences. I figured getting the PCB from DrLava's site would let me simply assemble the device as a kit rather than worrying about circuit layout. The other route available to me is the home-made PCB which would be... a lot of work.

 

[ElektroFreak]

So I finally finished reading this entire thread and the whole Spaghetti/Gary debacle is rather depressing.

What I do find encouraging, however, is that at its basic level the soundcard DAC requires nothing more than what amounts to audio samples. Being so simple in its essence I wonder why there aren't more open-sourced softwares for laser scanners.

If I ever write any laser show software (which I probably will, now that I think about it. I'm always looking for new ideas for cool software) I will probably release it for free. I don't know about anyone else, but the chances that I'll bother checking out any software that costs money (let alone $600 worth of money) is slim to none. The easiest possible way to get people to try out your software is to make it free. The only software I pay for these days is video games. And what I'm paying for isn't the software itself so much as the servers and services that are provided to make it all work. I find the concept of paying money for a large integer to be somewhat absurd. People are gonna reverse engineer it. If you spend all this effort to prevent people from picking your awesome creation apart, it's more likely to be wasted effort as a function of how useful the software is. Spend that time doing something productive...

You're right about laser DACs working with audio frequency waveforms, but there's a bit more to running a pair of galvos properly than there is to driving speakers. One of the main differences between the currently available free options and software/hardware combos such as Quickshow is that there are proprietary techniques used in the professional level options that produce cleaner output from the galvos and the lasers, and enhance the longevity of the galvos through optimizations to the waveforms. When I compare LFI Player, HE Laserscan, Spaghetti and Quickshow side by side, the difference is immediately noticeable. The image quality in Quickshow is smoother and cleaner and the galvos make less noise. The next best in that group in terms of overall output quality is LFI Player.

I love seeing people get the bug to create! Can't wait to see what you come up with in terms of software should you decide to write some.

Also, Bionic Badger's suggestion to use ebay for components is a good one. You should be able to find plenty of pots that will work, and at 1/10 the price you quoted.

 

[arcaneinsomniac]

You're right about laser DACs working with audio frequency waveforms, but there's a bit more to running a pair of galvos properly than there is to driving speakers. One of the main differences between the currently available free options and software/hardware combos such as Quickshow is that there are proprietary techniques used in the professional level options that produce cleaner output from the galvos and the lasers, and enhance the longevity of the galvos through optimizations to the waveforms.

That's interesting. I imagine I need to do a lot more reading on this subject, especially over on PL. I realize the need to optimize the display sequence so that we minimize the amount of extra distance the galvo needs to cover in order to draw a particular design. Is this what you are referring to? Now that you got me thinking about it, I reckon there are more factors at play. For instance I can see how sending a signal to a scanner that is too fast for the scanner to accurately track, can potentially damage it or at least cause excessive wear-and-tear. I can see a situation where particularly pathological cases of input may cause the galvos to "resonate".

Well, that's research for me to do when I get to that point, I guess.

Also, Bionic Badger's suggestion to use ebay for components is a good one. You should be able to find plenty of pots that will work, and at 1/10 the price you quoted.

I have gotten electronics from eBay. I just find that digikey is great because I get that package in the mail VERY quickly and it is VERY nicely packed. I just know that if I decide to get my trimmers on ebay I'm gonna end up waiting on them when everything else is here.

That said, I found 10 each of the exact same trimmer pots on ebay for $10, making them 50 cents each. Which is 5x cheaper. So I did just order that. Gonna see if I can get the seller to send it via some type of express airmail from HK.

And I ordered the rest, minus the trimmers, on digikey.

 

[LaserAdct]

I tried to use LFI player with my setup and it was a complete mess. Instead i tried HE-Laserscan 5 (the free edition) and everything worked perfect. With LFI my blanking was all messed up showing tracelines all over the place when drawing even simple things. I think this should be tweakable, but I never figured out how to. With HELS everything was in perfect order. The free shows that can be downloaded are quite impressive and is certainly something to try out since it is free. I cant even imagine doing shows like that in LFI which is probably the worst piece of software I have tried for use with a laser scanner. LFI is usable for single ilda frames, but wheres the fun in that. IMO that piece of software died a long time ago when people stopped developing it. It is OS and despite that nothing is developed for it anymore, that kinda tells a silent tale about how useless it is at current state.

 

[Bionic-Badger]

The power supply part of my scanner (spacelas pt-30) seems to have +24V and -24V. Is there a chance I can use that? Although that would mean the correction amp will be permanently stuck to the scanner hardware which I would want to prevent.

Just make sure your op amps can handle that kind of voltage. Ultimately, the it'll be the gain and offset resistances that determine what the output is. The voltage rails just give you more headroom.

The thing is I can't really find a diagram of the layout for the laserboy pcb. All the different correction amp designs have subtle differences. I figured getting the PCB from DrLava's site would let me simply assemble the device as a kit rather than worrying about circuit layout. The other route available to me is the home-made PCB which would be... a lot of work.

I just rigged up something like what I saw in the earlier posts here. My correction amp is actually just a breadboarded version. I haven't taken time to make a more permanent solution. There's really not much to it other than an op amp with offset and gain, and a clean power source. If you're going to invest in the time of making something special also consider creating an analog current controller for your lasers like Benm showed earlier.

 

[arcaneinsomniac]

I tried to use LFI player with my setup and it was a complete mess. Instead i tried HE-Laserscan 5 (the free edition) and everything worked perfect. With LFI my blanking was all messed up showing tracelines all over the place when drawing even simple things. I think this should be tweakable, but I never figured out how to. With HELS everything was in perfect order. The free shows that can be downloaded are quite impressive and is certainly something to try out since it is free. I cant even imagine doing shows like that in LFI which is probably the worst piece of software I have tried for use with a laser scanner. LFI is usable for single ilda frames, but wheres the fun in that. IMO that piece of software died a long time ago when people stopped developing it. It is OS and despite that nothing is developed for it anymore, that kinda tells a silent tale about how useless it is at current state.

This is very good to hear, that there does exist a free software that works well. I knew about He-laserscan but I was almost certain it would be just as expensive as other options. Thanks for sharing your experiences!

Though elektrofreak seems to say that LFI has better output quality. Hmm.

Just make sure your op amps can handle that kind of voltage. Ultimately, the it'll be the gain and offset resistances that determine what the output is. The voltage rails just give you more headroom.

Well, yes I think I meant to take GND and +24 from scanner and send it thru a rail splitter like you mentioned. But I think that would not work because my grounds would then be at different values between the amp and scanner which is a no-no.

I just rigged up something like what I saw in the earlier posts here. My correction amp is actually just a breadboarded version. I haven't taken time to make a more permanent solution. There's really not much to it other than an op amp with offset and gain, and a clean power source. If you're going to invest in the time of making something special also consider creating an analog current controller for your lasers like Benm showed earlier.

Youre right, it is a rather simple opamp circuit. I just really prefer to be careful and not get it wrong the first time around, though. I should probably first build the amp in my DIY LPM so as to familiarize myself with these circuits again.

For the Analog current controller I have always figured on getting flexmod P3's. From what I remember from Benm's diagram it looks like it is utilizing a BJT or maybe even a Darlington (somebody correct me) and i'm not comfortable with the power dissipated by such components (esp with 445's in the picture).

 

[ElektroFreak]

As with any software, LFI requires some learning. Oddly enough when I first tried setting HE Laserscan up it was a royal pain, but that was some years ago. The newer versions are much improved.. I've had good luck with LFI, but it did require me to figure out how to get everything going properly. Once I did that, it became my primary live-control software until I bought QS2. Both LFI and HE Laserscan have live control options that make them very powerful, and for MUCH more than just single frames. It's unfortunate that software gets bad reviews and negative comments when often the people are just frustrated that they can't get things to work easily.

 

[Bionic-Badger]

Well, yes I think I meant to take GND and +24 from scanner and send it thru a rail splitter like you mentioned. But I think that would not work because my grounds would then be at different values between the amp and scanner which is a no-no.

Actually, it shouldn't matter since the ILDA standard specifies differential pairs of inputs for its signals, not absolute grounds. Always good to test however because the scanner maker may not implement it that way.

For the Analog current controller I have always figured on getting flexmod P3's. From what I remember from Benm's diagram it looks like it is utilizing a BJT or maybe even a Darlington (somebody correct me) and i'm not comfortable with the power dissipated by such components (esp with 445's in the picture).

I doubt the Flexmod does any different. It probably uses a power MOSFET for regulating current, which still dissipates power by making the MOSFET act like a variable resistor. The only way to save power is to use fast PWM modulation, but often you can't get switching speeds fast enough to not be visible to the naked eye in conjunction with the scanner. It's really not that much heat dissipated actually. I've done it with about 0.9A and it gets hot, but you just use a heatsink.

There are other benefits to the P3s though, so it might be work getting. ~$40 each is a pretty hefty price though.

 

[arcaneinsomniac]

So there's no way to get fast analog current control without bleeding off extra power as heat? I guess standard methods to do it (buck circuit?) has inductances and properties which make it too slow to keep up?

So I got an email saying my order from Nautilus Integration has been cancelled. I guess DrLava doesn't have any more laserboy PCB's. I think my digikey order is in, though. I guess it's breadboard time.

 

[Guyfromhe]

So there's no way to get fast analog current control without bleeding off extra power as heat? I guess standard methods to do it (buck circuit?) has inductances and properties which make it too slow to keep up?

So I got an email saying my order from Nautilus Integration has been cancelled. I guess DrLava doesn't have any more laserboy PCB's. I think my digikey order is in, though. I guess it's breadboard time.

I don't know all the details about it but I think it has something to do with inconsistent power outputs and something or other from the lasers at different power levels thus screwing up any attempt to make analog colors... I could be wrong.