Senior Member
Police have no ideas what you are talking about. Most of them anyway! Class 4 lasers, you need to fill out an form and registred at Statens Strålevern and get it aproved! You can use class 3b lasere here without any problems, so long it is a safe setup. No audience scanning, witout all saftey requirements..Originally Posted by Amuse
And I do have some lasers equipment for sale..and I am from Norway!
Pangolin FB3 QS/LivePro/SMS2Laser
Riya MultiBus
Pangolin LD2000 Pro
Needs moar photons
thats a pretty silly statement... you cant guarantee a QM2000 will NEVER put out an unwanted static beam either... what if the driver crashes in some magical way? or you get a power surge that freaks out the card or some totally implausible yet possible event... scanfail should be in the projector, and you should have interlocks and a big red "holy crap" button... while a soundcard based dac does seem a lot scarier than other options, I really cant see how it would be THAT much more plausible for it to fail in the manner you describe
It's not silly at all...
The QM-2000 has been designed from the ground up with laser safety in mind. Every circuit is designed to fail "safe". True, this does not mean that some crazy confluence of power spikes, driver errors, and acts of god could not cause it to emit a static beam, but it's damn unlikely. Pangolin understands that their device is an important link in the chain of laser safety, and they spend a *lot* of time making sure they get it right.
Contrast that with the design of a cheap USB sound card. It's not a critical safety device. Not at all! When used as it's intended (driving a set of headphones), there is very little - if any - risk in the case of a device failure. Thus there is no need to design a sound card to fail safely. Likewise, the drivers for these cheap sound cards are notoriously unreliable. Several users (including me) have experienced system crashes (under Win XP pro) while running these devices. And that's not a big deal if you're using it just as a sound card.
But when you modify it so it will drive a laser projector, you just promoted that device to "safety critical" status. Now it *is* important that the device not fail improperly. But have you done anything to the sound card (or the drivers) to increase it's safety or reliability? No, you haven't. And that is exactly what Mccarrot is talking about.
Yes, there is always operator intervention to consider. But human reaction times are not instant, which is why a proper laser professional does everything he can to prevent problems before the show is ever put on. You don't just slap some gear together and then say "well, if anything goes wrong, I'll just hit the kill switch..." Try putting that in your variance and see what the CDRH has to say about it...
To imply that a sound card is even remotely secure compared to the QM-2000 is laughable. One is a professional device, the other costs $20 on E-bay and is disposable. While I agree that a sound card makes a great first controller for a hobbyist on a very tight budget, if you're doing professional shows you can afford a professional controller (be it one from Pangolin or another manufacturer).
Adam
Senior Member
Not disagreeing with your assertion that a real DAC is a superior answer (or that a manual scram switch is going to be far too slow if a static beam is hitting someones eye (seconds, you want well sub ms shutdown if it goes pearshaped that way), but I do take issue with the DAC being part of the safety system.
The DAC even if it is effectively standalone with on board processing (QM2K style) which was written to a suitable safety integrity level (most are not, and you cannot rely on a PC driver for this because Windows is both far from a realtime system, and is certainally not qualified for use in a safety critical system) is still at the far end of a (usually) cheap off the shelf printer extension cable, and is thus too remote from the action.
Safety critical systems should be designed to minimise the complexity of the parts within the safety critical area, in this case you do this my arranging for the projector to be safe for the audience irrespective of what is fed to it, not by designing it so that it is safe if it is not fed anything silly!
The 'failiure surface' that you have to be concerned about rises extremely quickly once you bring software or firmware into the picture and designing a hardware and software system that is known to fail safe (even in the presence of say a VHF walkie talkie coupling to a cable that is just the wrong length[1]), really is rocket science. The limits of formal verification (even abscent processor manufacturer errata notices[2], and even assuming the assembler and liker have no bugs[3]) are only a few thousand lines of code.
If ADAT (which has some truely nasty failure modes) is good enough then it surely follows that a soundcard is no worse (Provided you buy a good one)?
Regards, Dan.
[1] This has actually caused serious problems with theatrical scenic automation at theme parks and major venues, where the computers are in some cases running multi hundred HP drives, or motors flying people! The saving grace is that the safety systems do NOT involve the computer (except possibly as a status display).
[2] For example the HC12 had a doozy where every 4096th time it woke up from sleep on interrupt, it kind of didnt! An Intel Core 2 has 88 pages of erata last time I looked.
[3] A certain embedded vendor has a compiler that I discovered forgot what "volatile" meant when you turned the optimer on, nightmare!
ADAT's aren't used for any audience scanning shows that I'm aware of. Controllers are, and that's why they need to be part of the safety loop. (In a non-audience-scanning show, typically a projector mask is used to eliminate the possibility of beams reaching the crowd.)
Coupling from an errant VHF radio to a laser projector's output would be impressive, since the X & Y position data is resolved using a differential receiver on the scanner amp inputs. You'd need one hell of a coincidence to couple to just one leg (on each amp) and not the other to create a static beam that way. Still, if you are running single-ended receivers on your galvo amps then I could see where this might happen. But it's still highly unlikely. (And it's just about impossible with differential receivers.)
As for cable failures, that is addressed in the ILDA pinout. (Re: the interlock loop). If the cable fails, the projector quits. Granted, you could have a failure of a single wire inside the cable, but that's awfully rare, and in any case that would collapse only one axis (X or Y), so that still wouldn't yield a static beam. In my personal experience I've never had a cable failure, but I do buy decent cables, and I store them carefully.
I agree that hard projector masking is the way to go for non-audience scanning shows, and in fact this is the way it's done here in the US. But since audience scanning is far more prevalent in Europe, you can't use a mask to keep the beams out of the crowd... True, you should also have a scan-fail interlock if you're going to crowd-scan, but having a reliable controller is also important.
I agree that the controller is not the *only* safety critical device in the system, but I do believe it is prudent to consider all components (including the controller) when assessing the safety of the show.
Adam
Senior Member
Controllers, safety critical or not? (Was Cheap laser for concerts)
Just how much gain bandwidth do the opamps you are using have? CMRR at > 5 MHz is mostly non existent with a simple 4 resistor and opamp style diff amp, and not that much better even if you get clever (And besides, I see an awful lot of 'ILDA' cables that are NOT twisted pairs).
The real fix is common mode chokes, but I don't see much including them.
Actually, the place where this really tends to bite is when the RF gets picked up by something like a clock trace on a processor card, instant crash, and it can be a right monkey to find (BTDT).
The other thing that hurts us here is the tendency of a lot of builders to NOT connect the screen directly to chassis at the projector, AES 48 has much to say about this from an interference rejection perspective. See also papers by Neil Muncy, Bill Whitlock and Tony Waldron (they are written for audio, but we have many of the same issues). Basically not doing this makes the screen a NICE aerial for AM radio, and if the kit at the other end ties it to signal reference rather then case ground......
Hopefully! Designing an interlock circuit which actually works really reliably is not as trivial as it appears on first glance, and the tendency of a lot of kit to have badly thought out ilda loop through connections does not help.As for cable failures, that is addressed in the ILDA pinout. (Re: the interlock loop). If the cable fails, the projector quits.
That does of course help :-)but I do buy decent cables, and I store them carefully.
Well yea, but I tent to put the controller in the 'Production Critical' rather then 'Safety Critica'l class, the rig should be safe (even if the show is lost) irrespective of what the controller outputs (and in fact, irrespective of what the projector receives).True, you should also have a scan-fail interlock if you're going to crowd-scan, but having a reliable controller is also important.
One approach I have seen used to dim things when crowd scanning involved a set of 2 stopdown glass sheets that could be set in front of the aperture to attenuate the beam power when the Y axis dropped below a suitable point. A sort of mechanical intensity map. Combined with a scan fail, all is good.
Regards, Dan.
Needs moar photons
I was not disagreeing with you, im just saying no device is always going to fail safely, and no safety device is going to be better than a well trained vigilant operator... when it comes to critical electronic devices shit happens, especially when its least convenient cause thats just the way it works lol
that being said, I would not audience scan with a sound card dac personally, and would imagine anyone that is well versed enough to think they can audience scan safely would also realize that would be a poor decision
You think so? This is what I came up with on my first attempt. It seemed pretty trivial for me, and I'm not an EE...
Senior Member
what mw laser are you using
what mw laser are you using in the video? thanks and God bless.
Posting Permissions
Reply With Quote
buffo