Page 19 of 27 FirstFirst ... 9151617181920212223 ... LastLast
Results 181 to 190 of 264

Thread: To the laser show software companies

  1. #181
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Quote Originally Posted by m0f View Post
    Could it be a tuning issue ? (Edit, or as it seems to affect both x and y, grounding issue ? For reference I do run mine without the ground wire connected, and also the scanners themselves are mounted on plastic standoffs to prevent grounding through the case. Though I just did that to keep signal and power grounds isolated.)
    The tuning question came up before, but it's doubtful it's just that. Slight separation of the lines may happen, according to Zoof, the one person who ever tested this, but not that magnetic-like hysteresis that happens close to the zero-motion moment in either axis. It's a real killer, because obviously fine detail like the text in an ILDA pattern cannot be legible with a problem as severe as this, because motion must frequently stop in either axis when writing text.

    About grounding, no problem there. I'm familiar with audio wiring, and was using a DC coupled balanced output. It was even tested for jitter on the ILDA pattern where the draw exists the circle and starts on the detail above it. If that bight, unblanked, does not jump about, then jitter is not a problem, nor is noise. (Pointing accuracy may be, but in smooth continuous motion, even WideMoves are ok with that.)

    Edit:
    I looked at the WideMove driver board, it's grounded to the baseplate via its mountings, as intended. I don't think isolation will help. (It's usually done to eliminate ground loops, whose result is slight sine-wave modulations of intended scan position, but that's a problem I've never seen with these). I'm certain that the warped-line problem is mechanical, because it dominates any motion that is either stopping, or trying to start, regardless of the signal strength. Another certainty is that if this were in the signals, then this system, usually used for audio, would act like a noise gate! It would mute all low level signals. That definitely never happens.
    Last edited by The_Doctor; 11-29-2013 at 06:12.

  2. #182
    Join Date
    Oct 2007
    Location
    Vancouver, Canada
    Posts
    327

    Default

    I looked at the WideMove driver board, it's grounded to the baseplate via its mountings, as intended.
    There must be different revisions as strangely none of my driver boards ground through to the baseplate, they're all factory isolated. Only the scanner block itself, which of course is connected to the driver. It took me awhile to figure that out when I was trying to isolate everything, but in the end I only had to isolate the scanner block itself.

    You're right it doesn't look like the typical 50 / 60 herts type issue, but thought it worth mentioning as one of the only differences in setup. But every set of my dt's are rock solid stable so if it's not signal or tuning related, then at least it's not likely a common issue amongst DT's.

  3. #183
    Join Date
    Feb 2011
    Location
    New Hampshire
    Posts
    3,513

    Default

    What I am describing is very subtle and requires you measure the spot and then the lines. Nothing is perfect. No matter how high the quality of your scanning system there is inevitably going to be some miss tracking as a line is over written many times each second. This will come from multiple sources. First, the mirror bends and twists under the accelerations of the motor. This will cause distortions and defocus and will only serve to spread out the minimal static point. Second, the motor will have some compliance both electrically and mechanically and as the accelerations are applied the motor will not have a 100% perfect correspondence. Last, there is the accuracy of the feedback loop from the position sensor. This is due to the finite resolution and the time delay that both lead to small pointing errors.

    These factors are present no matter what scanner is used and can be minimized, but not eliminated. Lower quality scanners with weaker motors, thinner mirrors and shafts poorer position sensor resolution and sketchier drive electronics are going to have more of this error. The lines from my projectors are VERY SHARP, but I think they might be made as much as 2X sharper with better scanners. If I am correct, that this is a universal problem, then it doesn't matter if you have spatially filtered OPSLs and 1/20 wave optics, this degradation is still going to be an issue.

    That's why I raise this now. Has anyone else made these measurements? Is this just my funky scanner?

  4. #184
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Quote Originally Posted by planters View Post
    What I am describing is very subtle and requires you measure the spot and then the lines. Nothing is perfect. No matter how high the quality of your scanning system there is inevitably going to be some miss tracking as a line is over written many times each second. This will come from multiple sources. First, the mirror bends and twists under the accelerations of the motor. This will cause distortions and defocus and will only serve to spread out the minimal static point. Second, the motor will have some compliance both electrically and mechanically and as the accelerations are applied the motor will not have a 100% perfect correspondence. Last, there is the accuracy of the feedback loop from the position sensor. This is due to the finite resolution and the time delay that both lead to small pointing errors.

    These factors are present no matter what scanner is used and can be minimized, but not eliminated. Lower quality scanners with weaker motors, thinner mirrors and shafts poorer position sensor resolution and sketchier drive electronics are going to have more of this error. The lines from my projectors are VERY SHARP, but I think they might be made as much as 2X sharper with better scanners. If I am correct, that this is a universal problem, then it doesn't matter if you have spatially filtered OPSLs and 1/20 wave optics, this degradation is still going to be an issue.

    That's why I raise this now. Has anyone else made these measurements? Is this just my funky scanner?
    Have a look at the lower right part of the image in the first link on this post in Sam's Laser FAQ.
    http://www.repairfaq.org/sam/laserdps.htm#dpsldd317
    Is that what you mean? In that case it's not the mirror warping as far as I know, but maybe the thin shaft bending due to stresses of resonant disturbances. I did that test with WideMoves. As it's a fast drawn circle with sines in quadrature, the other issue I've mentioned is probably not relevant here, but you can see that each draw of the circumference is slightly off-track. The scanners were running as fast as I could get them to go, short of producing distinct noise and chaotic behaviour. I wanted the fastest clean-running scan I could get to resolve the shape of 1MHz square wave pulses from my driver... I managed to tune the business such that pulses occurred in integer relation to the sinewave period, but as you can see I couldn't do much about the tracking innacuracy. It's not that bad though, not far off a deviation of about 1.5 times beam divergence of a TEM00 mode diode laser. And the spots appear to be round, suggesting that the thin mirrors weren't warping.


    m0f, I haven't got the system set up as I used to, it's been a long time... I'd need to change the whole computer hardware and OS just to run the Layla 24/96 and its WDM driver. Which makes me feel a bit weird asking other people to try stuff for me, but that's basically what I do need. I'm not really trying to solve the WideMove error, I think that's past my solving. What I really need most is to see what happens when people try my rotating line thing on their known-good systems. THAT will be extremely useful because it will guide me to a better scanner choice, and a certainty that if I encounter trouble, it will be trouble I caused. Nothing would easily make me happier than never having to talk about this WideMove problem again. I'm looking for a way out, not to keep returning to the same weary circle of hell.
    Last edited by The_Doctor; 11-29-2013 at 07:17.

  5. #185
    Join Date
    Feb 2008
    Location
    East Sussex, England
    Posts
    5,248

    Default

    Edit:
    I asked Bill (Pangolin) in PM about this, but I got a terse three-line statement that means basically that unless I used Skype he wasn't going to talk to me. I don't know why. He suggested email in the public post, but in PM said he hates long emails. Mine wasn't long, not a lot longer than the length of this post. Anyway, the point of this is that if a person wants me to buy scanners from them, they MUST answer this question. And if they do not care to do this, then I will buy my scanners from someone who does. I'm not that fussy, so long as I do not see such an extremely objectionable problem at the outset, because it's a lousy way to embark on a long and expensive course of action.
    I think what Bill meant was that he would rather be able to respond by email, rather than via a forum PM. He is often out and about travelling, and so logging into PL and using the accursed PM system on a phone is a bind he could probably do without.
    So, as PL is NOT a proper way to contact a company about their product, why not just drop him an email? I do all of my comms with Bill, Justin and Alexey by email, and whilst they might tire of constant questions, I do get a (sometimes terse ) response.

    If I end up having extended dialogue with anyone from any forum, I move it into the email domain asap, as its simply much easier to manage, and search and reference.
    Frikkin Lasers
    http://www.frikkinlasers.co.uk

    You are using Bonetti's defense against me, ah?

    I thought it fitting, considering the rocky terrain.

  6. #186
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Quote Originally Posted by norty303 View Post
    So, as PL is NOT a proper way to contact a company about their product, why not just drop him an email? I do all of my comms with Bill, Justin and Alexey by email, and whilst they might tire of constant questions, I do get a (sometimes terse ) response.
    Fair point. Got to admit it annoyed me, but for now I think the best thing I can do is let that motivate me to putting the problem here, asking for people to try that rotating line thing. Two reasons: 1, gives more people useful data, 2, gives me a shot at asking better questions come email time.

  7. #187
    Join Date
    Feb 2011
    Location
    New Hampshire
    Posts
    3,513

    Default

    Doc,

    Please forgive me, but I don't understand. I think you are talking about a specific tracing aberration with the wide moves scanners. I can't address that. The errors I am talking about are small enough that when hundreds of re-tracings are drawn there is the appearance of a wider illuminated line. The errors are not so course and repeatable that line separation occurs. Can you try the measurements I talked about? They are not difficult to do.

  8. #188
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Quote Originally Posted by planters View Post
    Doc,

    Please forgive me, but I don't understand. I think you are talking about a specific tracing aberration with the wide moves scanners. I can't address that. The errors I am talking about are small enough that when hundreds of re-tracings are drawn there is the appearance of a wider illuminated line. The errors are not so course and repeatable that line separation occurs. Can you try the measurements I talked about? They are not difficult to do.
    Did you see that image? While I was focussuing on the warped-line issue in this thread, when I saw your post it reminded me of something else. That's what is in the image. I'm fairly sure it related closely to what you said. I can't measure it. It's an old image now. But you can see it. What I can tell you is a few known measures to set scope for what that image shows: scan mirrors to wall = 210 cm, circle diameter = 60 cm. I can tell you that much because my room hasn't changed. I'm not sure this matters though, more important is the relation of divergence in the beam, to deviation of tracking between turns of that circle. I imagine that unless the rotor, shaft and mirrors are stiffer, similar deviations would be seen on most scanners when pushed to their fastest clean scan of a drawn circle for any given scan angle within their ability.

    Edit: Not exactly scientifically quantified, but by 'clean', I mean don't make them scream. If they sing a bit, they're likely not pushed so hard that erratic deviation is inevitable, but if they scream with a ragged edge to the sound, then all bets are off. I never let them run that hard because I doubt they'd look good or live long that way.
    Last edited by The_Doctor; 11-29-2013 at 08:10.

  9. #189
    mixedgas's Avatar
    mixedgas is online now Creaky Old Award Winning Bastard Technologist
    Infinitus Excellentia Ion Laser Dominatus
    Join Date
    May 2007
    Location
    A lab with some dripping water on the floor.
    Posts
    10,036

    Default

    Doctor,

    Bill possibly wanted you to use Skype for human reasons. As in solve your problem quickly over the phone. Not some evil nefarious plot, its Thanksgiving Weekend here, a chance at a long Holiday. He could solve it quicker with voice and get home to his family.

    Here, calculate your scan width, which is the first step to seeing what is going on:

    http://www.pangolin.com/userhelp/scanangles.htm

    Steve
    Qui habet Christos, habet Vitam!
    I should have rented the space under my name for advertising.
    When I still could have...

  10. #190
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Quote Originally Posted by mixedgas View Post
    Doctor,

    Bill possibly wanted you to use Skype for human reasons. As in solve your problem quickly over the phone. Not some evil nefarious plot, its Thanksgiving Weekend here, a chance at a long Holiday. He could solve it quicker with voice and get home to his family.

    Steve
    Not nefariousness. I never thought that. I was just annoyed at the impasse. It's as much down to me. I hate phones, so it was like the irresistible force and the immovable object.. As far as I know, Skype needs a phone plugged into the BT hub for voice transfer (and likely won't run on W98 anyway). I think my best shot is to find out what results people get with known good systems, drawing that rotating line. The results of a bit of feedback there will help me ask better, more concise questions later.

    Edit:
    Steve, I looked at the scan angles and all sorts already. I could revisit that but best not I think, for now. Higher priority is to find some scanners that look like not doing this (They'll need good accuracy for small angles). Then, when my eggs are not all in one basket called WideMove, I might find it less stressful to thrash over old ground. What I can say is that when scanning a circle I have seen the top, bottom and sides flattened. Increasing speed, or scan angle, will reduce the problem, but it is a large scale issue, and is never eradicated unless the combination of speed and angle is such that momentum forces a smoothing out of the mechanical cause of the mis-shaped draw. I have some saved pictures of scoped source, feedback, and photo of the scanned light. Sensor output matches very closely with drawn shape, implying that the scanner amps 'see' the distortion perfectly, but cannot correct it. The source is a very clean quadrature sine, drawing a perfect circle. Very few combinations of speed and scan angle ever got the real draw to look that good.
    Last edited by The_Doctor; 11-29-2013 at 10:41.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •