Printable View
I am not going to be demonstrating my latest projector @ MIT this weekend and so I will make a video. The installed EMS 8000 scanners preform very well, but as you will see they are the limiting factor by a large margin in the final performance of the projector. I agree with kecked that a higher performance scanning system is required to go from here.
@Kecked, we had the Boeing guy do an analysis of 6215 with 3mm mirrors and he found that you'll get into wobble at around 45K. This confirmed what we are also seeing in real life. If you TUNE and operate the scanners to 30K, you will not notice this. But then the only benefit to having 6215 will be the angle at which you can project.
@Planters, I'll be acquiring EMS 8000s soon and will be able to fully understand the implications myself. (As you've learned about me, I never believe the hype...) Tom is indeed always improving. This much is impressive. He's never a stopped target. The only problem I have is the choices he makes at each stop are not very impressive (to me)... I'm sure once I announce my findings (which I typically do in private directly to him) he will work on them some more, and come up with EMS9000...
Bill
I keep hearing talking about these new scanners and how they are better than other 30K scanners because they can deliver that speed at larger angles. So what is the angle for the 30K pattern?
A few quick shots. Note shaft thickness of approx : 2.9 mm / 0.110 inch and that generous application of epoxy does not slow things down.
Steve
Thanks Steve. This picture shows a few things perhaps not intended. For one thing, it shows how the scanners can be rotated all the way around without breaking. (There is no "shaft stop".) Our actuators and scanners have two possible resting positions, located at 180-degree angle. One really cool thing to see is -- if you apply power to the scanners now, the mirrors will instantly flip into the proper orientation. Normally they will stay there too. In Steve's case, I guess he rotated them...
For another thing, Steve's pictures clearly show how we support the back of the mirror with the shaft.
And finally, one thing we haven't talked too much about is the "Darth Vader Shaped" X mirror. This allows us to have exactly the same inertia as the Y mirror, without increasing mirror width or length (either dimensional increase being bad for cross-axis wobble).
By the way, we're going to reduce the front dimension of the X-Y mount by around 0.1 inches, so the final package will be a bit smaller than what you're seeing here.
the bolt through mounting hole is a pleasant surprise.
just when i was getting good at making blind holes and countersinking in my baseplate. tell me it's sized for 1/4-40.
Quote:
Originally Posted by mixedgas
Great question. Perhaps you missed it, but check out this video, because your question is answered within the first few minutes:Quote:
Originally Posted by johnjack
http://www.youtube.com/watch?v=8iWjpz2YEO0
On these particular amps, we were able to get 14.5 degrees with the ILDA test pattern at 30K. That's impressive, and more than I've seen from Cambridge 6800s or 6210s driven with amps that don't use an H-bridge.
But I honestly hate the question of "angle with the ILDA test pattern" because, unless you've been around for a while, and read a number of posts, this might incline people to think that 14.5 degrees is the maximum angle for all patterns. This "angle with the ILDA test pattern" only governs a single feature of the ILDA test pattern. Most other features will continue to look good even if the angle is increased. I think this will be visible on the video I made for Norty here:
http://www.youtube.com/watch?v=gImSsPJCxw4
As can be seen in previous posts by Norty, these scanners can scan a lot wider than just 14.5 degrees. Norty is getting angles of 50 degrees and in the first video I mentioned above, the angle we were seeing is around 53 degrees or so.
For reference, if you accomplish an angle of 53 degrees, you will have a "one for one" projection situation. What I mean is, if you are 10 meters from the screen, then you'll be able to produce images that are 10 meters high and 10 meters wide.
Bill
Well there are a lot of mounting options including "from the top" and "from the bottom". (I just noticed on this mount we don't have any holes "from the front" but that is also typical for us...)Quote:
Originally Posted by swamidog
And yes, you can fit a #4-40 screw through those counter-bore holes, or I think M3 or maybe larger metric...
Bill
Mounting holes, two of them. One under the scanners, one through the block long ways from the vertical. Appear to be clearance holes for M3, which is fine with me. My company uses M3 for nearly everything, so I'm used to it. Its a nice thread pitch and easy to tap for. I've never had much trouble finding M3 at a hardware store, nation wide(USA). 4-40 would fit as well.
Both holes are top accessible. No blind holes, no need for bottom access. I tapped the one hole for 8-32 so it sits on a standard optical post. I was able to do that and retain the 4-40/M3 functionality. I would not do that for anyone else's mount, for heat sinking reasons. The 1/2" diameter post, two inches long to a large baseplate, was enough heat sink at normal room ambient. These run cool.
You can also unbolt the side and flip the entry by design. You have to interchange Motor+ and Motor- on the cable, on one galvo. That's it, no other magic required. So no need for finding a different left or right hand block if your design requires it.
Having no stops is a serious advantage, for shipping and handling. If the mirror is rotated, the amp surges a bit, and flips the mirror, and the galvo starts scanning. Which is a function of the position sensor design. If you over drive it, you can go a long way without fear of damage. This is great for setup, if your software does not retain the angle settings.
Also the mirrors are far enough away from each other that you cannot "crash" the mirrors into each other. This alone is worth the price of admission.
I rotated the mirrors to get good shaft views. The scratch on the top of the mounting block is mine, from clamping for a quick test.
I flipped the cables during testing to interchange the axis connections. The "Darth" Mirror's inertia closely matches the paddle mirror, so there was no harm done and the tuning was close enough, nothing oscillated.
These are the smaller mirror set, great for tight beams.
Steve
I think Imeasured the through holes at 3.3mm, which meant they were ideal for tapping for M4*0.7, which I did and screwed in from underneath.
There are also a series of smaller blind threaded holes on mine, but not quite M3 - so must be an imperial size.
I chose to screw from underneath, partly because getting a hex bolt through the centre hole from the top would mean removing both galvos from the block.
If this hole was moved forward (toward the 'front' of the block) and to the corner, it would be possible to access it from above with both galvos in place, and put it more diagonally opposite the other hole.
However, I 'think' it might be placed thusly, because you can fix the centre, and rotate the block for final alignment, and the beam should still be centred on the mirror.
P.S. without wishing to derail this thread further, but related to the glut of threads and discussion about 'speed' and 'kpps' currently, do we need to revisit 'why' we use the ILDA pattern at 8 degrees and 30k as a 'standard'?
I thought it would be obvious that it is simply a common benchmark by which to judge performance, which tests a number of different parameters/properties, but it would seem that some people think that it defines max angle for graphics, or isn't then applicable when not scanning beams at larger angles, and other such strangeness.
I liken it to some of the computer graphic chip benchmark programs - it taxes a scanner set in a particular way, that is consistent and repeatable across any scanner platform.
I seem to recall there was a similar derailment discussion on the ill fated EMS thread of recent history.