Scannermax 506s

[Pangolin]

Neither would display graphics perfectly at 30k

Please let me know what "flaws" you detected, because I certainly didn't see any at my office -- at least none using typical graphics and at scan angles you'd typically be using.

Now, if you're talking about running "automated graphics" (for example automated text) at full size (on your scanners this is more than 50 optical degrees) and at 30K, not even Cambridge scanners will do that. Therefore the results will depend on the exact graphics you're talking about. In a video I made, comparing the performance of the '506 with DT40s, our scanners could be ran full size and at 30K with no noticeable "flaws". This test was done with a Kvant scanner amp, but the results should have been at least similar with your amps:
https://www.youtube.com/watch?v=8iWjpz2YEO0

The graphics in this video were well-made (hand made by arguably the best professionals in the business), and so such graphics should not show "flaws" on the Compact 506 or any well-made scanners. I can say the same about graphics made by our LC-Max, LC-Flash and BEYOND 3D.

Bill

[Pangolin]

Regarding temperature, is there a safe window to run these things at?

Remnance (i.e. flux density) of NdFeB magnet decreases 0.1% per degree C temperature increase. So an increase of +50C will decrease magnet flux density output by 5%. Running the scanners cooler will therefore increase performance somewhat.

Setting that concept aside, you never want to run the scanners at a temperature hotter than you can hold your hand on it. This is a "rule of thumb". The pain threshold for humans is at 50C. So if you can't hold your thumb on the body of the scanner or X-Y mount, then they're getting too hot.

Over the weekend, I was able to get a competitor's scanners there by simply running modest test patterns through them (without bolting the X-Y mount to a large projector base plate) but the only way you'd ever get ScannerMAX scanners hot is by running high-frequency square waves (like in some abstracts -- and I'm not sure software other than Pangolin would easily allow this) or high frequency sawtooth (such as those found in raster images).

On our Saturn series, it is the low coil resistance which leads to low heat (Heat = I squared R, so lower R means lower heat). On the Compact 506, it is a combination of lower coil resistance plus the high coil inductance, plus a kind of trick that we do in the output amplifier, allowing the conversion of "reactive power" into torque. We're exploring taking this concept even further right now...

By the way Stiffler, your amps don't have a coil temperature calculator, but instead have a "polyswitch" (polymer based positive-temperature-coefficient resistor that kind of acts like a resettable fuse) put there by Jian. That "polyswitch" should open before allowing the scanner amplifier to put so much current into the coil that it would heat up the scanner.

Bill

[Stiffler]

Hey Bill,

Regarding my comments about graphics not being perfect-

I ran all the test patterns through both projectors, and the grid pattern was especially tough for both scanners. I had to decrease the size of the projection substantially in both projectors to get the corners all square, but I'm sure I was still well above 8 degrees. This is more of an observation than a complaint, and since these projectors are mainly used for atmospherics, it's a moot point. Another observation I made was that neither of these scanners would project a perfectly round circle, even after I spent a fair amount of time working on geometric correction; see the last picture. Once again this is just an observation, but I am curious about the possible causes, and potential fixes. It could just be me doing a bad job of dialing in the settings, I don't have much practice.

There is one test in particular that I am looking forward to: position repeat ability. Specifically, the ability to hit the same spots (mirrors) with target beams over and over again. With the DTs, it's neccessary to reset them every time the projector is turned on, and it was suggested that this due to thermal drift.

[Pangolin]

the grid pattern was especially tough for both scanners. I had to decrease the size of the projection substantially in both projectors to get the corners all square, but I'm sure I was still well above 8 degrees.

Sure. That test pattern was not meant to be ran at 50 or 60 degrees, but it also wasn't meant to be ran at 8 degrees either. I think around 20 to 30 degrees it should look fine. But again, if you run this same pattern side by side with Cambridge model 6210 scanners (costing 10 times as much as the Compact 506 costs in volume) you'd see literally exactly the same thing.

The main limitation here is an "error square rooter" circuit which, in these simple amplifiers, is made with two diodes. Sacrifices must be made here because the voltage threshold of the diodes changes with temperature, and also a diode does not accurately represent a square root function. I'd be willing to bet money that with our little Mach Mini digital amp and the '506, these corners will square up at larger angles. Still, this particular pattern at 50 degrees? No way for that small, inexpensive scanner...

Another observation I made was that neither of these scanners would project a perfectly round circle, even after I spent a fair amount of time working on geometric correction; see the last picture.

Yes, well "you're doing it wrong" First, you should use a grid for geometric correction. Nothing else will do. Second, the circle will become what I call "diamond shaped" (like a rotated square) if you're scanning too fast. The reason is because the scanners will be running "ballistic" rather than within the controlled, small-signal bandwidth region of the servo. Basically you're just scanning the circle too damn fast and too damn big (at the same time). If you scan slower (or smaller), then this will round right out.

See this too has to do with the error-square-rooter circuit. It's not easy to put into words, but that's the cause of the problem. If a circle isn't round it means you're operating pretty far outside of the small signal regime and way into the large signal regime.

Imagine a car driving around the city who has to pick up his daughter from school and get her to piano class clear across town in only ten minutes. "California stops" will be involved

but I am curious about the possible causes, and potential fixes.

Both explained above...

It could just be me doing a bad job of dialing in the settings, I don't have much practice.

By the way, if it's our software that is causing this, then I would want to completely understand the circumstances. Our software (especially QS) is made so you just walk up to it and start playing. Unless your base settings are far off (and as Swami and Galvonaut will tell you, we try to hide these settings from you -- hehe -- for your own damn good) then it should only be extraordinary circumstances that something like QS would do this...

There is one test in particular that I am looking forward to: position repeat ability. Specifically, the ability to hit the same spots (mirrors) with target beams over and over again. With the DTs, it's neccessary to reset them every time the projector is turned on, and it was suggested that this due to thermal drift.

Well there are several drift sources here. One is that the components used on the amplifier are not low-drift components. Just look up the pots on Digikey and you'll find that they have 100 PPM temperature coefficient. The problem is, everyone in the whole world who makes scanner amps uses the same pots, because pots with lower drift are exponentially exponentially exponentially higher priced (take a look)! This is why we like Digital so much...

Another drift source is the position sensor inside the scanner, and components used. On the DT40 Jian uses a design from 1992. This is the only design Cambridge did not patent. I expect the drift of our Economy Compact 506 sensor should be much much better, although we have not yet fully characterized it.

Bill

[Stiffler]

Thanks again Bill for the very detailed answers!

[mixedgas]

Good News! The Amps arrived at the importer today. They are in New York. They will be here Tuesday if all goes well.

I asked the importer to open the box for pics, they look good. Two twists over the sample pair, a Polyfuse self resetting protective device instead of a Picofuse, and Rt Angle pots, so when mounted vertically you can adjust them normally.


Steve

[Stiffler]

No pics for us? BOOOO! lol

[masterpj]

110$ for a 506 scanset minus amps? *I just recall reading that somewhere.
I am interesting in buying a pair if that's the case.
I have an LM scanner or LW50k scanner amp I can modify for the job.

[Stiffler]

110$ for a 506 scanset minus amps? *I just recall reading that somewhere.
I am interesting in buying a pair if that's the case.
I have an LM scanner or LW50k scanner amp I can modify for the job.

That's the cost of the amps, the galvos are more.

[mixedgas]

One amp pair just left for Pangolin, the rest of them are coming here and should be here Tuesday.

Steve