The Curve for a 6200

[mixedgas]

Attached is a factory plot that is roughly, very roughly, available Angle vs Sinewave frequency for one, tiny , fast, galvo with a very small low inertia mirror, optimum tuning, and no mirror mount. A triangle wave might be 60% of This and a Square Wave 25-35% of this.

Steve

[rns0504]

Attached is a factory plot that is roughly, very roughly, available Angle vs Sinewave frequency for one, tiny , fast, galvo with a very small low inertia mirror, optimum tuning, and no mirror mount. A triangle wave might be 60% of This and a Square Wave 25-35% of this.

Steve

These are mechanical degrees I believe? I just ran the numbers off their datasheet and assuming a max body temp of 50C, the "theoretical" calculations we use match pretty well.

We had an engineer do some theoretical calculations, some simulations and verified with some "real world" tests to make this page below. It's not perfect, but its good for estimating... needs to be updated with new 15 and 20mm scanners...

https://scannermax.com/pages/performance-calculator

[buffo]

Hey Ryan - I was messing around with the performance calculator you posted (https://scannermax.com/pages/performance-calculator), and I'm getting inconsistent results.

If you enter an angle in the calculator on the left, the resulting frequency value does not match the plotted graph on the right. If you enter a decimal value (like 2.5) for the angle, the results are even more inconsistent.

Is this a bug with the calculator, or am I using it wrong?

I started playing with this after thinking about the ILDA test pattern, and specifically, the fact that the size of the center circle is ~ 31% of the full pattern width and height, meaning that if you were scanning the full test pattern at 8 degrees, the circle would be ~ 2.5 degrees. Thus, if the scanners were perfectly tuned to 30K, that's 30,000 points per second divided by the 12 points that make up the circle, or 2500 Hz.

So, ignoring the fact that there's already significant positional distortion at 2.5 degrees, as the scanners never make it all the way out to those points that make up the circle, I would expect the calculated value (and the plot) at 2.5 degrees to be less than 2.5KHz.

At 3 degrees, the calculator reports 1250 Hz, but the plot (Compact 506 - 3 mm, Sine Wave) shows 2628 Hz. At 2 degrees the calculator reports 1481 Hz, compared to 3304 Hz on the plot. And finally, at 2.5 degrees, the calculator reports 1467 Hz, which seems far to close to the result for 2 degrees.

Am I missing something obvious here?

Adam

[mixedgas]

Ryan, Thank You for the Link...

Yes, Mech Degrees,

check your email.

Steve

[mixedgas]

Adam gets it. I posted this to help with technical folks, usually who don't understand why galvos do not have an infinite bandwidth. We don't teach inertia in high school physics in America anymore, for the most part.

A couple of times per year I need to explain why the test pattern is what it is. It is not easy.

Steve

[buffo]

We don't teach inertia in high school physics in America anymore, for the most part.

Lol! Priceless....

Adam

[JohnYayas]

The graphs are enlightening. I knew there were limitations and dealt with them in software but I have never seen it plotted. Thanks.

[rns0504]

Adam,
Thanks for pointing that out. Something seems to be wrong in the interpolating "algorithm". I had our webmaster remove those portions for the time being. The graphs themselves should be approximately correct, within the stated margin of error.
As a sanity check, see below from yesterday, where I ran a set of C506 with a "very generic" 30K tune, not optimized at all, at about 2.8kHz with 2.5 degrees optical. Getting a bit warm for sure, but still getting the job done

Steve,
Thanks for the email and documents - I will have to check it out soon!

[buffo]

Hi Ryan,

Thanks for checking into that. I had a feeling something was off, because the plotted graph seemed to match everything I had ever heard from you (or Bill) regarding the performance of the 506s.

Also, way to go the extra mile by setting up a custom bench test to verify this. The screenshot you posted is the first I've seen. (I really need to make another trip down to Orlando one of these days so I can see the new building and have a peek at all the shiny new toys you've got in the shop, lol!)

Regarding the indicated "servo board temperature" of ~ 50 degrees C, is that temperature sensor on the amp, or is it on the top of the scanner itself? (I would think that scanner body temperature would be the more crucial measurement...)

Adam

[buffo]

The graphs are enlightening. I knew there were limitations and dealt with them in software but I have never seen it plotted. Thanks.

Dude! Haven't heard from you in ages - nice to see you're still in the game.

Can you elaborate on how you compensated for the limits of scanner performance in your Spaghetti software? Are you talking about setting a hard limit on the magnitude of change between any 2 points? What would that look like? (Interpolating extra points between them to reduce the delta?) Genuinely curious...

Adam