To the laser show software companies

[White-Light]

Norty,

My understanding is P. is working on two amp designs; the "credit card" due out in the next few months and designed to drive the 506 (conventional position sensor?) and the "digital" with a significantly further time horizon.

I believe the one I linked to on the Scannermax FB page may be the more expensive one. Its stated as being DSP (Digital Signal Processing).

However, you need to contact Pangolin if you really want time scales.

[Pangolin]

There are two 506Cs, one has the new patented Saturn position sensor on it, and the other has a backwards compatible sensor.

Hehe. BOTH have our patented position sensor. It's just that one of them uses parts from Digikey, and the other uses specialized custom-designed parts made in Japan and California. Signal to noise ratio on both are similar (and superior to conventional galvos). The difference is stability over time and linearity.

I expect that our pro sensor will not drift at all regardless of time and temperature. It's designed to be super-stable. The economy sensor uses parts suspended on posts. I expect that it WILL drift over time and temperature -- but I also suspect that it should not drift any more than the original Cambridge model 6800 did, and people have been blissfully using 6800s for years, so...

Another difference is linearity. Our PRO sensor is guaranteed to deliver 99.8% linearity even over very large angles. The economy sensor only has 99% linearity over around 48 degrees optical, and linearity is decreased for scan angles over 48 degrees. But in my own tests, I do not see any ill effects of this, plus everyone uses geometric correction anyway, so...

BUT, one thing I have learned is that "people don't know from angle"... The human body did not come equipped with any angle measuring devices. Because of that, people always way overestimate or underestimate the angle they are using. I've found this even with industrial engineers. I also developed a statistic over a period of around five years which proves that -- on average -- people in the lightshow business only use around 30 degrees optical (although those same people were convinced that they were using 60 degrees). Sure, there are people who are using 60 degrees or even more. But not most people. So...

The Compact 506 was designed to be a cheap scanner -- nothing more. It's not designed to be a great scanner -- just a cheap scanner. Nevertheless, in our side-by-side testing that everybody saw at ILDA, projected images were indisinguishable from Cambridge 6210 at the same scan angle and with the same content. But if you want a great scanner, then we'd recommend our Saturn series.

By the way, I too was surprised to learn of the *RETAIL* price of HB amps. As Norty might have said, the scanners are a small fraction of that price. I'm in the midst of developing the mods for Kvant dual-axis amp to drive these scanners with optimal performance. I believe the price of Kvant amps is lower. Still, if you can wait, then just wait for our own small amp (internal name given is "Mach Mini").

BTW, Pangolin are looking at competitive pricing versus DT40's it seems, but in a recent email to me quoted a retail price of $750-$900 for the DT40's.

I've responded to let them know that DT40's are not that expensive from some of the Chinese sellers.

Yes. Justin is doing his own market analysis. Ebay pricing for DT-40s was in the $750 to $900 range according to his findings, but we tremendously appreciate your feedback and market insights, which are no doubt more authoritative than a quick eBay search.

We are committed to making a cheap scanner with good performance. Once we get all of the market analysis in hand, then we'll know a lot better what the price will be FOR INDIVIDUALS. For OEMs, the price will always be lower because we amortize any support needed over at least hundreds of units if not more.

Bill

[planters]

I have and that is where my ideas about the differences in these amps and the time frames originate. The challenge is to remain patient while assembling a picture from bits and pieces of data. For example, the small step signal responses for the 506 and the Saturn can be found on the Scannermax web site. But, if you look at comparable scanners (ie 3-5mm aperture) from Cambridge and Nutfield, the 506 @ 250 usec is no where near the Nutfield @ 200usec and the best Cambridge is down @ 150 usec close to the anticipated Saturn 5 with comparable mirrors. I can't imagine that the amp isn't an important component of each system otherwise I would expect that a couple of generic amps would be supporting much of the industry. Power supplies are a good example. I like Mean Well, but there are a dozen others out there above and below these, as far as price, that deliver acceptable performance.

I want to upgrade a couple of satellites with something better than the 30K scanners I obtained from Omar last year. These have surprised me in that they work every bit as well as the DT I owned and unlike the DT they haven't broken. I have thought about EMS 7000s but the 6.8mm mirror for a 5mm clear aperture seems a little like buying a 2x4 stud when I need a 2 inch by 4 inch piece of wood. I will just have to wait to see how P. puts together a package.

[Pangolin]

For example, the small step signal responses for the 506 and the Saturn can be found on the Scannermax web site. But, if you look at comparable scanners (ie 3-5mm aperture) from Cambridge and Nutfield, the 506 @ 250 usec is no where near the Nutfield @ 200usec and the best Cambridge is down @ 150 usec close to the anticipated Saturn 5 with comparable mirrors.

In some ways, numbers on data sheets frustrate me... For one thing, all readers make the assumption that the publishing company is actually telling the truth, or measuring things in exactly the same way as every other company. This certainly is not the case on both counts!

The numbers published on our data sheets and Cambridge data sheets can be relied upon to be true (at least by themselves, without considering literally anything outside of those numbers). I'm not so sure about the others... For example, for years I saw companies who made scanners that looked like 6800s quote "25000 dyne centimeters per amp" and "0.015 gram centimeters squared" (exactly like the numbers on the 6800 data sheet), but then when I obtained samples of these scanners and run the numbers myself, I've seen the numbers be off by a factor of 2 (always for the worst). So many companies just copy the numbers from other companies, just to seem similar or competitive.

Other companies are simply specifying things differently. When we specify step time, we specify it as "stop to stop" -- meaning, the time the mirror is fully stopped on both ends of the step. Many companies measure the step time from 10% to 90% as would commonly be done for scope rise-time measurements. This is of course far faster than "stop to stop", particularly since actually most of the time is spent in the first 10% and last 90% of the motion!

And there are things that the data sheets flat out don't tell you. For example, no data sheet quotes something called the "Motor Constant". If you know how to calculate it, then you can calculate it yourself, but none specify it. If they did, we'd be the clear winner, followed by Cambridge, and other than us two, many others would look pretty bad in comparison.

Finally and most importantly, no data sheet tells you the resonance characteristics of the scanner, thus, no data sheet tells you what the MIRROR will do when you try to make the motor go that fast. 150 microseconds? Sure! (for the motor) But then what happens once you've made that step? Does the mirror continue to vibrate for ten cycles?

The numbers on our data sheet are conservative. For example, our torque constant on the Compact 506 is spec'd at 23,000, but actual units measure more than 23,500. I like to under-promise and over-deliver. But this is certainly not generally the case, and there are some companies whose data sheets themselves are contradictory (adding up certain numbers does not give you certain other numbers).

I can't imagine that the amp isn't an important component of each system otherwise I would expect that a couple of generic amps would be supporting much of the industry.

Actually this was the case for many years! For many years there were two dominant amplifiers on the market: The Turbotrack, and Cambridge CB6580. After that, some folks copied both of these and generally started making scanners too. That's why substantially all scanner amps look the same in their design, so all deliver similar performance.

I mentioned it in this video:
http://www.youtube.com/watch?v=BiaM0adidIo

The conventional moving magnet galvanometer was invented in 1976. The Turbo Track was released in 1993, based on ideas I developed for a servo in 1989. The original Cambridge 6800 was developed and sold in 1992 -- then the patented and improved version came out in 1995. But the original one was never patented, so that's what people copy. Basically, everyone is using 1989 servo technology together with 1992 scanners (actually invented in 1976).

Sure, the voltage transformer hasn't changed much in 100 years, but geez, doesn't anybody besides me think there are better ways of doing things?

Bill

[EyeMagic]

Hello Planters,

We now ship all EMS-8000 scanners with aperture optimized mirrors. The X mirror for 3.8mm aperture is 6.8mm wide, for 5mm aperture is 8.6mm wide and for the (new) 6mm aperture is 10mm wide. Material is fused silica for stiffness and lower weight.
The respective X and Y mirrors are modeled to have exactly the same inertia so they do not affect tuning. (X mirror is wide and short).
The new mirrors work with our new mount that sets the X scanner at -16deg from horizontal, so the beam hits the Y scanner with less elongation.

The 8000 scanners are the (strongly) updated 7000, with more efficient motor, stainless steel body and a new servo amp (rev.K).

Price remains the same.


All the best to all !

Tom Kamaras
EyeMagic

[planters]

Just the two guys I wanted to talk to.

Bill,

Your points are well made and more than plausible. At this point, all we (out here) have to work from IS the spec sheets. In some markets the experience of end users or third party testers can be used to evaluate the product or at the least a manufacturer (Pangolin/Scannermax) can run its product against a competitor's model and see how they perform in side by side testing. I have no doubt that you would report the results honestly, but without the amp/motor available as a integrated package (for sale/shipment) this will have to wait. I am now concerned that if I cobble together one of your motors and a compromised amp the results will be unfair (for Pangolin)
I am finishing a 10W projector for the purpose of hosting the best scanner available, when it's available. I live only 60 miles from Nutfield and was promised a side by side of their QS-5 and the CT6215. They also say their specs are very conservatively rated and their systems are lower cost than CT. I would love a short list of half a dozen real world tests of assembled and running systems that could distinguish a superior scanner for projection systems.

Tom,

Thanks for your response. Your company's website is... well,weak.There is no mention of the 8000 series and no product specifications. I only learned that you even have an 8000 series available from an off hand comment by a member here that stated he had one and that certain modifications had been made on the 7000 series and he thought that the newer model was better. He was happy. Now I am pleased to learn that you are using an 8.6mm mirror for the 5mm aperture model and I'm thrilled to learn that a 6mm aperture model is now available. The fact that you are using generous mirror sizes as well as a tilted, optimized motor orientation is a pretty big deal and you should show this on your site. I would love to take your 8000 series with the 5mm aperture mirrors to Hudson NH when I visit in January.

[mixedgas]

One could say that a real scanner shootout is also best done with a signal generator and a scope with FFT while looking at motor current and position signal.
While a test pattern will give you speed/angle information, it is not going to give you information on a damped, sudden, stop, ie settling time.

And how the amp is tuned matters, while doing this.

Steve

[swamidog]

his estimates are off by nearly a factor of 2.

Yes. Justin is doing his own market analysis. Ebay pricing for DT-40s was in the $750 to $900 range according to his findings, but we tremendously appreciate your feedback and market insights, which are no doubt more authoritative than a quick eBay search.


Bill

[norty303]

his estimates are off by nearly a factor of 2.

Thats basically what I told Bill/Justin in my email

[planters]

Steve,

I've heard you propose this before and I don't doubt the rigor of that approach. However, I phrased the question the way I did because while the information that would be obtained from that analysis would be useful for all the applications of these scanners including tracking and machining (and targeting). For projection, what I think would be more widely convincing is a few identical test patterns run on each scanner system that highlight the most important capabilities. The problem with the numerical approach is the effect of a particular value on real world images might be to over or under estimate the significance of a particular value. If the performance of one system is for all practical purposes the same or close to another then cost can be used to choose which one is the better value.

A test pattern could be evaluated for line width to test pointing repeatability. A grid of points could be measured for spacing fidelity (like a Hartman-Shack test) as the grid is enlarged to check for linearity and the points can be measured to compare the stiffness/ringing of the motor/mirror system.

Your point on tuning is a good one. The trade offs to achieve the best aesthetic compromise is going to depend on the operator and may not be reproducible. How important is this. EMS claims their scanners are factory tuned ( I have been impressed with their performance right out of the box). High end manufacturers like Nutfield and CT presumably deliver optimized systems to their customers. Is this going to be a problem?