New EYEMAGIC Scanners EMS7000

[planters]

Back on Feb 20, I asked the question; " Its been a couple of months since the group buy and I'm wondering if there has been a review of the performance of these scanners?"

Actually, had to ask again a few days later and then the discussion began. During the interval around the turn of the year the major topic was "what PSU to get?"

I still think it's strange that with so many participating in the group buy and with other large companies taking delivery of the new scanners, there has been almost no discussion of their comparative performance. If I had gone ahead and purchased and received one of these you better believe I would have had it on the bench and powered up as soon as I had a free evening. I own an EMS 4000 and a simultaneous comparison would be easy even if it wasn't quantitative.

The discussion seems to be drifting into a bit of an emotional argument and that's unfortunate, but so what. Why don't we begin to focus on some plain facts, pics and testimonials? I'd love to see them.

[Pangolin]

I kindly ask Mr. Benner to copy/paste the text from my emails that says "how dare you question me'' (OR EVEN IMPLIES THAT) , as he stated in his post.

If he finds it, I will personally ask a BIG ''SORRY, I AM WRONG''. But if not, then he has to leave the conversation about EyeMagic and Tom Kamaras and come back next time he has something true to say about others.

Or even better, why doesn't he post the entire conversation that we had regarding the new scanners and let the readers decide ? I kindly ask him to do that.

I am traveling now, as is Justin. We are preparing for the Prolight & Sound tradeshow, and then we will be going on a tour of most of the factories and studios throughout Europe. We'll be out of the office a total of 32 days! As such, I don't have access to the email thread in question.

When I return to the office, I will be happy to post the entire email thread. But before I do so -- Tom, are you absolutely sure you want everyone on PL to permanently be able to see the attitude you expressed in the email thread? Take another look at the thread, refresh your own memory about it, and then let me know.

Maybe all this thing is because Pangolin wishes to make a cheap 60K scanner and we made it first ?

First of all Tom, we don't wish to make anything "cheap".

Second of all, Tom, based on the set that I received, it does not appear as though you have made a 60K scanner.

The numbers (3,8mm and 5mm) for the beams are what our customers have said that they have used with our mirrors with perfectly acceptable results

So, in other words, unlike Cambridge and other companies, who publish specifications based on science and what the materials can actually do, you publish data based on "what our customers have said they have used"? To me the biggest problem is that many customers don't really even know what they have!! It could be that customers *think* that have a 5mm beam, but in reality it might be smaller, particularly shortly outside of the laser where the beam is actually still focusing (i.e. related to the Raleigh range of the laser).

I end this post, asking for one favor: Connect an LD2000 to a set of EMS-7000, display the ILDA test, the hysteresis test, the lasermedia test, the grid test or whatever test and just publish the photos. After all, what your customer wants to see is sharply displayed images, not forum posts or ''Google hits...''

Actually Tom, shortly after I received my pair, I did a complete report about them, with all of the specs, microscope photos, even lubrication specs of the bearings. I held off on publishing my report because it would be seen as a "big guy going after the little guy" or something like that.

Tom, I'll publish the report if you'd like, but it's not very flattering...

I am sure there are some people reading this who think that this is all about competitiveness, or that I'm just a mean guy. That's not the case!! What gets me is that I expect published information to be based on PROVABLE SCIENTIFIC FACT -- not marketing hyperbole.


Best regards,

William Benner

[j4cbo]

A single bearing in our scanner costs as much as your entire scanner.

I take issue with this sentence. Your scanner does not exist. I understand that Pangolin has been engaged in R&D related to designing a scanner for well over a decade, but there is no product that is released that customers can buy right now. It's just not possible to compare a shipping product to vague research activity.

You said you "prefer clarity", but your supposed science there is also about as clear as mud. As I understand it, here are your claims:
(1) The inductance (relation of I to dV/dT) of the EyeMagic scanners is 140uH - fact
(2) Regarding inductance of the drive coil in a galvo, "lower is better" (arguably true) and "100 uH is about the highest inductance anyone would ever want to see" (really?)
(3) The permeance coefficient of the EyeMagics differs from the Cambridge galvos. You claim that this is "related to how well a scanner can handle heat", but this is at best a second-order relation.
(4) [ ... many missing steps; name-dropping of test equipment ... ]
(5) Therefore, "by the numbers", the EyeMagic scanners can only scan the ILDA test pattern at 40kpps.

You haven't proven anything. Bill. Not that there aren't valid complaints about some of EyeMagic's numbers - the beam aperture vs diameter issue in particular - but really...

Once you have the specs, it's just a matter of applying a few simple mathematical formulas to see how fast a scanner can go.

Would you care to provide, or at least enumerate, these formulas? Can you say in more detail where exactly that 40kpps number came from?

[drlava]

Well, this has become a really interesting discussion. Apparently they did analyze this quite in-depth, to the point of reverse-engineering. However, given that electromagnetic actuation is one of Bill's loves, it's not all that surprising - from an intellectual or a competitive point of view. What I'm hearing is that with the rotor flatting, at small scan angles before the torque drops 'too much', due to the triangular torque-angle curve, perhaps these are very fast, maybe even 60k fast, but at wider angles it falls off to 40k. Is that right? I can imagine this would be an acceptable tradeoff for many applications, if accurate, even if it's not 'in the way scanners should be tested'.
The report, including scan angles, lube oil, permeance calculations, etc would be quite interesting and fun to read for us techies who aren't necessarily up with the state of the galvo design art.

[planters]

Bill,

If you have independent test reports of the EMS 7000 scanners, you should just present them. It isn't really an issue as to whether Tom wants this or not.

AT THIS TIME why discuss the Pangolin development work when it's possible to compare these delivered scanners to currently available models from Cambridge, China and even from Eye Magic, HEAD TO HEAD. Ultimately, sophisticated test equipment will pinpoint the reason for the performance limitations, but for the vast majority of the end users it is the relative real world performance that counts. We need to see some activity in this area.
Why is it that we have threads on this forum that discuss the most minute progress with an important red diode on an hourly basis (because it's extremely interesting), while there is almost complete silence on these new scanners for months! I'm not suggesting anything nefarious, it just seems weird.

[heroic]

I have a set of EMS4000s for which I paid a hair under $1700. Given that I can buy the bearing for an ultracentrifuge for about $700, and that's designed to turn at over a quarter of a million RPM for several years at a time, without need for servicing and with necessarily tiny runout (concentricity better than 1 micron over the 18mm bearing) I have to ask, Bill, what the heck kind of bearings are you using?

[Solarfire]

Starting this post with the bottom line. “This thread is absolutely useless without pictures”.

First of all, datasheets for such systems (closed loop positioning) are at best a guideline and by no means a guarantee for superior performance. The only real data relevant too the subject at hand is the visual aspect. IMO the only real test of relevance here is a side by side test, best with an ILDA Y adapter, feeding both systems with the same DAC and same software. At the end of this test the cost/performance factor will clearly answer all questions which have bin raised here.

Just a few analogies to the worth of datasheets vs. actual performance. For those familiar with the JBL 100 speakers, their measured data is crap compared too many other speakers, but a simple listening test overwhelmingly shows superior performance in clarity and dynamics. The JBL100 is legendary for its outstanding performance compared too many other speakers costing many times more. To this day the L100 is still sought after, in some cases for more than the original cost from more than 35 years ago.

Tube amp vs. silicon. To this day there is still no substitute for the qualities of the electron tube, the only major reason for using silicon is the cost. Tube amp data vs. silicon also results in crap, but a listening test quickly reveals qualities which just can’t be reached with silicon.

So back to the bottom line, I finally got my PSUs and will be doing final assembly and testing here shortly (with pictures).

During this recently developed discussion about kpps, raised the question for me, how exactly is kpps defined? Yes I know kilo points per second, but what is the distance/angle between 2 points? Is this a static or dynamic value? Is acceleration and deceleration somehow accounted for?

[planters]

Solarfire,
Glad to here that you will be testing these scanners soon. Were is everybody else?

Furthermore, on the kpps question, many of the industrial scanners are not presented with kpps statistics. Is there a simple equivalency between kpps and some more broadly recognized performance measurement?

[Solarfire]

Solarfire,
Glad to here that you will be testing these scanners soon. Were is everybody else?

Yeah.. I had a lot of pre work/development to do for these; I developed some quick swap precision sleds for these so I can drop them into different power level projectors. Scanners, amps, IRISColorSafe and PSU all integrated on a 220mm x 150mm footprint.

As to where everybody else is, good question. Has anyone not gotten theirs yet?

Furthermore, on the kpps question, many of the industrial scanners are not presented with kpps statistics. Is there a simple equivalency between kpps and some more broadly recognized performance measurement?

Another good question that I couldn't find an answer to. I mean there is a major difference if I have a 1000 points, point on point on a line or 1000 points unequally spaced on a square. So unless kpps is defined somewhere it’s actually a useless value that can be laid out as one pleases. I couldn’t find a definition anywhere.

[j4cbo]

So unless kpps is defined somewhere it’s actually a useless value that can be laid out as one pleases. I couldn’t find a definition anywhere.

"ILDA kpps", referring to a scanner, means that the scanner being fed the ILDA test pattern at that point rate will draw the pattern with the circle just inside the square. The ability of a scanner to do this of course depends on the scan angle, so a scan angle must be specified too. This has a rough correlation with the -3dB point of the scanner's frequency response and its slew rate limit, but those aren't as intuitively easy to deal with when asking questions like "how quickly can I run my DAC if I want to draw this image with these scanners?" The test pattern is an attempt at quantifying scanners' response to something approximating typical laser graphics.