Announcing: "The Radiator"

[Lasertize]

30Kpps after some bench tests corresponds to a 2400 Hz small angle bandwidth response at the very small angle in the circle of the test pattern, , and as you open up the scan angle angle that frequency response falls off like mad in a non-linear way.
Wide angle jumps are slow, small angle jumps can be fast. The galvo shaft has inertia, and takes time to respond. It resonances that can cause bending mode failures and bearing failures if the galvo is abused.


There is not a linear relationship between PPS and displayable abstract Cycles per Second, that is dependent on scan angle and the galvo's non-linear frequency response. The galvo is, in simplest terms, a complex low pass filter with several integrals, and with several shaft resonances.


So, say you have an 1000 data point frame, at 30,000 PPS your time for one complete point is 1/30,000 = per point time in seconds, is 3.33 x10-5 seconds..., thus one 1000 point frame is completely scanned every .0333 seconds.


PPS is simply used to tune to the galvo amp to a common response to the standard test frame at a specific angle , so that images from different users are interchangeable.

So 30Kpps means the galvo can be given a point update rate of 30,000 points per second while scanning the ILDA test pattern at 8 degrees.

ILDA rated speed is related to making the feedback and PID loop tuning of different scan system hardware consistent, so artwork is interchangeable. Along the way it became a marketing point, as the first widely used feedback galvo system was 12K PPS, then 20Kpps, then 24Kpps, then 30Kpps, and a few systems have went faster. So for a long time the standards were 12K and 30K, and finally the Asian copies of scanners became less expensive, so most users have a 30K scan set.

Very little of this relates to how an abstract console behaves. In fact, with a console, one has to be careful not to stress the scanners with fast waveforms. You listen to the scanners, monitor their temperature, monitor their current draw, and keep the scan angles small when using a console, as the first thing the clone galvo manufacturers removed from the galvo amps was the "coil temperature calculator" circuit that protected the galvanometer scanners from overheating.

Only two manufacturers of scanners publish a scanner response vs frequency curve table, which is a complex data table that might tell an engineer a few things, but tells an artist basically nothing.

The ILDA test method was a compromise designed by a panel of very bright people to allow laser show companies to interchange images by tuning the hardware to a common response standard. It was NEVER intended to convey the complex response of a galvo scanner. PERIOD.


Steve

my IQ was 170 now is 171,
thank you MIXEDGASS that is the best so far here so to dumb it down, we were all way off the page.
good schooling there sir!

[mixedgas]

Factoid, for most newer Galvos, the stator or field coils are a few thousandths of an inch away from the moving magnet rotor. What often happens when you overdrive the Galvo is the copper wire and rotor expand from heating and touch each other, resulting in coil damage or shorts. The coil is wound direct on the steel laminations, the only spacer is the wire's enamel insulation in many designs. I have a friend who just lost a scanner (very low cost model) to the ball bearings failure to be fast enough to follow his overdriven abstract waveform. A bearing ball skidded, thus gouging the bearing race , deforming it. The shaft can only partially rotate now, so the Galvo is useless. It now sounds like a croaking frog when it spins.

When it comes to scanner driver amps, coils, rotors, and bearings, you get what you pay for.

Your not way off the page, the model you were taught in school/online is a vast simplification.

I've been at this a while. I had the Privilege to be taught by industry pioneers and the second generation, so to speak.
I also ate lunch more then once with the factory engineers of three scanner manufacturers at ILDA conferences.

Sadly you have to destroy a pair of scanners some times and tear them apart to learn what they can, and can't do.

Generally you will hear and see something just before they start to destroy themselves. Make sure your waveforms are not fast square waves with very few points, for one thing. (keep scan angles small, avoid fast square waves, or sine waves above the first resonance. )

Many Galvo amps have dual tunable notch filters in the feedback loop to allow you to use the full bandwidth with less fear. Some middle of the road amplifiers have PTC varisters as pseudo-resettable fuses in the output stage. Your image size will become smaller, erratic and distorted if these are overheated as you overload the system. It serves as a warning that you are pushing too hard.

I only end up typing this series of answers about ten times a year. :-)

One thing you can do is put image size potentiometers on your projector, to form a means of limiting users from driving the scanners to the "stops", though not all scanners have mechanical limits these days. The scanner gets really suicidal if you try to drive it past sixty degrees optical, which is 30 degrees mechanical. I've watched more then one new operator try for 180 degrees... Optical deflection is twice the mechanical scan angle, due to physics. I often start with the image size potentiometers turned way down when creating new images, no more then 5-10 degrees optical.

For consoles:
Scan a small angle circle at say 10 degrees optical. Start at say 30-60 Hz and make sure your waveform has at least 256 XY point pairs, if it is not truly analog. . Slowly increase the circle's frequency or PPS rate. DO not touch the image size or abstract amplitude knobs while doing that. Note the frequency at which the image size decreases say 30 percent. Stay below that frequency in all of your X-Y component waveforms and resist the temptation to open up the amplitude knob to increase drive to restore the image size. That should keep you fairly safe till you have an idea of what not to do. Is this is Limiting in artistic possibilities, Oh yes. Is it 100% protection from mistakes, No.

Steve

[Lasertize]

Mixedgas,
I have a Clubmax3000/fb4, on some of the show abstracts, the Y scanner drops the image to the ground leaving the X an unintelligible image on the ground. I have dropped the scan angle as low as 15* even that it has gotten better, it still happens. scanners are 40kpps. As long as I don't do the more complex abstracts, they work fine. Aaron Pangolins tech support manager thinks maybe there's a amp problem with the Y scanner.
It's a new projector from KAVANT bought in July 23/20. how do I prove the failure? the ILDA test pattern tests fine.

[mixedgas]

Mixedgas,
I have a Clubmax3000/fb4, on some of the show abstracts, the Y scanner drops the image to the ground leaving the X an unintelligible image on the ground. I have dropped the scan angle as low as 15* even that it has gotten better, it still happens. scanners are 40kpps. As long as I don't do the more complex abstracts, they work fine. Aaron Pangolins tech support manager thinks maybe there's a amp problem with the Y scanner.
It's a new projector from KAVANT bought in July 23/20. how do I prove the failure? the ILDA test pattern tests fine.

Drops the image to the Floor???? yes/no? That means someone is abusing something, generally, and your close to system destruction. Sounds like something in the system is close to failure and it is protecting itself. Generally the galvo slamming full angle to one side is a loss of one of the PSU voltage rails, so the amplifier looses position control, can't center the rotor, and the deflection is extreme. One would hope you physically position or physically mask your projector at a show to prevent in-advertent audience exposure. I never trust the galvo hardware to do that function.

Potential Answers: Scan angle too wide and system overheating, image has poor design, PPS set way to high for image size, Galvo PSUs are undersized, PTC fuse is tripping from too wide or too complex an image, or PSU overcurrent limit is tripping, scanner is resonating from frequency components above resonance. Bad image position offset / image size in Zones or show script is in there too, This list is not all inclusive.

Send a video to Pangolin.

Pangolin has a scan angle vs distance table on their website, probably in the WIKI. Saves you from doing trig by using a tape measure. How wide were you scanning? What show content were you using? Is this an image you made yourself?

Steve

[Lasertize]

Drops the image to the Floor???? yes/no? That means someone is abusing something, generally, and your close to system destruction. Sounds like something in the system is close to failure and it is protecting itself. Generally the galvo slamming full angle to one side is a loss of one of the PSU voltage rails, so the amplifier looses position control, can't center the rotor, and the deflection is extreme. One would hope you physically position or physically mask your projector at a show to prevent in-advertent audience exposure. I never trust the galvo hardware to do that function.

Potential Answers: Scan angle too wide and system overheating, image has poor design, PPS set way to high for image size, Galvo PSUs are undersized, PTC fuse is tripping from too wide or too complex an image, or PSU overcurrent limit is tripping, scanner is resonating from frequency components above resonance. Bad image position offset / image size in Zones or show script is in there too, This list is not all inclusive.

Send a video to Pangolin.

Pangolin has a scan angle vs distance table on their website, probably in the WIKI. Saves you from doing trig by using a tape measure. How wide were you scanning? What show content were you using? Is this an image you made yourself?

Steve

MIXEDGAS,
this the link to the scan failure:

https://www.youtube.com/watch?v=ahnqEFEyu 4

and this the link to the same show one week before scan failure:

https://www.youtube.com/watch?v=VYHpSq1EINc&t=59s

scan angle is 8* at 6 feet distance Kpps is set to 20k as a pangolin recommendation and the abstracts are from other pangolin shows provided in the software purchase.
from the great real information you have provided, it looks like an amp failure to me. I am glade the projector is still under warranty.
Thanks Steve
Emory

[mixedgas]

If that was scanned at 8' optical I'm not so sure you should have a problem. Not enough info to tell, but other then the one very dense abstract I'm not seeing anything that screams "scanner killer". I do see some frames / tracks that are very dense for a 20K pair.
Ok, I have many projects to continue and I'm only on line as I wait for things to run in the lab at work on a late shift. I will be back in a few days.

The video linking to the "scanner killer" came up unavailable. Probably should take further projector failure questions to the LSS or Help forum, and leave Radiator more or less unblemished.

Steve

[tertior]

Thank you for your explanations; I get it. The confusion came from the Radiator manual page 19: "It is recommended to use a speed between 30Hz and 60Hz to start with. High speed playback can put more stress on your laser projector and also reduce the detail and resolution of the final projected image." But when I see the Shape A photo and the speed is 428.43 HZ hence my question. Now it's clear

Thank's

[shivahoj]

Hi guys .
How do i sync this via usb to a midi dignal from a pc. Also i see it should sync to audio but when i run a signal in i get a low level on tne audio test and no visible signs of sync. Any help?
Thanks
Frazer

As said in the manual, the USB port is NOT carrying MIDI signals. You have to use the standard 5 pin MIDI connectors. Get yourself a cheap MIDI<-->USB converter cable, this works for me without any drivers in Windows, Linux and Mac. Apart from that, I also am finding out about controlling Radiator, with mixed results.

About the low audio level: Most consumer audio devices may not generate enough output voltage to sufficiently drive the audio circuitry. I tried this with several MP3-players and mobile Phones. These are made for low impedance headphones.
Now I am using a MOTU 828mkII USB interfacehttps://motu.com/products/motuaudio/828x (basically an 800€ sound card) that can output a whopping 10 Volts analog on 12 outputs, and also has the added benefit of being DC-Coupled, so You can drive Radiators CV inputs with DC-Voltages or Milli-Hertz signals. (right now I only could generate 1 Hz signals in Audacity, I don't know any free software to do this, but there is a commercial "silent way" software from "expert sleepers", a company that builds eurorack devices.)
The above works nicely for me, I could easily overload Radiator's Inputs, but the "best" signal is when I turn the Warp Type to "VU Fast" and the Warp amount to about 50-70%. Then the figures react nicely to the music.
Maybe You could use the output of an amplifier (even the speaker outputs, WITH EXTREME CARE !!), to drive the audio inputs of the Radiator ?

About the "sync":
Can you please explain what You want to achieve?
There is no syncing of anything with audio. Radiator can REACT to music, but not sync (by itself).
But do You know You can hold the Shift key and tap the ON(Tap) buttons on each of the LFO's to punch in the speed?

[shivahoj]

wow, I read the entire radiator manual, It's impressive, packed full of features. well thought out and designed! Roberta, M certainly likes hers!
What is the max time length of a recording allowed?
I am finding it hard to cash in my metal assets knowing the value is increasing every day, how about that Platinum, all ready over 1k$ an ounce, It's out performing gold and silver, regardless I can't seem to let any of it go knowing the future predictions values, even my bundle of gold backs are worth more than I paid for them.
I am pinching pennies so by January 2021 I'll be able to buy your dream machine "Radiator" YAE!
Reading the Radiator manual on the MIDI in and outs, was not sure if my APC40MKII would plug up to it, which would make my wife user assessable to the Radiator if it will work in that manor (she is blind).
the APC is a good lay out for her to use with out having to Braille every control point, this may be stupid Question, it's a honest one.

I am working on the same problem, I also want to control the radiator via APC40MKI and MIDI(and QLC, a free DMX lighting program ), right now I am in the middle of understanding the MIDI NRPN concept.
AFAIK there has to be a program(or device) that reads the APC's Midi input, and translates it to 3 or 4 Byte Commands sent to the Radiator. I haven't tried it, but You might want to look at Bome Midi translator.

Some of the things in radiator can be controlled with regular CC Commands, e.g. sending control data (0-127)via CC1 would change Hue, etc.
but others would need a more complex message:for the LFO1 impact on the color Hue level, you need to send 9 on CC98, then 1 on CC99, then 0-127 for the actual control on CC6, or, if a fine resolution is needed, 0-127 fine grained control on CC 38, and only after that the coarse value on CC6.

i have not yet figured it all out (only about half of it), but this is what I think is it. I have managed to connect a simple MIDI controller(http://www.doepfer.de/pf.htm) right to the radiator, and can control the simpler aspects of radiator with it.
I will continue working on Radiators remote control abilities, feel free to DM or ask in thew forum if questions arise.

[shivahoj]

Since I am planning on a slightly larger setup with QLC+ (DMX and Artnet, several moving heads, foggers and Wash lights), Video(Resolume Arena) and a Radiator(via ILDA to a Laser, plus the HDMI signal fed back into resolume with a Blackmagic capture device), I want to automate as much as possible on Radiator.
Right now , in Corona times, I have a lot of time thinking about new ways to do stuff...

I already tried to control Radiator via Network, see this post: https://photonlexicon.com/forums/sho...184#post355184
but swamidog told me this is not the easy way to go.
Now I hooked a simple MIDI fader device(http://www.doepfer.de/pf.htm) to the radiator, and I can control SOME things on the radiator. Most direct CC-controlled items work fine, these are all that have a number under CC in the manual.
I didn't check them all, but the Bank Change (CC32) does not work as expected.it should accept values from 0-9, giving access to 10 Banks of 100 presets each. No matter what I send on MIDI CC32 here, the bank does not change on the tiny screen (or maybe it don't change anyways, because it is organised on 100 Banks of 10 Slots each?)
Furthermore, in each bank, how (with which MIDI CC command) do I select one of the 100 presets and put it out to the laser?
It NEEDS to be CC, because the software I chose as the central hub(QLC+ 4.12, because it is free and available for Mac , Win and Linux) can send out CC messages OR PC(program change) OR Note Velocity, but ever just one of the three. And I am stuck to CC, because I want to automate all the CC and NRPN(non-registered Parameter Number) values, and have to use CC for that.