Avoiding scanner damage

[Zoof]

Is it possible to do damage to scanners like the dt40pro's when applying some bad signals to the amp through a DAC like the easlylase?

I understand that galvo's, especially the high speed ones, can easily be damaged by wrong tuning them and by signal spikes. Assuming they are setup propperly, tuned right and the DAC is good, is it possible to cause damage the scanners through the input signal... for example by scanning a square at the max size defined by four points, one in each corner, and scanning this at 40kpps?
My idea is to write my own code and experiment with abstracts and my own animation code. As always experimental code is subject to errors or just 'bad ideas'.

thanks, Matthijs

[Pangolin]

Hi Zoof,

If the scanners are heat-sinked properly AND if both the scanners and drivers are designed properly, then the driver will protect the scanner from damage. BUT, this makes a few assumptions...

It is absolutely possible to damage these moving magnet scanners if they are not properly heat-sinked AND if you input a signal that has a high duty-cycle. Examples of a high duty-cycle signal are raster images (very rare in this business, especially if you don't have an LD2000 system) and abstract imagery with a lot of squarewave content or high frequency continuous sine waves.

Unfortunately, the 6800 scanner (and its derivatives) do not have a thermistor, or any other temperature-sensing mechanism. The scanner driver makes some assumptions that the scanner is properly heat sinked, and can dissipate 15 watts of power, and that the maximum outside temperature of the scanner body itself will never exceed 40 degrees C (or so). Heck, in Spain during the summer, the outside ambient temperature alone is 40 degrees C!!

From the picture, I see that the DT40 has a heat sink on top, but I am not sure if this is really sufficient, especially if you do not have air flow over top of them. I am also not sure if the power dissipation capability of these scanners is as good as a real Cambridge 6800 or 6210 (my guess is that this is not quite as good). If my guess about the coil is correct, then I am also not sure if Jian has made component changes to account for this not-so-good-ness. If not, then his driver will assume that the scanner can dissipate 15-watts of power, and will not adequately protect the scanner in the face of rouge signals.

Jian and I have engaged in some private emails so some of this will come to light in the near future. Also, if Jian has not yet made the appropriate component changes to protect the scanners from rogue signals, he seems like the kind of guy who likes to learn and would surely make such a change quickly after it comes to light (again, if such a change is needed).

Bill

[Zoof]

If the scanners are heat-sinked properly AND if both the scanners and drivers are designed properly, then the driver will protect the scanner from damage. BUT, this makes a few assumptions...

It is absolutely possible to damage these moving magnet scanners if they are not properly heat-sinked AND if you input a signal that has a high duty-cycle. Examples of a high duty-cycle signal are raster images (very rare in this business, especially if you don't have an LD2000 system) and abstract imagery with a lot of squarewave content or high frequency continuous sine waves.

Hi Bill,

I gather from your info that mainly the heavy duty cycles need to be avoided and that it is unlikely that a single bad signal is going to destroy the scanners.... if setup correctly.
From what I know the feedback circuit is going to generate the larges currents while keeping the scanner on track. So an input signal generating large accelerations will do the most harm. E.g. when the scanner is moving in one direction at its highest speed then a large input signal applied in the other direction will be the most harmfull single signal that can be applied. When the scanner is lagging behind (because the scan speed is too fast) this effect could be the greatest. And this is the situation to check for with respect to single bad signals and when the AMP should protect the galvo. Is this about right?
With respect to heavy duty cycle I guess a software check could be performed to give some indication on scanner (ab)use.
It's not that I'm planning such wild things with the scanner but it would just be a shame to make these kinds of mistakes.

Matthijs

[yaddatrance]

Josh has been running the DT40's since earlier this year... From our experience, the heatsink is unneccessary
if the unit is mounted on a plate... and the scanners itself runs cool. The scanamp itself does generate a bit of heat...
If you find the scanners are running hot running the ILDA 12/24k pattern at 30k, you may want to retune your system,
the gain may be set too high.

The rule of thumb is... if it sounds terrible, it probably IS terrible.

It is most certainly possible to destroy scanheads with "bad" signals... but most likely not
in the case of your 40k square wave... most scanamps will pop a fuse fairly quickly
due to the current generated by the load before "real" damage occurs. The scanner destroyers
are usually frames/shows with a few "bad" points and a low enough duty cycle that the fuses
don't pop and the scanner is forced to regularly overwork itself.

If your projector hardware is to spec (i.e. doesn't overdrive the amp) and the amp isn't
near its max gain... even the worst bad scan won't kill it as long you catch it within a second
or two, so keep your finger near the "Oh &@#!" button when testing out new concepts...

For our developers, we gave each of them an analog scope to develop on... the 60-200Mhz
Kikusui's are nice surplus scopes with analog modulation (often called Z) and X-Y mode which
don't cost too much... This way they can choose to continue to view a bad frame to see where
they went wrong... I actually consider it an important step since not all errors are visible on the
computer onscreen simulation...