Fb3...straight Beam?

[highvolt]

I guess they do this to keep the brightness of all the frames in your show equal. a 10 point image will have the same brightness as a 1000 point image?

hmm, I didn't hook up a scope (yet). I expected 100% duty cycle when using 500 points. Therefore, I thought the output voltage was not enough.

Anyway, the multimeter is also correct in some weird way I think. Since it gives the average voltage, it will correspond to the average brightness a user can see. The first dot (20 points at 1.7v) sure looked half has bright as the 500 dots.

While I like the FB3 a lot, I have the feeling that i'm losing a bit of laserbeam power (and power is money)

Anyway, I've sent Bill and Ante from pangolin an email. I'm very curious about this, hehe

[h2so4]

my analogue modules have not arrived yet, but I have the same problem with TTL ones. Exactly what you guys are talking about. I have not tried with a cope yet, nor with a multimeter, but the visual result is clear: if I draw one point(/circle/square/...) it is very dark, but if I select all, copy and paste lots of times, as data to the dac (point count) increases, image (and beams) gets a lot brighter...

@Highvolt: Interesting...keep us informed about Ante's answer, maybe we could learn the truth lying behind this!!! Thanks...

[fuzcub]

Perhaps the minimum 3% off time measured by dazebtwn is needed by the flashback hardware for housekeeping (shuffling data to/from the onchip ram , talking to the USB and DMX daughterboard chips, setting up the next frame to be scanned. etc)
B.

[ante]

Hi folks,

Highvolt asked me to comment on this topic, and I thought it might make sense to reply to this forum, rather than using private e-mail.

Fuzcub was very close to the complete answer, actually. Flashback indeed needs some time, between the frames, for reading and parsing the next frame's header, as well as certain housekeeping tasks. That means that there will always be some fixed gap between the frames, no matter how many points the frame has. However, this gap time is not all that big, as one could think, after readnig this topic. The gap time really depends heavily on several factors, which I will list later on, but let's just say, for now, it takes around 820 microseconds in direct feed mode, and around 1.25 millisecond in standalone mode. Let's do a simple math for 1000 points frame (an average real-life frame) at 30K scan rate:

duty = 1000 pts / 30000 pps = 33333 us
gap = 820 us
frame = duty + gap = 34153 us
loss = gap / frame = 820 us / 34153 us = 0.024 = 2.4%

So, the light lost due to this interframe gap is 2.4%, which is not all that bad, on my opinion. Sure, it would be better to have 100% yield than 97.6% yield which is what we have, but I think this is a small tradeoff for all the tasks we can do within that gap time.

In standalone mode, this gap is slightly longer (3.6%) due to more time spent on communication with DMX daughterboad.

Let's see how all this looks on the scope, and prove our math holds water:





Note that the color signal amplitude is 5V sharp, and it does not depend on FB3 power supply voltage. FB3 has it's own +-5.8V DC-DC converter for analog subsystem, which supplies output opamps (not just x and y singals, but also color signals). And, even if we would supply them with 5V, the rail-to-rail opamps, used on FB3, would be able to provide very close to 5V output. Note that the only part of the FB3 that really needs 5V is DMX daughterboard. So, if you use the FB3 without DMX daughterboard, you can actually power it as low as 3.6V, and it will still work, and it will still drive color signals to full 5V. There is absolutely no reason to power FB3 with 5.8V, and actually, this can even damage the FB3 (more specifically, the DMX daughterboard).

Highvolt, you also mentioned, in your mail, that if you short the serial termination resistors, you get something like 300mV more on output. This is not ok. There are two possible reasons for that. Either you have low impedance color inputs on your laser projector input, which causes voltage drop on 100 ohm resistors, or you have a common mode problems. Implementing a high impedance differential receivers for all color inputs (on your projector) would solve all any of these problems.

If you ask me, ANY device with ILDA input SHOULD have high impedance differential receivers on inputs. It's so easy to make it, and it solves so many problems.

Ok, let me now list all the factors that influence the absolute gap size:
- operation mode (direct feed mode vs. standalone mode)
- minimum points at output (cue properties -> vector settings)
- blanking shift (each shift adds a point at the frame start)
- existence of DMX daughterboard (DMX daughterboard makes the gap longer for couple of microseconds)

There are two more factors that do not influence absolute gap time, but they heavily influence a relative gap size (the procentual representation of gap within the frame):
- number of points in the frame
- scan rate

Note that there are more factors that might influence an overall frame brightness, but they have nothing to do with the interframe gap. One such factor are blanked lines within the frame. It is obvious that you cannot get the full laser power out of the projector, if you display the frame with many blank points. Another such factor is the quality of laser diode driver. If the diode needs a long time to reach its nominal output power, then you get less power with fragmented frame. Dazebtwn showed very nicely this problem on his last post. However, dazebtwn, I must comment on one thing that you mention in your post. You mentioned laser warm-up time. Warm-up time is not the time, laser needs to reach its full power after blank, but rather the time, it needs to stabilze the temperature after being turned off for a while. Smart laser diode drivers keep laser diode powered just below lasing treshold, which allows the diode to reach the full power very quickly. Anyway, if you do beam targeting, be sure to keep scanners show that beam as long as you can afford. In practice, that would mean, keep the frame rate as low as it gets (without experiencing flickering). You can do that by increasing the number of points in the beam, or by lowering the scan rate.

Gottaluvlasers mentioned that beam targeting with LiveQuick takes too much time. Let me check on this, and maybe we can find some simple solution, and implement it.

Well, maybe I forgot something, but I hope I answered to most of your questions. However, if you have any further questions regarding Flashback, or anything else, really, please do not hesitate to ask. I will try to follow this great forum and see if I can contribute in any way.

Ante
Pangolin Team

[gottaluvlasers]

just my opinion, but i think this post should be "stickied!!"

A LOT of wealthy information in here for MANY questions that have been asked in the past. with the scope screen shots, it REALLY helps to understand what is REALLY happening.

Ante and Dazebtwn

thank you VERY much for those excellent explanations and the O-scope screen shots. im sure it took some valued time out of your work day to set those little experiments up.

thanks!

-Marc

[allthatwhichis]

That's got to be the best first post I have ever seen! Welcome to the board ante!

[FourDee]

I know there is a way to adjust this, you can do this with the 'idle current' pot.
There is a topic about this on laserfreak.net (in german)
I'll look it up when I have some time

Edit(I've had some time):

See the image below
You can see the shunt resistor, usefull for monitoring the diode current. (0,33 ohms, so 1mV = 3mA)

Pot 1 : adjustment for the diode TEC regulation (don't touch if you don't know what you're doing)
Pot 2 : adjustment for the crystal TEC regulation (idem)
Pot 3 : Diode current
Pot 4 : Threshold current (this is the one you're looking for)

You can adjust the threshold current without influencing the max diode current. Turn on your laser without modulation. Adjust pot 4 untill it just faintly begins to emit light. Then dial it back untill the emission just stops. This will give you full 0-5V modulation range.

ps: I'm not responsable for possible damages caused to your laser by following this tutorial (just making sure ...)

greets
FourDee