Pangolin ? laser not perform at maximum

[rfourt]

Guys, just to clarify: The problem is *NOT* with the FB3. It's with laser manufacturers that don't follow the ILDA spec for high impedance inputs on the modulation inputs for their laser driver modules.

You will experience this same problem on *numerous* controllers if you happen to have one of these drivers that has a low input impedance.

I sent a PM to Rfort because it appeared that this thread was devolving into a urinary olympics, and the real message (that you need to measure the voltage with an Oscilloscope) was getting lost in the clutter. As I suspected, he did not catch this information when it was posted earlier.

The bottom line is that people are *way* to quick to blame the controller, be it Pangolin or another brand. As I said above, this low impedance problem affects NUMEROUS drivers, including Laserwave. Well, that's not really fair, since LaserWave has since changed their design to solve the problem. But last year, it was a problem even for them.

More to the point, Pangolin performed an extensive research project on laser diode drivers back in the summer of 2008 (this is the project that Francesco Loon was working on for them while he was here in the States), and they found several other problems with the drivers they tested. Hayden's post above enumerates most of the common problems rather well, but not many people seemed to be listening to him.

As for a list of drivers, the problem is that they're inconsistent, even among drivers from the same manufacturer. So a list isn't very useful. If you have this problem, a quick solution is to use an op-amp as a unity-gain buffer on the color input. I've posted a schematic in the gallery that shows you how to hook this up. It's incredibly simple, and shouldn't take you longer than a minute or two to hook up. Here's the link: (The circuit uses a TL-082 op-amp.)

Adam

Once again thank you so much buffo.

Although about the list i think it can be a nice idea anyway. Yes the list might be litle inconsistent but it gives us a litle idea.


And to highlight again for future readers. The Error is NOT becouse of FB3 or the pangolin software. Im really sorry i doubted "pangolin Bill" about his answer, but thats becouse (as he said) i had not the propper understandings about the tech-stuff behind the error, and it was all jibberish for me, specially as English is not my first language.

I guess people like me with only litle knowledge about the true tech behind lasers etc, are excactly what i myself fear the most as a support technician in my own "Computer Bussiness". I always say, the most dangeruous/enoying customer is the one who think he knows 150%, who goes out and buy a hardrive and install it but have no clue about the settings behind it. or when the newspapers have there monthly headlines and gives away a free CD with 500 Memory boosting, fixing, tuning CPU throttling, maxmimizing, super tuneing -shitty-crappy free apps (spware/adware). That has the headline: "SPeed Up Your COmputer for Free" Then we know that:

- ok, the next 2 weeks will be full of support work. :P

Btw.. Did the New Matilda boards.. solve this also, as the same as the drawing from buffo?

[danielbriggs]

Btw.. Did the New Matilda boards.. solve this also, as the same as the drawing from buffo?

The MatILDA buffer boards are essentially like Buffo's diagram, but with a few more bells and whistles.
If you can wait a few weeks when I'm next at my other residence (where I've stored my 1W Viasho), I can test it for you then.
Alternatively if you want to give it a try, then 1 Power Board should suffice; x1 ILDA in, 1x ILDA buffered out.
MatILDA uses similar high impedance J-fet's; there is also ESD protection clamps on each output line to offer protection for >5V signals applied, preventing damage to the laser head.

Cheers,
Dan

[taggalucci]

The MatILDA buffer boards are essentially like Buffo's diagram, but with a few more bells and whistles.
If you can wait a few weeks when I'm next at my other residence (where I've stored my 1W Viasho), I can test it for you then.
Alternatively if you want to give it a try, then 1 Power Board should suffice; x1 ILDA in, 1x ILDA buffered out.
MatILDA uses similar high impedance J-fet's; there is also ESD protection clamps on each output line to offer protection for >5V signals applied, preventing damage to the laser head.

Cheers,
Dan

I'm so glad I bought some of these - especially when I get to hear of more great reasons to have them

I'd be interested in a method of A/B testing (i.e. before and after) to identify drivers which may be susceptible to these issues. Any suggestions?

[buffo]

Good point about the MatILDA buffer boards... Either one of those, or one of DZ's original color correction boards will accomplish the same thing as the simple circuit I posted. (Well, obviously a MatILDA board or DZ color board will do many other cool things too, but one thing they *both* do is buffer the input signal to overcome the limits of a low impedance driver on your laser.)

In order to test your driver, you must first have an oscilloscope. Then, with your projector disconnected, check the voltage on the color signals coming directly out of your controller. You should see +5 volts in a series of pulses as the frames are being drawn. If you don't see this (with just the oscilloscope connected), then you know you have a problem with your controller. (An unlikely scenario, but possible, so check this first.)

Now, assuming you do see 5 volt pulses, then you *know* your controller is working correctly. So the next step is to connect your projector and measure the voltage on the color signal again. If you see a DROP in the voltage with your projector connected, you *know* that the input impedance on your projector is low.

Now, how low is it? Well, it doesn't matter. You still have to correct it by buffering the signal. However, if you are determined to calculate it, then you need to know a little bit more about your controller. On the QM-2000, the output of the color signal has a 100 ohm series resistor, so based on the voltage drop across that resistor you can calculate the input impedance of your driver. But what if you have some other controller and don't know what the output resistor value is? Well, then it's a bit harder to calculate. (This is why I say it really doesn't matter... You still have to correct the problem, and the solution is the same no matter what the input impedance actually ends up being.)

The larger point here is that the actual value of the input impedance isn't all that important. What is important is that it be at least 2 orders of magnitude larger than the output resistance of your controller, so that you don't drop any significant voltage across that output resistance. ILDA picked 10K as a typical value for input impedance, with the understanding that it would be enough to ensure you see a full 5 volts on the input even with up to a few hundred ohms of resistance on the output of the controller.

Bottom line: If you suspect you are having this problem, do the test with an oscilloscope as described above. And if you have a full 5 volts with nothing hooked to your controller, and less than 5 volts with your projector hooked up, then you know you need to buffer your color signals to solve the problem. Either use the circuit I linked to, or one of DZ's color boards, or a MatILDA board, or some other op-amp based unity-gain buffer circuit.

Finally, as I stated in a previous post, credit must be given to DZ for the initial idea behind my simple buffer circuit. It is, in fact, based on a similar circuit that he uses in his color board... (And I suspect it is also very similar to what is used in the MatILDA board...)

Adam

[Doc]

I totally agree about all of the above and I have installed a buffer to alleviate that problem. However, I found that crowbarring of the mod signal is only a very small factor. Much more improvement can be had by temp tuning the lasers during modulation via the FB3, whilst displaying a typical frame rather than at CW (like most are set up at the factory).

You will lose some top end CW power but that is irrelevant for our purposes.

[lucerne]

You will lose some top end CW power but that is irrelevant for our purposes.

That's absolutely correct.

There are different ways to calibrate the temperatures of your laser heads. You can calibrate them for a highest possible output power during CW operation or for "best effort" during modulated operation.

Two weeks ago, we did this with a 300mW specified 457nm laserhead from CNI. We made two different calibrations:

1) First only CW operation, calibrated for maximum possible output power. We measured around 610mW, not bad for a 300mW laser head
Then we added a modulation signal (rectangular 0V/5V, 50% duty cycle, at the beginning only 1-5kHz). Ideally, you would now measure 305mW, this would be the RMS value of 50% modulation signal. Of course we didn't... We measured only around 110-120mW RMS.

2) That's why we recalibrated the driver during modulated operation. The first calibration was useless without an AOM. It took time and it really needs some experience to calibrate a driver during modulated operation.
Finally, we got around 440mW during CW operation and 210-220mW RMS with modulated operation (again: rectangular 0V/5V, 50% duty cycle, 1-5kHz). That's exaclty what we want!

Then I did some more measurements:
30% duty cycle: around 130mW
50% duty cycle: around 210-220mW
70% duty cycle: around 300-310mW

After that, I also made measurements up to 20kHz:
30% duty cycle: around 120mW
50% duty cycle: around 205mW
70% duty cycle: around 285mW

Conclusion: You loose some power, when you calibrate a driver in modulated operation. BUT you loose even more power, when you have a CW calibrated driver and use it in modulated operation! By the way: Don't forget, it was a 300mW specified laser head and we still get 440mW out of it

As soon there is enough time, I will also do some measurements with a fast photodiode.

Regards, Leander

[allthatwhichis]

Doesn't this all come down to controlling laser modulation is a bitch and you can't blame your laser controller hardware or software for any odd laser behavior?

You are sending electricity through a crystal to get light out of it. There is no way to do this "perfectly" everytime and the responsibility to get the closest thing to perfect modulation should be on the laser manufacture not controller manufacture.

[Doc]

Doesn't this all come down to controlling laser modulation is a bitch and you can't blame your laser controller hardware or software for any odd laser behavior?

You are sending electricity through a crystal to get light out of it. There is no way to do this "perfectly" everytime and the responsibility to get the closest thing to perfect modulation should be on the laser manufacture not controller manufacture.

^^^ exactamundo ^^^

[lucerne]

I really don't know exactly what you're trying to tell me.

[Doc]

I really don't know exactly what you're trying to tell me.

I think Allthat was refering to the thread starter, not the explanation.