Power Meter Calibration and Comparison

[jayrob]

I've had a thought.  Before shipping it out of country, we should have it shipped back to me and I can do a midway reading to verify the lasers outputs halfway through the experiment on my meter again.  I'll of course pay to have it shipped to me like everyone else. What does everyone else think?

I like this idea too, and can pretty much guarantee that the only ones that may not like it, are the poor couple of guys over seas that have been waiting the longest!

danq, you may have a point there with the temperature difference. I don't know about over 1 1/2 mW per degree?? Igor, what do you think?
Jay

 

[IgorT]

danq, you may have a point there with the temperature difference. I don't know about over 1 1/2 mW per degree?? Igor, what do you think?
Jay

Hmm.. It was warmer at Kenom's. Normally that would mean lower efficiency. But with PHRs that can mean a higher wavelength and lower losses in acrylics. I have seen PHRs climb in power as they warmed up. Scopeguy has a PHR from me, which does 155mW cold and 164mW at only 4°C more!

It is a possibility... PHRs are unpredictable diodes.. I used to think that their efficiency climbs with heat or that it's a mirror alignment thing, but it's possible that the wavelength climbs, causing reduced losses in optics. I have seen from 27% to 43% more power behind the lens on long vs. short wl PHRs... If a short wl. diode shoots up in wavelength as it warms up, the power after the lens could be increasing, while the raw output would actually drop with heat.



Kenom: As far as i know your 160mW laser measured 155mW on Electron's meter. So your meter is showing 3% more.. My meter shows 2% more than his, so our readings should be near identical - 1% appart..

Here is where it gets weird! Glenn's meter shows 7% more than mine! So he should have recorded more than you! I have calculated a correction factor for his meter once, but i don't remember it right now.

Looking at the data, Dave, Glenn Hydro and Jayrob measured pretty much the same. Some small variances are to be expected.


However Glenn's meter should be showing the highest numbers! Or did he use the correction factor i gave him? If not, then something strange is going on. It might be a good idea that the lasers go to you, Kenom..

It doesn't make any sense, that the laser would not let itself set higher than 75mW. That by itself worried me when i first heard it. But the consistency between all readings but yours are only making it more confusing! The lasers better go back to you, Kenom.... And even if you measure 75mW and it seems correct, we still need to figure out what went wrong that you couldn't set it higher - a thing like that could cause power to vary with circumstances.

 

[Kenom]

OK it goes to rckstr next right? It could be just as easy for him to open up the bluray and check out the circuit inside it. I'm no expert on electronic troubleshooting. Only basic stuff. If we really want to modify the laser with a new driver or fix the driver it will kind of nullify the readings thus far unless it comes back to me prior to the fixing so that the power can be tested first before it's fixed.

When it comes to scientific experiments accuracy is intregal. You have to make sure that it's done the same way each and every time and that your baseline doesn't vary or it nullifies the whole thing.

 

[danq]

After I do the basic testing, I'll crank up the heater and get it to 80 in here, and test again - that may tell us something.

DanQ

 

[Kenom]

Indeed it may. Will be interested to see the results...

 

[IgorT]

OK it goes to rckstr next right?  It could be just as easy for him to open up the bluray and check out the circuit inside it.  I'm no expert on electronic troubleshooting.  Only basic stuff.  If we really want to modify the laser with a new driver or fix the driver it will kind of nullify the readings thus far unless it comes back to me prior to the fixing so that the power can be tested first before it's fixed.  

When it comes to scientific experiments accuracy is intregal.  You have to make sure that it's done the same way each and every time and that your baseline doesn't vary or it nullifies the whole thing.

Indeed...

But if you send the laser to Rkcstr, and he measures the power (and the current), finds and repairs the problem, sets the laser to the same power he measured (or better yet to the same current), and then sends to you, we could still confirm if there was a drift in power or not, and have a fully functional laser from that point on.

It doesn't really matter tho, who gets it first, what matters is, that the power gets verified and the laser repaired.

 

[Kenom]

This is true. He'd have to make sure that the readings he takes prior to fixing it are pretty precise so that he can get it back to that point after the repairs are completed.

 

[Zom-B]

I think it's too dangerous for the accuracy of the measures if rckstr messes with the drivers. Even as an overseas person I think too that it is best to send it back to Kenom.

The difference in power vs 4 degrees can it be due to a kink? Last GB I receicevd 9 out of 10 diodes with a low kink (<120mW) of which one had a ridiculously low kink below 60mW. In a previous GB only 5 of 12 diodes had a low kink. Does anyone have a graph of this particular laser diode?

 

[IgorT]

I think it's too dangerous for the accuracy of the measures if rckstr messes with the drivers. Even as an overseas person I think too that it is best to send it back to Kenom.

The difference in power vs 4 degrees can it be due to a kink? Last GB I receicevd 9 out of 10 diodes with a low kink (<120mW) of which one had a ridiculously low kink below 60mW. In a previous GB only 5 of 12 diodes had a low kink. Does anyone have a graph of this particular laser diode?

It could be due to a kink. But the 75mW limit is indication of a problem for sure!

I said Rkcstr, because he might have an easier time finding the problem on the driver, since he made it.. If Kenom checks the power first it would be better, to make sure it didn't change. I agree.

 

[daguin]

Did I miss something?  Has the output started to decline?  Has it gone dark?  Are the measurements varying so wildly that we believe that something broke?

Even if there IS something "wrong" with the driver, as long as it is the same thing wrong with the driver for everyone, it should not matter (unless the output is varying because of it).  Remember, this is a test of our meters.  As long as the output is consistent, even if it is not what we wanted to begin with, the tests of the meters will still be valid.

If 75mW was unacceptable, the experiment should not have started.  However, just because the variables are not what was planned or predicted is NOT a reason to terminate an experiment.  We are not dealing with human trials here.

Unless I missed something, like I stated above, the most extreme reaction should be to have Kenom measure them again.  To open the case and do ANYTHING to the driver WILL invalidate the work.

It is possible that the outcome will simply show that the PHR at this current is too variable to get an acceptable guideline set.  However, if any variable or component is changed, the experiment must start over.

Peace,
dave

 

[Zom-B]

I think it's too dangerous for the accuracy of the measures if rckstr messes with the drivers. Even as an overseas person I think too that it is best to send it back to Kenom

...to test if the driver is stable or is too variable. Is what I meant.

 

[IgorT]

Even if there IS something "wrong" with the driver, as long as it is the same thing wrong with the driver for everyone, it should not matter (unless the output is varying because of it).  Remember, this is a test of our meters.  As long as the output is consistent, even if it is not what we wanted to begin with, the tests of the meters will still be valid.

You are missing something Dave.

If there is something wrong with the driver, so that it does not allow you to set it higher than 75mW for some reason, this does NOT mean it will be necessarily be the "same wrong" for everyone.

It could mean, that the driver is out of regulation for some reason. It could mean many different things, even that the current is gradually dropping. And all of them invalidate the experiment, unless they are eliminated.

That's why i "objected", when Kenom said it does not allow him to set higher than 75mW. Maybe i wasn't "loud" enough...
Replacing the pot should not have resulted in the 75mW limit. The limit is a sure sign of a problem of some kind. All the usual suspects are the kind that cause the current to vary.


It is important to verify:
- If the power is where it was when Kenom first measured it
- If the current is where it was set
- If the driver is REGULATING the current
- If it's not, why not
- If there is a problem, was it that kind of a problem that could lead to variation of the current or was it the "same wrong" every time (this can only be determined by finding the problem by an expert)

Also, if Kenom measures the power and it's 75mW, and Rkcstr measures the current and checks for regulation, that does not invalidate the experiment. Current can even be checked from the outside, simply by hooking the laser to a variable PSU through a DMM, and varying the input voltage, to see if the current stays constant. If the power starts changing, that means the driver was not regulating, and the current was dependant on different circumstances, even the diode temperature - Kenom's temperature was higher = more current.

If it is determined the driver is not regulating, then the experiment HAS to be restarted. If the current is not where it was, it has to be restarted regardless of the power.

If the diode is in a kink, the experiment would also be invalidated, as they go in and out of the kink with temperature sometimes.

Also, a driver can be replaced without invalidating the experiment, if it is done properly, and if the experiment is not already invalidated by the previous driver varying.


If there is a problem, it is better to find it NOW than when the lasers return from Europe.

 

[daguin]

You are missing something Dave.
If there is something wrong with the driver, so that it does not allow you to set it higher than 75mW for some reason, this does NOT mean it will be necessarily be the "same wrong" for everyone.
It could mean, that the driver is out of regulation for some reason. It could mean many different things, even that the current is gradually dropping. And all of them invalidate the experiment, unless they are eliminated.
That's why i "objected", when Kenom said it does not allow him to set higher than 75mW. Maybe i wasn't "loud" enough...
Replacing the pot should not have resulted in the 75mW limit. The limit is a sure sign of a problem of some kind. All the usual suspects are the kind that cause the current to vary.
It is important to verify:
- If the power is where it was when Kenom first measured it
- If the current is where it was set
- If the driver is REGULATING the current
- If it's not, why not
- If there is a problem, was it that kind of a problem that could lead to variation of the current or was it the "same wrong" every time (this can only be determined by finding the problem by an expert)
Also, if Kenom measures the power and it's 75mW, and Rkcstr measures the current and checks for regulation, that does not invalidate the experiment. Current can even be checked from the outside, simply by hooking the laser to a variable PSU through a DMM, and varying the input voltage, to see if the current stays constant. If the power starts changing, that means the driver was not regulating, and the current was dependant on different circumstances, even the diode temperature - Kenom's temperature was higher = more current.
If it is determined the driver is not regulating, then the experiment HAS to be restarted. If the current is not where it was, it has to be restarted regardless of the power.
If the diode is in a kink, the experiment would also be invalidated, as they go in and out of the kink with temperature sometimes.
Also, a driver can be replaced without invalidating the experiment, if it is done properly, and if the experiment is not already invalidated by the previous driver varying.
If there is a problem, it is better to find it NOW than when the lasers return from Europe.

Where is any indication that there is an inconsistency problem?  In my opinion, "maybe it might" is not enough.  Are we seeing any indication that the driver is out of regulation?  Are we seeing any indication that the laser is in a kink?  All I see we have now is that we do not know why the driver wouldn't give a higher current, NOT that the driver is problematic to the purpose of the experiment NOR that the diode is in a kink.  Beyond that, the experiment is showing some variability in the measurements of our meters.  Isn't that what the experiment was designed to test?

The pre-test and post-test will tell us if some extraneous variable was missed.  That is what they are for.  Since a question has come up, then an added mid-term test may tell us now if something has changed.  However, I still believe that DOING anything to the driver will invalidate the experiment and require a restart.  "Measuring" the current, is not "doing" something to the driver.  However, how does one "check for regulation" without altering something?  I disagree that we could replace the driver (or any part of the driver or re-adjust the driver) and still remain valid.  Any alteration of a significant component of the experiment will invalidate the results thus far.

Peace,
dave

 

[Zom-B]

There are just too many unknowns, too many things that 'just don't feel right'. The driver does things it theoretically shouldn't. And the measures deviate too much for comfort, even considering temperature. Did i miss something?

It's better to be safe than sorry. If we send it back to kenom, he can rule out these problems and we might have lost a couple of weeks. If the driver is faulty or the diode power in a kink, we loose everything if we don't send it to kenom.

 

[daguin]

Granted. Send them back to Kenom. IMNHO it is being done simply because confidence has been questioned, not because we see a problem. If Rkcstr measures the same current and Kenom measures the same output, will that be enough? I have a feeling that it will not. If confidence has been lost, even if they test the same, those that have lost confidence won't believe the results.

Peace,
dave

 

[IgorT]

It's an electronics problem Dave. You may not see it, but the 75mW limit is physically impossible with a fully working driver. And the driver is the only thing keeping the current constant, and the power somewhat constant.

And it is possible to check for regulation without taking the laser appart. I described how in my previous post - the current is measured and the input voltage varied. If the driver is in regulation, the current will stay the same. Since this is a linear driver, the current can be measured between the PSU and the laser.

And i only recommended a driver transplant IF there is a problem - as in the driver not being in regulation / a blown capacitor bypassing current / random unknown problem. If so, it invalidates the tests done so far anyway.


If we do the testing after Europe, we have to repeat everything. The problem can be tested for without invalidating the testing by:
- Measuring the power
- Changing the temperature of the laser, and remeasuring the power
- Observing the beam profile for mode-hopping
- Measuring the current, while varying the input voltage from a variable PSU
- I don't know how to check for a blown cap (which could bypass some current even if the driver was regulating), without opening the laser, however Kenom's first power measurement would show this.

All of this can be done by Kenom.


I'm hoping you have a variable PSU, Kenom?