This is interesting...

[Seoul_lasers]

alright, I just got a new Ophir20C laser head that is new and calibrated. I also have a laserbee II 3.2W pro from J.Bauer that I recently got re-calibrated. I decided to measure a known laser of mine to see if there was any difference between peak and average output between the two sensors. What I found was quite remarkable.

This is a 400mW rated 405nm Olike laser. It has a single glass lens.

Ophir 20C = ~484mW pk (490mW pk cold). 480mW av after 1 minute sample

On the Laserbee, only after 2 minutes do the peaks start to level off and a more realistic peak and average form. The Peak and average after 2 minutes between the Ophir head and the Laserbee is = (1.055x)
Whereas I can obtain a result after 60seconds with the Ophir head.

I have to note that the sample taken is for 1 minute.

I have tried my Daguin 405nm 12x Kryton groove again the difference is
Ophir20C 703mW peak 689mW av
Laserbee II 665.8mW peak 660mW av


1W Olike 445nm

Ophir20C 921mW pk 890mW av
Laserbee II ~873mWpk 843mW av


.... etc ...etc





Comments?

 

[RyanElectro]

Yeah I always see rhd's builds with his ophir sensor to be higher than the laserbee during the "first 45 seconds" when the LaserBee is still getting is reading.

However, I'm surprised you only get an accurate result after 2 minutes. That's very long.

 

[Seoul_lasers]

Yup, well the reading is still off by a factor of 1.055 even after 2 minutes. What I mean to say is that the pattern starts to emerge where the peak and average have a distinct pattern. As you can see it takes nearly a minute for the laserbee to get warmed up.

Peak for the Ophir happens in the first 10seconds and then levels off whereas, the
response time is nearly 1 minute for the laserbee?

The RHD build was crazily low, peaking at only 2180mW, and on the Ophir peaked at +2600mW heh..

I decided to measure the current off the Laserbee sensor, and got 802mV peak (690-750mV) repeatedly from my Hioki DMM set on mV setting after ~5 seconds.
The reference voltage is supposed to be 3.3V? Would this not mean that the peak 802mV = 2646mW? Something seems REALLY screwy!

 

[rhd]

This likely isn't an indication of anything wrong with either of your sensors, but rather, an issue of which peaks the two sensors can identify, and where in your laser's duty cycle they peak.

If a laser peaks in the first X seconds, where X is the amount of time it takes the slower of the two LPMs to get a reading, then the slower LPM will miss the true peak.

An interesting comparison - for example with my build - would be to LPM it on the LaserBee first. Then, once the LaserBee's reading has stabilized (at 45 seconds, or however long it takes), move the beam to your Ophir. See what the Ophir measures.

If the readings are substantially different then, you've got a calibration problem with one or both.

 

[Seoul_lasers]

This likely isn't an indication of anything wrong with either of your sensors, but rather, an issue of which peaks the two sensors can identify, and where in your laser's duty cycle they peak.

If a laser peaks in the first X seconds, where X is the amount of time it takes the slower of the two LPMs to get a reading, then the slower LPM will miss the true peak.

An interesting comparison - for example with my build - would be to LPM it on the LaserBee first. Then, once the LaserBee's reading has stabilized (at 45 seconds, or however long it takes), move the beam to your Ophir. See what the Ophir measures.

If the readings are substantially different then, you've got a calibration problem with one or both.

Yes... did this and there is distinct difference. This is what I suspected was the issue, the sensor on the laserbee misses the peak, too slow... however the average is still way off as well this is my primary concern. The average is running 2050-2060... Ophir sees ~2300-2360mW averaging somewhere in the middle of those two values. I am going to trust the Ophir20C as it is totally new.

 

[rhd]

Yes... did this and there is distinct difference. This is what I suspected was the issue, the sensor on the laserbee misses the peak, too slow... however the average is still way off as well this is my primary concern. The average is running 2050-2060... Ophir sees ~2300-2360mW averaging somewhere in the middle of those two values. I am going to trust the Ophir20C as it is totally new.

Ahhhh, I see your mistake.

The LaserBee can't actually produce an "average", because the warm-up slope gets factored into the averaging.

 

[Seoul_lasers]

Ok...this makes sense then. This why the
Two graphs are miles apart

 

[RyanElectro]

Will keep this in mind when selling a build, on high powered lasers >2W, the difference can be as much as 300-600mW more. That's 15-30%.

 

[rhd]

Will keep this in mind when selling a build, on high powered lasers >2W, the difference can be as much as 300-600mW more. That's 15-30%.

Nope - that's not it either.

The "peak" should be the same (or relatively close) for any properly calibrated LPM.

The "avg" will not be accurate on ANY type of LPM that includes the initial "pre reading" slope in calculating "average power".

Just think about this logically. The slope at the beginning of a LaserBee 2.5W USB LPM's chart doesn't actually represent the power of the laser at that point in time. Your laser isn't "building up the courage to go full power", lol. The LPM just doesn't have a reading yet. So if you use that portion of the curve to calculate "average", how would you expect the figure to mean anything? The implication of that reasoning is that the same laser would have a higher "average" power on an LPM that gets a reading faster.

That's absurd

 

[RyanElectro]

Nope - that's not it either.

The "peak" should be the same (or relatively close) for any properly calibrated LPM.

The "avg" will not be accurate on ANY type of LPM that includes the initial "pre reading" slope in calculating "average power".

Just think about this logically. The slope at the beginning of a LaserBee 2.5W USB LPM's chart doesn't actually represent the power of the laser at that point in time. Your laser isn't "building up the courage to go full power", lol. The LPM just doesn't have a reading yet. So if you use that portion of the curve to calculate "average", how would you expect the figure to mean anything? The implication of that reasoning is that the same laser would have a higher "average" power on an LPM that gets a reading faster.

That's absurd

But nobody uses the curved part to determine reading. So what's happening is that the laser is actually putting out a much higher reading during the first 45s and then "settles down" to a max peak (and slowly but surely gets lower and lower from there)?

 

[lasersbee]

RHD is correct...

The Average reading on the EagleEye data logging software takes
all the data input to it to calculate an Average over the time of the
total test.

Once you start the data logging and the input goes above zero all
the above zero data is averaged out. Any data readings of Zero
are not factored in.

To get a more accurate Average reading we do the following.

1) take a reading to find the time it takes to get a max reading
of your Laser under test. Note that time.

2) after a cool down period of your Laser place the same Laser
onto the LaserBee Sensor for that same amount of noted time
and then start your EagleEye data logging software.

The averages will be much more exact since the ramp time was
not recorded and averaged in..

@ S_l as I've mentioned to you a few times already... reading
raw mV data off the LaserBee II Thermopile sensor will not give you
a linear output...

Where did you get your OHIR Sensor... there were issues with some
OPHIR sensors that were sold...



Jerry

 

[Seoul_lasers]

It would seem so. Problem is that the laser bee sensor just isn't sensitive enough.
I wonder what it would take to adapt an Ophir into my current laser bee?
Obviously I'd need a dual rail psu to power the sensor and a CPU upgrade?possibly...

 

[Benm]

I suppose its mostly a firmware issue: Sensors like the one on the laserbee have a pretty slow response time by themselves. Using methods like trend analysis its is possible to estimate on what power level the bare reading should settle.

Looking at the graph in the PDF, i think the prediction method is used, but it predicts quite a bit short of the final result. This could be due to a slight drop in laser output power during the first 10-20 seconds, though that's not really evident from the ophir reading.

If you do need a really accurate measurement just wait a few minutes for everything to settle. It would be nice if the power meters had an indicator for this steady state so you could be sure.

 

[lasersbee]

But nobody uses the curved part to determine reading. So what's happening is that the laser is actually putting out a much higher reading during the first 45s and then "settles down" to a max peak (and slowly but surely gets lower and lower from there)?

The EagleEye Data Logging Software does use the curved part to
calculate the average.
When I see a chart I use the level part of the graph to "eyball" the
average.

For Example on this sample Chart the Average is shown a 718mW
but if you look at the power line between 55 seconds and 205 seconds
the Average is a lot closer to 749mW than 718mW.

Why is the Average power Number so critical when it can be more
easily seen on the actual graph...

I would think that peaks and dips in power were more important..

Jerry

 

[Seoul_lasers]

Well, I tried Jerry's method. It works for the 405nm, but again, the problem is the 3-5 seconds of dead time for the warmup on those 445nm lasers. Still averaging +300mW too low.

 

[Trevor]

It's possible to write a software differentiator I suppose - it would get you closer, but isn't a perfect solution.

Trevor