FS: Kenometer USB Firmware Upgrade / 0.1mW Precision / $30

My thought was to go straight to the horse's mouth... after all you did
write the firmware so I thought you should know how it does what it
does. You are claiming 0.1mW precision...

Click to expand...

Why do I claim this? Well, I recently purchased a 543.5nm HeNe on ebay. The seller verfied its output on a Coherent LaserCheck as 2.6mW.

Using my firmware, my Kenometer USB reads it as 2.6mW.

I was trying to say the software "invented/manipulated" the original
data to something that looks odd on the graph in your fist post.

If you look at the original data RED line at ~00:00:02 it dips to
~80mW... The BLUE line at the same time slot goes to ~104mW...
That's one he!! of a difference... It happens again at 00:00:11 and
00:00:21...

Just looking at that chart... it looks like the firmware finds the noise
factor value... then seems to add that to the overall average of the
non modified actual data with noise...
That is what I meant by invented data... Somehow it just doesn't look
right to me...:undecided:
That's why the only way to see if the output is indeed valid is with a
known calibrated LPM as a base.

Click to expand...

Those are two entirely different trials. The differences in overall trend can be accounted for by remembering that the Novalasers X150 is a DPSS pen. Between the two trials, I did not wait long enough for it to completely cool. It cannot be expected to behave exactly the same.

That data was not "made up." Quod erat demonstrandum.

-Trevor

@Bluefan: Won't work? Oh? See results with HeNe earlier in my post. :tired:

Hm, tell you what. I'll go ahead and take a measurement of my HeNe, demonstrating precision to 0.1mW. I'll provide you with a graph of output. I'm also going to get a copy to someone to field test and verify my "claim."

My code is proprietary. It is closed-source. Deal with it and leave my thread alone.

-Trevor

You'll need to verify the linearity, a sweep of intensities and determine the noise and drift present. If that works out with good enough precision I agree that it works. Some magic closed source algoritm should raise concerns, as pointed out. Closing the discussing this way is your choice, I'm not selling so I don't have to answer questions, I only ask them.

Bluefan said:

You'll need to verify the linearity, a sweep of intensities and determine the noise and drift present. If that works out with good enough precision I agree that it works. Some magic closed source algoritm should raise concerns, as pointed out. Closing the discussing this way is your choice, I'm not selling so I don't have to answer questions, I only ask them.

Click to expand...

I'm going to commandeer a room for a few hours tonight to thermally stabilize the meter and take a measurement of my HeNe to prove the precision. At high powers this really only gives smoother output as opposed to on low-power systems that require more precision in measurement. 0.2mW is 10% of 2mW, but only 0.01% of 2W.

I don't feel compelled to reveal to great detail exactly how my code works. The way I see it, it will only force Radiant Electronics and Jerry to keep up.

magic closed source algoritm

Click to expand...

"Any sufficiently advanced technology is indistinguishable from magic." - A. C. Clarke

-Trevor

Will this new method cause an averaging over a longer period of time than the default firmware? Only reason I ask is because I want to see the transient spikes if there are any. I am sure there has to be some amount of averaging but I hope it isn't longer than the default. If I were to control the power source and cause it to go up and down by 5% of the power output so it was up one LCD update and down the next and repeat that for a minute would I see the power curve on the graph go up and down by 5%? If not then what is the maximum frequency that the power can change that the new firmware can accurately show?

twhite828 said:

Why do I claim this? Well, I recently purchased a 543.5nm HeNe on ebay. The seller verfied its output on a Coherent LaserCheck as 2.6mW.

Using my firmware, my Kenometer USB reads it as 2.6mW.



Those are two entirely different trials. The differences in overall trend can be accounted for by remembering that the Novalasers X150 is a DPSS pen. Between the two trials, I did not wait long enough for it to completely cool. It cannot be expected to behave exactly the same.

That data was not "made up." Quod erat demonstrandum.

-Trevor

@Bluefan: Won't work? Oh? See results with HeNe earlier in my post. :tired:

Click to expand...

I had 3 LaserChecks in the shop at one time and there were
variations of 5-10mW between them. Not even checking them
against our Newport... just between each other...

If the 2 different trials were separate and you used a non-stable DPSS
Laser then the 2 graph on your 1st post mean nothing. It is like comparing
Apples to Bananas...

No Laser behaves exactly the same on more that one test...
Not even the Labby Mounted Lasers we use to calibrate the
LaserBees... There is always a variation over time and temperature
of the Laser...

I don't want to know your source code... and I never asked for it.
You make a claim of 0.1mV precision with a 10bit ADC and I question
it... If you can't explain how you do that VooDoo that you do so well
then I'll have a hard time swallowing it...

Like I said you probably know more about this than I do....

BTW....was that...Deal with it and leave my thread alone.
directed at me..????


Jerry

xStatiCa said:

Will this new method cause an averaging over a longer period of time than the default firmware? Only reason I ask is because I want to see the transient spikes if there are any. I am sure there has to be some amount of averaging but I hope it isn't longer than the default. If I were to control the power source and cause it to go up and down by 5% of the power output so it was up one LCD update and down the next and repeat that for a minute would I see the power curve on the graph go up and down by 5%? If not then what is the maximum frequency that the power can change that the new firmware can accurately show?

Click to expand...

That can be tweaked based on your needs and how much electrical noise is in the system that needs to be attenuated. Since I don't expect a huge number of sales, I can custom tailor each one.

The ripple in my USB is huge - +/-5mW or more, as you can see on the graph. I've got the USB firmware tuned to output five samples per second (as it was before I went playing around with the firmware). If the Pros are far less noisy (which I suspect they are), I think a much higher sample speed can be achieved.

What's the current max sampling speed (it's configurable if I remember right?).

I had 3 LaserChecks in the shop at one time and there were
variations of 5-10mW between them. Not even checking them
against our Newport... just between each other...

Click to expand...

...maybe we need a new professional optical LPM standard then? Were these all new and recently calibrated? I assume they need calibration like other LPM varieties?

If the 2 different trials were separate and you used non-stable DPSS
Laser then the 2 graph on your 1st post mean nothing. It is like comparing
Apples to Bananas...

No Laser behaves exactly the same on more that one test...
Not even the Labby Mounted Lasers we use to calibrate the
LaserBees... There is alway a variation over time and temperature
of the Laser...

Click to expand...

I wasn't saying it would behave the same. I'm telling you that it was two separate trials and that it wasn't "made up" data. It's the same laser with and without electrical noise attenuation.

I don't want to know your source code... and I never asked for it.
You make a claim of 0.1mV precision with a 10bit ADC and I question
it... If you can't explain how you do that VooDoo that you do so well
then I can't swallow it...

Click to expand...

Tell you what. Let me get some stuff together and I'll hook up a AAA battery to the input and show you that I have much higher resolution than 5mV at >1V input (which will also prove higher resultion when Vref is 1.1V).

I'll even do it on video. Is this a fair way to prove this?

Like I said you probably know more about this than I do....

Click to expand...

It's amazing what you learn from tinkering with firmware. I recommend it.

BTW....was that...Deal with it and leave my thread alone.
directed at me..????

Click to expand...

It's directed at anyone who wants me to give out proprietary details about my code.

-Trevor

EDIT

I just did the AAA battery test. With 5V as reference voltage, I got these results:

http://www.safelasers.org/photos/KenometerUSB/precision.png

That's +/- 0.1mV precision. The battery is quite obviously outputting 1.6088V. Based on this test we can assume that at <1.1V the precision is even higher.

That's on a 10-bit ADC.

-Trevor

Like I said... I don't want to know your Source code... I don't need it
I use 16 bit ADCs in most of my products..

One of the Laserchecks was brand new that I bought directly from
Coherent at a ridiculous high price the 2nd was about 6 months old and the
3rd was about 1.5 years old. We also don't know the age of the LaserCheck
that was used to test your Laser...

If you want to do a test to show the precision... I would suggest
using a quality DMM and a variable (at least 0-5V) power supply and
your USB LPM..
Then we could see the relationship between the true 0-5V varying Voltage
displayed on a quality DMM and the output of the USB LPM on the Video..

Once without your Firmware and once with your Firmware...
That would make me a believer....


Jerry

Delivered. Here's the AAA battery on the old firmware. +/- 4.88mV (rounded here), as expected.

http://www.safelasers.org/photos/KenometerUSB/precision_old.png

I unfortunately don't have a variable 5V supply on hand to test with. Though testing other ranges seems superfluous, since Vref will always be 5V when above 1.1V. With 1.1V as Vref, the numbers only get better.

-Trevor

Sorry... but that Graph of one Voltage point (1 AAA Battery) says
nothing to me about the precision of your firmware...
If you are merely testing for noise or noise reduction a steady 1.5V
AAA battery will suffice.
But if you are testing Laser Power over the stated 0 to 5Watt range
of the Kenometer USB LPM then you need to show that claimed 0.1mV
precision over that range.
Without a Laser the only way to do it is with a 0-5Volt Variable
Supply...


EDIT
http://www.safelasers.org/photos/KenometerUSB/precision.png
If I take 10 samples....and divide them by 10....

1608
1612
1609
1605
1609
1608
1606
1610
1606
1608
------
16081/10 = 1608.1

That does not equate to a precision of 0.1 units per sample.
It is only the average of those values... IMO
How do we know that the actual value without noise was
not really 1605....??

Like I said in one of my 1st posts... all you seem to have done
is reduce the noise that the PRO and USN circuits produce on their
own... You have not made a more precise LPM...
And like I said in the same post.... I did that 6 months ago with
a resistor annd a cap strategically placed in our PRO...

I think I clearly stated my thinking on this... no sense to repeat
myself...

Good luck with your sales...

It is another good way to reduce the inherent noise in the Kenometer
USBs and PROs circuitry that use the OPHIR Heads...


Jerry

Couple things here.

• A AAA cell bypasses the circuitry that introduces noise. This is a precision test only, and it does demonstrate precision at that voltage level.
• There are two ranges, 0-1.1V and 1.1V to 5V. The ADC precision will be constant through those ranges. The 10-bit precision to 1.1V is 1.07mV, and after 5V the precision is 4.88mV.

That graph does, however, demonstrate beyond a shadow of a doubt that my firmware is far more precise. +/-4.88mV precision before, +/- 0.1mV after. Even if the linearity of the ADC changes by a relatively large amount (+2 LSB, yielding a variation in readings of +/-19.5mV), a precision of +/-0.3-0.4mV will be retained. That's a worst case scenario.

I just checked your site; the Laserbee LPM's (I/II) have precision to 1.0mV after 1W. You said you are using a 16-bit ADC? Or does "most" include none of your LPM's? :thinking:

-Trevor

twhite828 said:

I just checked your site; the Laserbee LPM's (I/II) have precision to 1.0mV after 1W. You said you are using a 16-bit ADC? Or does "most" include none of your LPM's?

-Trevor

Click to expand...

No need to be an a$$...

There is no reason for the LaserBee to display better than 1mV...
Even if in the firmware it is calculating at 16bit accuracy..

BTW... the noise associated with the Keno USB and PRO does not
come from the OPHIR head's circuitry....


Jerry
(8000-138-10386)

There is no reason for the LaserBee to display better than 1mV...
Even if in the firmware it is calculating at 16bit accuracy..

Click to expand...

I see. :thinking:

BTW... the noise associated with the Keno USB and PRO does not
come from the OPHIR head's circuitry....

Click to expand...

I know. It's the DC/DC converter. The simple test I did completely bypassed it.

-Trevor

EDIT:

http://laserpointerforums.com/f42/pic-microcontroller-16-bit-serial-voltmeter-49890.html#post687824

http://www.photonlexicon.com/forums/showthread.php/11105-Kenometer-Pro-Review?p=145636#post145636

Ah, so you switched for the Laserbee II. Makes sense. What made you decide to not use more of the precision? Were the values past the decimal changing too fast to be usable?

lasersbee said:

There is no reason for the LaserBee to display better than 1mV...
Even if in the firmware it is calculating at 16bit accuracy..

Click to expand...

Hey Jerry, where is your 16 bit ADC? Cause I don't see one on any of the laserbee series power meter circuit boards.

lasersbee said:

EDIT
http://www.safelasers.org/photos/KenometerUSB/precision.png
If I take 10 samples....and divide them by 10....

1608
1612
1609
1605
1609
1608
1606
1610
1606
1608
------
16081/10 = 1608.1

That does not equate to a precision of 0.1 units per sample.
It is only the average of those values... IMO
How do we know that the actual value without noise was
not really 1605....??

Like I said in one of my 1st posts... all you seem to have done
is reduce the noise that the PRO and USN circuits produce on their
own... You have not made a more precise LPM...
And like I said in the same post.... I did that 6 months ago with
a resistor annd a cap strategically placed in our PRO...

I think I clearly stated my thinking on this... no sense to repeat
myself...

Click to expand...

EDIT: Nevermind. This is taking far too long to explain. I'll let numbers and comparisons to calibrated LPM's speak for themselves.

-Trevor

MarioMaster said:

Hey Jerry, where is your 16 bit ADC? Cause I don't see one on any of the laserbee series power meter circuit boards.

Click to expand...

Guess old man time is catching up with you MM...
You will need glasses if you cant see it...:crackup::crackup:

BTW... how many of those circuit boards do you have in front
of you... I don't remember you buying any or me sending any
to you....:thinking:

Jerry