Page 2 of 2 FirstFirst 12
Results 11 to 18 of 18

Thread: Part 2 - Laser Power Meters

  1. #11
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Tempting.. Thanks for that reminder, I saw heads for those on eBay a few times, and will keep looking for them now. I considered them before, but had no source for calibrating. Do you know if they're putting out a buffered voltage, linearly proportional and adjusted from the head to match more than one meter? If so, I'm thinking that any accurate voltmeter will do. That would save SO much money...


    Marconi, those three Laserchecks... do you know if they deviated not only 10% from each other, but also, in one or more case, deviated more than 5% from accurate values found by better meters?

    Also, how much did you have to pay for your 3% one, and is there any chance I could get one for that cost? I've got money ready but if I make the wrong move it will be a while before I can afford to try again. I'm already having a hard time figuring out whether to get meter or diodes first.

  2. #12
    Join Date
    Feb 2005
    Location
    Florida
    Posts
    802

    Default

    Hey Doc,

    Yes, The Laserchecks differ from one another..
    I am really not sure why..
    Sorry not 3 percent accuracy..rated only 5.

    But my experience has shown much worse.!!!

    That was the test labratory enviroment conditions.
    The first and second units I bought In 2000
    +/- 5 percent.
    They read 12mw different from each other at around 200mW at 658.
    The third one I bought brand new Dec 04 from edmund qouted
    +/- 5 percent ..reads a bit higher than the other two by 15mw
    around 200mw at 658.
    This is comparing them to both Scientech 365 and 373.

    That is the reason I went with the Scientechs
    Calibration is easy and repeatable. measurements are fairly quick.
    (I can tune-up lasers with then)...It'l read down to a few tenths of a milliwatt.
    with up to hundreths resolution.
    Plus, now I can measure directly up to 30Watts.

    My first Lab meter is a Liconix 55PM that is good to 300mW but is
    wavelength sensitive...but very quick tuning.
    I had it Cal'd at SE Laboratories in July 2000 for 488, 532, and 658nm.

    All-in-All the Laserchecks are good , portable meters that give you an idea of the power involved...
    But, If you want a solid , Kewl looking unit that will last a while..
    Depending on what you are doing ...
    I'd get a Calorimeter or Pyroelectric power meter.

    Oh, the calorimeters are not buffered.. straight DC from the peltier..
    Altho, the newer ones maybe..dunno.
    You could make-up a meter, but why bother..they are so cheap on ebay..
    as well as the heads...just keep waiting, they show-up every now and then.

    "now , where did I hide the key to the Tardis"
    "My signature has been taken, so Insert another here"
    http://repairfaq.ece.drexel.edu/sam/laserfaq.htm
    *^_^* aka PhiloUHF

  3. #13
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    That, was hidden from you. :twisted: And from me, it would seem..

    Thanks for that info, I think I'll bypass the Lasercheck idea and go for a decent head when one shows up. So long as I can get it calibrated with some kind of certification, that would be better than a Lasercheck and far more useful. Did the calibration cost you a lot?

    Buffering won't be needed straight from the Peltier, it's low resistance and a stiff output already, but I haven't looked in detail to see how linear they are. I'm hoping I don't need awkward circuitry to match it to a linear input DVM. If I don't go for a Lasercheck, that takes the urgency off a buy I was thinking of. Which means maybe I should put diodes next. I doubt I have enough money to make it possible to join you in buying the Sony 250 mW diodes at any less than you already have to pay, but if you're willing to help me get some anyway, please let me know. Right now I have no idea of cost or where to find... All I do know is if I had to buy from any source I do know of for strong diodes, I'd have to sell a module for two thirds of what the diode would cost me, which is a non-starter. I know one good source for diodes but only up to 100 mW at 658. There was a 500 mW diode but horribly expensive and now also discontinued, so I don't think they'll be getting more like it unless they can sell lots.


    Edit:
    I just decided to gamble on this:
    http://cgi.ebay.co.uk/ws/eBayISAPI.d...tem=7597253735

    The seller says he'll take it back and refund all but postage if I can't use it. It's weirdly lacking in labelling, and he said he might have asked for more if he knew more about it. Apparently it works though, and looks like a Scientech 360 detector, according to a bit of Googling.

    Either I got something that is going to disappoint in some silly way, or I just got very lucky, I still don't know which. That's good money for the Reluctant Gambler, though.

    I guess the white sockets on it are for a calibrating heating element on the detector plate. I hope I can find out which head this is though, so I can find out the calibration constant. That way (I read from Sam's LaserFAQ) I might be able to calibrate by simply knowing current and voltage into that pair of sockets.

  4. #14
    Join Date
    Feb 2005
    Location
    Florida
    Posts
    802

    Default

    Ah, found the key...I'm back

    Well, the calibration for the Liconix I believe was around 100usd, but that was
    a while back, memory laspe thing.

    The nice thing about these heads is:
    It is so easy to calibrate ... all you need to do is feed the internal resistor
    enough voltage to dissipate 1w of energy...
    Ohms law comes into play here and the correction factor which is usually 1.08
    without looking thru my manual...maybe spec has one handy.
    I seem to remember that 6.1 volts at 152ma = .927 * 1.08 = 1.001W
    some may be a little different depending on the resistance of the resistor.
    But you get the jest.. Thats it...done .

    Find the heads with the internal cal resistor and you'll be set..Plus, You can check the calibration anytime you like.

    As for the Meter themselves..the 365 unit has a reset button to make-up changes in ambient temperature..one push and its back to zero.
    and 4 scaling ranges.. Spec and I have the manuals if you do pick one up.
    Oh, its microprocessor controlled too..It'l do differential measurements as well.
    But the bare bones meter is fine too..like a 373 etc.

    There are many manufacturers that do the same thing.
    Oh,., gotta go.. They're rerunning the new series again on Sci-Fi
    Dunno if I like it yet but, Hey.. I'm sure it'l grow on me.

    Nitey

    edit: I see we posted at the same time...Yes thats a 360001 head, Good for 30W Its what you want.. Your part way there
    "My signature has been taken, so Insert another here"
    http://repairfaq.ece.drexel.edu/sam/laserfaq.htm
    *^_^* aka PhiloUHF

  5. #15
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    I decided to abandon the idea of the differentiator. It makes sense in theory but is stupid in practise.

    I started by building the main circuit, a dual op-amp, (LF412) one half as follower to make a split rail ground from a single supply, so that the negative swings from the differentiator would work when light input was removed from the sensor, or a laser suddenly dropped power. The second half was a simple inverting gain stage to get 1 volt per watt for reading on any voltmeter.

    I had a second dual op-amp, one half intended for a non-inverting amplifier for the differentiator, which would then feed an inverting adder stage built round the second half.

    I used a voltage into the adder to provide a global offset adjustment. This works fine, until I set up the differentiator, which needs a gain of 40 to raise the peak differentiator value to match the main signal value for input to the adder. That makes the total offset error greater, so big that it's twice the value of the main signal from a 200 mW laser! (And impossible to accurately tune out without manually selecting components for every board built).

    Correcting it is unfeasible, it would be like spending X on the engine of a car, then having to spend 40X on the development of suspension that can absorb vibrations from a badly made engine.

    If it were possible to get a differentiator signal that was strong enough to match the main signal before any processing, this might have worked, but with a factor of 40 it's stupid to try, the correction for the differentiator would have to be an advanced (expensive) high precision op-amp with laser trimmed offset reduction and next to zero thermal drift. Pretty stupid to do this to 'improve' a laser meter that otherwise works to better that 1% accuracy with a single cheap dual op-amp and less than five standard 1% metal film resistors.

    EDIT:
    I just realised that the idea was unsound even in theory. :twisted:
    That's because the only way the raw differentiator signal could be strong enough to come close to matching the main signal well enough to allow similar accuracy in processing, would be if the rise time of the thermal sensor was so fast that there would be NO need to provide a differentiator to accelerate it.

    Consider this an excercise in footshooting. I present it here as a cautionary tale. If you want a sensor with a fast response, use silicon. Nuff said!

  6. #16
    Join Date
    Feb 2005
    Location
    Florida
    Posts
    802

    Default

    Hey Crow , no problem I understand completely...

    Actually, I've found the responce to be fast enough for just about all my measurements. Of course this is using the Scientech meter.

    Also, I hope the heater is intact so that calibrating would be easier.

    If you give up on that head..theres more to be found..

    I've been watching the thread on Alt.lasers as well., sounds like your on your way , despite the strangeness of that head.. Sam seems eager to help.

    Keep it up.. You will have an accurate measuring device when its all through.
    "My signature has been taken, so Insert another here"
    http://repairfaq.ece.drexel.edu/sam/laserfaq.htm
    *^_^* aka PhiloUHF

  7. #17
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    I like Sam. He helped me before, and was one of two people who successfully encouraged me to try usenet. I try to help him, at times, but he doesn't much from me, if anything.

    Looks like it being a different sensor than we thought, Sam confirms there's one that fits the description, (380101) and being a volume absorber type, it responds slower. I could still live with it, if it wasn't inaccurate across the surface. I did buy another cheaper one, still waiting for a friend to send it (Had to accept it for me from a seller who wouldn't ship abroad). That one really is a 360001, so I'll wait till that arrives before trying mods on this one.

    The basic circuit I had very quickly, but the gain on the differentiator bit caused a huge offset error, too big for my global offset idea. I'll build a differentiator seperately and test it to see if it behaves. If it can be done with no more than an LF411 or even half LF412 with external offset trim, this might still work, but it looks like the loss of accuracy might make it worthless even with dedicated offset trim with extra parts. Not too keen on lots of op-amp stages to do a simple job, either... I got it down to 4 including generating the split rail supply, adder, and differentiator. Without the differentiator I can use single rail and just one stage.

    So long as I get very little variance accross the detector, I'll get easily better than 1% accuracy. The circuit won't be the weak link. I'm not sure what the usual tolerance is for these heads though.

  8. #18
    Join Date
    Jan 2006
    Location
    Switzerland
    Posts
    129

    Default

    Quote Originally Posted by The_Doctor
    I like Sam. So long as I get very little variance accross the detector, I'll get easily better than 1% accuracy. The circuit won't be the weak link. I'm not sure what the usual tolerance is for these heads though.
    It is typically a few percent. If you calibrate your head with a heater, it can come down to 3%. At least this is what I found quoted in the data sheet of Scientech heads. I once calibrated a 360201 head in this way and found it to be accurate within specs.

    However, the heating method requires to know the appropriate correction factor, which is there because not all light will be converted into heat, and because the heating element won't heat the detector in the same way as the light does. This correction factor is typically a few percent and is usually written on the documents that come with the detector (and most often are lost in surplus deals).

    Below is an excerpt of the description of the cal procedure, out from a Scientech manual:

    Calibration of Large Aperture Calorimeters:

    A. Refer to Figure 1. Connect a DVM to the white jacks of the calorimeter.

    B. Measure the resistance of the substitution heater making sure to subtract the resistance of the patch
    cables from the total resistance measurement.
    Note: When measuring the substitution heater resistance of a 200 mm calorimeter, R1 and R2
    must be connected together in series.
    Compare this resistance to Rc in the calibration data in the front of the manual. The two should agree
    within 2%. If not, contact Scientech.
    C. Calculate the voltage equivalent to laser power using the following formula:
    V = (Rc x C x W)1/2
    where:
    V = voltage applied to the heater coil
    Rc = substitution heater resistance from step B
    C = Cal coefficient 360401 = 1.018 360801 = 1.000
    380401 = 0.974 380801 = 1.008
    380402 = 1.024 380802 = 1.008
    384UV5 = 1.021 388UV5 = 1.002
    W = desired laser power in watts
    D. Connect the DVM to the calorimeter’s DIN connector.
    E. Apply the calculated voltage (V) to the electrical substitution heater.
    F. Record the voltage reading of the DVM (Vc).
    G. Calculate the calorimeter’s output sensitivity (S) as follows:
    S = Vc/W
    where:
    S = calorimeter’s output sensitivity
    Vc = voltage output from the calorimeter in mV
    W = desired laser power output.
    The measured sensitivity should be ± 3 % of the calorimeters original sensitivity value.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •