Morgan
0
- Joined
- Feb 5, 2009
- Messages
- 2,174
- Points
- 0
Hi All,
I did some testing and measuring the other day whilst putting together a 445nm build and found what I think are some interesting results. The build uses a Flexdrive and the idea was to simply max it out and see what happened. The host is a 18650 host with a custom heatsink, an A-130 diode with an Odicforce 3 Element glass lens.
Optical output was measured on my LaserBee I LPM, (the sensor was open to the air Jerry!
)
I did a few runs at different driver outputs and measured the current by proxy taking a mV reading across a 1 Ohm resistor in series with the diode. The diode was installed in a module clamp and large 10mm thick aluminimum plate for heatsinking.
The setup was run until stable and allowed to cool to ambient between tests.
The results are in the document below. [EDIT: Not just the graph but below that is an Excel spreadsheet](If anyone can tell me how to get them in an Excel format into the actual post them please drop me a PM, thanks)
During the last test I measured the output current from the driver at a couple of points and it seems this was dropping all the time as it heated. The temperature was measured from the surface of the inductor coil whilst using thermal plaster to sink the chips on the other side. I sinked the chips after reading posts that this was the best thing to do but maybe sinking the inductor could improve the situation?
What I find really interesting is that the diode temperature remained at a reasonably stable ~30 Deg/C. I measured temperature directly from the exposed back of the diode using an IR thermometer and touching it confirmed it was not even warm to the touch so although I'm surprised by this, I can't ignore the fact it was running very cool.
Another query is how an LPM behaves in a test like this? Does the fact the radiator does warm to around 27Deg/C affect the reading? I will ask Jerry to comment here as it's his baby but this will help me in the future and hopefully others.
During the test the optical output peaked at 953mW but when the resistor was removed and the build completed, the reading was hammering it's way past 1,000mW and my LPM overloaded. On a further test run, we appeared to hit the thermal protection at around 4 minutes as regular short flashes increased in frequency until I stopped the test after 5 minutes. This is not too surprising considering the increased output from the test and the enclosed environment the driver was now in. The test was getting close to maximum operating temperature anyway.
In conclusion, I think that results like the ones above and similar output drops seen in other testing, (such as DTRs 1.6W SS 18650 build), may largely be due to the driver rather than the more normal losses due to heat produced by the diode itself. Of course, the heatsinking in the host alone is not as good as during the test but I can see that taking the heat from the diode and it remaining at a stable temperature does not prevent power dropping when using our other common components. In a handheld build it is likely a combination of the two but there are now two seperate problems to address and balance. I'm now thinking a longer host with 2 Flexdrives and two 18650 may be in order to hit a more stable high output?
EDIT: Having now exceeded my measuring capabilities I sent this laser away for a reading and the graph came back today. Here is a 30 second run plot and I'm quite pleased -
Many thanks for reading.
M
I did some testing and measuring the other day whilst putting together a 445nm build and found what I think are some interesting results. The build uses a Flexdrive and the idea was to simply max it out and see what happened. The host is a 18650 host with a custom heatsink, an A-130 diode with an Odicforce 3 Element glass lens.
Optical output was measured on my LaserBee I LPM, (the sensor was open to the air Jerry!
)I did a few runs at different driver outputs and measured the current by proxy taking a mV reading across a 1 Ohm resistor in series with the diode. The diode was installed in a module clamp and large 10mm thick aluminimum plate for heatsinking.
The setup was run until stable and allowed to cool to ambient between tests.
The results are in the document below. [EDIT: Not just the graph but below that is an Excel spreadsheet](If anyone can tell me how to get them in an Excel format into the actual post them please drop me a PM, thanks)
During the last test I measured the output current from the driver at a couple of points and it seems this was dropping all the time as it heated. The temperature was measured from the surface of the inductor coil whilst using thermal plaster to sink the chips on the other side. I sinked the chips after reading posts that this was the best thing to do but maybe sinking the inductor could improve the situation?
What I find really interesting is that the diode temperature remained at a reasonably stable ~30 Deg/C. I measured temperature directly from the exposed back of the diode using an IR thermometer and touching it confirmed it was not even warm to the touch so although I'm surprised by this, I can't ignore the fact it was running very cool.
Another query is how an LPM behaves in a test like this? Does the fact the radiator does warm to around 27Deg/C affect the reading? I will ask Jerry to comment here as it's his baby but this will help me in the future and hopefully others.
During the test the optical output peaked at 953mW but when the resistor was removed and the build completed, the reading was hammering it's way past 1,000mW and my LPM overloaded. On a further test run, we appeared to hit the thermal protection at around 4 minutes as regular short flashes increased in frequency until I stopped the test after 5 minutes. This is not too surprising considering the increased output from the test and the enclosed environment the driver was now in. The test was getting close to maximum operating temperature anyway.
In conclusion, I think that results like the ones above and similar output drops seen in other testing, (such as DTRs 1.6W SS 18650 build), may largely be due to the driver rather than the more normal losses due to heat produced by the diode itself. Of course, the heatsinking in the host alone is not as good as during the test but I can see that taking the heat from the diode and it remaining at a stable temperature does not prevent power dropping when using our other common components. In a handheld build it is likely a combination of the two but there are now two seperate problems to address and balance. I'm now thinking a longer host with 2 Flexdrives and two 18650 may be in order to hit a more stable high output?
EDIT: Having now exceeded my measuring capabilities I sent this laser away for a reading and the graph came back today. Here is a 30 second run plot and I'm quite pleased -
Many thanks for reading.
M
Last edited:
