chefla
0
- Joined
- Mar 24, 2011
- Messages
- 169
- Points
- 18
Hey guys,
I just finished my first 445 build. It is using one of daguins diodes, the direct press copper heat sink from Moh and a DX Aurora C6 host. Nothing special about those components. I have used them with an LPC-826 diode earlier this year. Check the link in my signature for details.
The one thing that is unique is the Micrel linear driver. I searched the forum and nobody seems to have used the MIC29152 so far. Well, maybe they have but did not care to share their findings? :thinking:
First up, why would I use a linear driver instead of a boost driver for a 445 build? Here are my reasons:
My goal for this build was to drive the 445 diode at 1100mA with a linear driver and two CR2 3.0V or CR123A batteries. During testing I used several sets of batteries to see how the diode current would change with the input voltage. I was hoping to find an IC which allowed me to have a 445 with full power regardless of the battery state.
For the comparison I had Micrels 2941, the 29152 and the LM1085. I do have a pack of various linear driver samples, but I had not found the time yet to test them all. So I picked the ones that looked most promising in their spec sheets.
I jotted the readings down on a piece of paper and did not do a proper chart or table. Sorry for that. But here are the most significant numbers which should give you an idea of the drivers performance:
(the numbers indicate diode current, diode voltage, battery voltage)
LM1085:
790mA@4.00V - Vin: 6.36V
844mA@4.19V - Vin: 6.55V
926mA@4.17V - Vin: 6.67V
1040mA@4.48V - Vin: 7.40V
MIC2941:
1084mA@4.36V - Vin: 6.56V
1094mA@4.40V - Vin: 8.02V
MIC29152:
1087mA@4.40V - Vin: 6.58V
1146mA@4.48V - Vin: 7.58V
Based on these numbers I picked the MIC29152 for my build. I used it together with 3x 3.3 Ohm resistors in parallel to get a 1.1 Ohm resistor which can handle the current without smoking. The resistors are rated at 0.6W and they need to dissipate 0.47W each, so I figured they should do fine without overheating.
I added two diy copper heatsink fins to the sides of the IC and put a thin layer of thermal compound on the back. The batteries are pressing the IC onto the aluminum piece that is normally containing the round LED driver PCBs. This allows for some additional heat sinking.
A big plus is the tab being connected to GND. In case you have enough space on your heatsink, then go for the TO-220 package and bolt the driver directly to your heat sink. By the way, there is no thermal epoxy used to ease the exchange of the IC in case it gets damaged. And, in case you want to build the same thing, make sure the bottom of the pill is flat. Mine had a slight mound, so I had to do some grinding in order to get it level.
I did a 3 min stress test to see if the IC would go into thermal protection mode, but there was no blinking or dimming effect of the laser noticeable. So I would say the driver is doing ok. But, the thing that worries me is the copper heatsink. It is working fine for my red builds. But for this monster the heat is building up rapidly. I think for my next 445 I will use a bigger host together with a beefier heatsink. I may get in touch with yobresal and buy one of his big hosts with the nice finned heatsinks…
Once I get my LPM, I will update this post with power numbers. But I would estimate this laser to be in the +1000mW range.
EDIT1:
I got my LaserBee today. I measured this laser at 1080mW with used batteries. I will charge the batteries overnight and update tomorrow the post with a second reading.
EDIT2:
With fully charged batteries I measured 1100mW. I like it!
Please let me know what you think. If you also have some experience with Micrel LDOs then please share it with the community.
Thanks for reading!
And here are the pictures:
That is how I started. Will everything fit?
Yes, it does! All wired up, heat sinked and ready to get installed.
Installed, showing the back side with the resistors.
Driver connected to the diode.
Fired up without a lens.
Beam shot.
The finished laser.
I just finished my first 445 build. It is using one of daguins diodes, the direct press copper heat sink from Moh and a DX Aurora C6 host. Nothing special about those components. I have used them with an LPC-826 diode earlier this year. Check the link in my signature for details.
The one thing that is unique is the Micrel linear driver. I searched the forum and nobody seems to have used the MIC29152 so far. Well, maybe they have but did not care to share their findings? :thinking:
First up, why would I use a linear driver instead of a boost driver for a 445 build? Here are my reasons:
a) I can build it myself (boost drivers require too many components and you need to make your own PCB, no thanks!)
b) I can use cheap batteries (the tailcap current is approximately the same as the current flowing through your laser diode)
c) I don’t have to worry about current ripples or spikes (one 10uF MLCC is enough to stabilize the driver)
d) It is cheap (I usually use free samples or buy the ICs for under $1 per piece)
e) It has a small footprint (ok, a boost driver can also be very small but has to be done as SMD)
The major downside of a linear driver:The heat dissipated by the driver. The voltage of the CR2 3.0V batteries drops under load to 6.1 Volts. Given the 1.1 Amps and the 4.5 Volt on the diode, the driver circuit dissipates around 1.8 Watt. So the efficiency of this setup comes to approx. 73%. Not really bad, but still a lot of wasted energy. A boost driver should easily exceed 80% in efficiency.
My goal for this build was to drive the 445 diode at 1100mA with a linear driver and two CR2 3.0V or CR123A batteries. During testing I used several sets of batteries to see how the diode current would change with the input voltage. I was hoping to find an IC which allowed me to have a 445 with full power regardless of the battery state.
For the comparison I had Micrels 2941, the 29152 and the LM1085. I do have a pack of various linear driver samples, but I had not found the time yet to test them all. So I picked the ones that looked most promising in their spec sheets.
I jotted the readings down on a piece of paper and did not do a proper chart or table. Sorry for that. But here are the most significant numbers which should give you an idea of the drivers performance:
(the numbers indicate diode current, diode voltage, battery voltage)
LM1085:
790mA@4.00V - Vin: 6.36V
844mA@4.19V - Vin: 6.55V
926mA@4.17V - Vin: 6.67V
1040mA@4.48V - Vin: 7.40V
MIC2941:
1084mA@4.36V - Vin: 6.56V
1094mA@4.40V - Vin: 8.02V
MIC29152:
1087mA@4.40V - Vin: 6.58V
1146mA@4.48V - Vin: 7.58V
Based on these numbers I picked the MIC29152 for my build. I used it together with 3x 3.3 Ohm resistors in parallel to get a 1.1 Ohm resistor which can handle the current without smoking. The resistors are rated at 0.6W and they need to dissipate 0.47W each, so I figured they should do fine without overheating.
I added two diy copper heatsink fins to the sides of the IC and put a thin layer of thermal compound on the back. The batteries are pressing the IC onto the aluminum piece that is normally containing the round LED driver PCBs. This allows for some additional heat sinking.
A big plus is the tab being connected to GND. In case you have enough space on your heatsink, then go for the TO-220 package and bolt the driver directly to your heat sink. By the way, there is no thermal epoxy used to ease the exchange of the IC in case it gets damaged. And, in case you want to build the same thing, make sure the bottom of the pill is flat. Mine had a slight mound, so I had to do some grinding in order to get it level.
I did a 3 min stress test to see if the IC would go into thermal protection mode, but there was no blinking or dimming effect of the laser noticeable. So I would say the driver is doing ok. But, the thing that worries me is the copper heatsink. It is working fine for my red builds. But for this monster the heat is building up rapidly. I think for my next 445 I will use a bigger host together with a beefier heatsink. I may get in touch with yobresal and buy one of his big hosts with the nice finned heatsinks…
Once I get my LPM, I will update this post with power numbers. But I would estimate this laser to be in the +1000mW range.
EDIT1:
I got my LaserBee today. I measured this laser at 1080mW with used batteries. I will charge the batteries overnight and update tomorrow the post with a second reading.
EDIT2:
With fully charged batteries I measured 1100mW. I like it!
Please let me know what you think. If you also have some experience with Micrel LDOs then please share it with the community.
Thanks for reading!
And here are the pictures:
That is how I started. Will everything fit?
Yes, it does! All wired up, heat sinked and ready to get installed.
Installed, showing the back side with the resistors.
Driver connected to the diode.
Fired up without a lens.
Beam shot.
The finished laser.
Last edited:
