Thermal impact of Aixiz modules

[arcaneinsomniac]

I couldn't find a subforum for this topic so that is why I am posting in General.

First off, I'm curious about what metal these Aixiz modules are actually made from. I'm worried about their ability to transfer heat from the LD to the rest of the (preferably aluminum or copper) host in an effective fashion.

My experience with electronics has been primarily with computers. So I have become quite anal over the years about that all-important thermal interface between a CPU and its heatsink. Everybody knows that you're asking for trouble if you do not use some form of thermal interface compound material to join the surfaces.

So here's the problem. A lot of laser builds have no straightforward soldering point in order to connect the case (battery minus terminal) to the V- in on the driver. In my experience the only clean solution (if using a case neutral diode) is to connect the case pin on the diode to the V- of the driver board. Because pins like the ones on the diodes are pretty much the only little protruding bits of metal that I can effectively solder!

This is great if the diode has good electrical contact with the Aixiz module (it's a press-fit. Check.), the Aixiz module has good electrical contact with the heatsink (this is the interface I have a problem with), and the heatsink has good electrical contact with a part of the host that is directly connected to the battery's negative terminal (I have never had a problem here. Check.).

Now I understand that a CPU thermal interface has to transfer 100W of heat effectively and we're talking about well under 20W for the beefiest 445nm diode we've got, but simply using some set screws to secure the module inside the heatsink is not likely to provide more than 5% of actual contact area assuming we don't have precision engineered parts. My suspicion is that even though our typical heatsink solution will work well enough to prevent our diodes from melting, it's going to cause the diode assembly to heat up a lot and stay really hot.

Let's see here: Google tells me that 12mm*pi*8mm = about 300mm^2 area, 5% of this is 15mm^2 contact area. Let's see. assuming the materials are both aluminum i've got one quantity left in the Q=kA/L equation where k = thermal conductivity of Al, A = 15mm^2 and I'm not sure what L has to be...

So I do not have the means to measure the temperature that the diode/Aixiz assembly heats up to in order to see if this is a problem. Ideally I want to measure the effectiveness of that naked metal thermal interface to see if a better thermal interface would be helpful.

When I use thermal compound in the aixiz-to-heatsink interface, thereby (hypothetically) dramatically increasing thermal conductivity at that juncture, it destroys the circuit and I must find a way to connect the driver V- input to the host in another way. This is often quite difficult. What I tend to do a lot is squish copper wire (soldered to the V-) into some thread or junction somewhere where manual tightening can be applied. It works, but is not such a "clean" solution!

The reason I am overanalyzing this issue is because I have a 500mw Mits 635 diode sitting here that I want to keep cool as much as I can, but I have no direct-press heatsink solutions available. I do on the other hand have an Aixiz module and an Aixiz module receptive heatsink that Mr. flaminpyro kindly built for me a while back.

The issue is that once I press the Open Can 500mw Mits 635 diode into an Aixiz, it's not ever coming out of that Aixiz. And I'm just trying to find an adequate thermal solution which does not involve my copper wire hack. Like, yeah, I'll put up with the copper wire hack if it means I get to keep the diode a bit cooler. But maybe there is an even smarter solution I haven't stumbled into yet!

Whew. Thanks for reading all the way to the end! :beer:

 

[Distillate]

First off, I'm curious about what metal these Aixiz modules are actually made from. I'm worried about their ability to transfer heat from the LD to the rest of the (preferably aluminum or copper) host in an effective fashion.

I believe I've read the aixiz modules are made of brass. I assume they are zinc plated. [EDIT: aixiz modules are nickel plated brass] There's always DTR's copper & aluminum modules..

I also think using the set-screw for heat transfer sounds iffy and plan on using the copper wrap method to improve heat transfer when I start builds. What if one of DTR's modules was wrapped with fine copper wire to make a tight fit and thermal compound was only used on the top third of the module, seems like the copper wire on the back 2/3 of the module would create your neg. connection.

Edit: I've also wondered about bimetallic corrosion when bonding aluminum to copper using a steel set-screw, etc.

 

[ARG]

I think the aixiz modules are zinc plated brass. Copper or Al modules are a much better option.

Edit: A second to slow

 

[Hiemal]

I have a really simple solution.

Make a square laser diode module, and then the contact area would rise up quite considerably. Easy peasy.

Or, forgo the module altogether and just go direct press, where the diode is pressed into the heatsink itself, rather than a module then heatsink.

 

[arcaneinsomniac]

the copper wrap method

It seems like carefully wrapping copper wire around a cylinder which must be inserted inside of another cylinder will result in nothing more than frustration for me. And how would you go about ensuring the tight fit?

I imagine the copper's softness would help produce a decent contact surface so at least there's that.

I do like adhering to a de-facto "standard" which is what the 12mm dia. modules have established. having interchangeable laser modules and heatsinks is a plus. I guess my squished V- copper wire solution is a pretty workable one for now.

I have a really simple solution.

Make a square laser diode module, and then the contact area would rise up quite considerably. Easy peasy.

Or, forgo the module altogether and just go direct press, where the diode is pressed into the heatsink itself, rather than a module then heatsink.

Square LD module. I like this idea! Too bad this means I'd have to have a mill in addition to a lathe. Gah!

Yeah the direct press heatsink (or direct press host!) is somewhat of a holy grail. It offers low modularity though. However since I have a fondness of solarforce flashlight hosts (what with the surefire host compatibility) a press-fit P60 dropin style heatsink is pretty good as far as modularity goes.

 

[arcaneinsomniac]

What are peoples' opinions about thermal-compound at the diode-pressfit junction? I like to do this too sometimes, and I have never tested if it makes for a bad electrical connection. In fact my last Ehgemus build I did this (it was diode pressfit into host, and I actually used arctic silver on the diode. As a result I had to do the copper wire trick. The diode mounting area had a nice ring mount thing I could squish that wire into, which is why it worked well just fine)

 

[ApexProxy]

Here is what I have found.

I never use aixiz brand modules because their main material is brass which has fail properties for what we need. Instead I use these:
https://sites.google.com/site/dtrlpf/home/diode-modules

When I press a diode into a module I have gotten into the habit of putting the smallest residue of thermal compound inside the pocket where the diode will be pressed into. This does not effect electrical conductivity here.

After pressing the diode and soldering everything up. I then smear a thin amount of thermal compound around the outside of the module and put it in the heatsink. The setscrew makes easily enough of a solid connection right through the compound into the module for any negative connection required. This only works if your heatsink makes electrical connection to the host. Also, the heatsink and setscrew cant be anodized for this to work.

I always solder a lead from the driver to the case pin to utilize the connection through the setscrew as long as I have a continual negative driver. Then, if you are using a host that uses a pill such as a C6 it is very easy to solder a bridge from the negative ring contact on the driver board to the pill itself and then the connection is also made through the pill to the negative case. Double connections means you will run into continuity issues half as often

Keep in mind this works for any diode that has an isolated negative pin and any driver can be used. In the case of an LPC-826 you will have to make sure your driver is continual negative.

 

[arcaneinsomniac]

I always solder a lead from the driver to the case pin on top of utilizing the connection through the setscrew as long as I have a continual negative driver.

I'm gonna nitpick: Isn't it true that if you use the case pin to the driver, then that *depends* on the setscrew's connection? Because the case pin (inside of the module) must connect to the back of the battery through that setscrew!

It's a good point though, the setscrew is a high pressure contact point. Should be good electrically so long as the screw is not anodized, and that ain't something a nail file can't fix.

Thanks for sharing your knowledge Apex

 

[ApexProxy]

Read the bit above that sentence.

Also, if you forgo the thermal compound around the module you can get a good connection from the module to the heatsink directly. I only put compound on the modules of 2W+ 445nm builds.

 

[arcaneinsomniac]

Read the bit above that sentence.

Wait so you said "solder a lead from the driver to the case pin on top of utilizing the connection through the setscrew"

If you don't solder the lead from driver (i assume Vin-) to the case pin, then the setscrew connects the 12mm module (and the diode case) to the battery-. How does that help connect driver board Vin- to batt-?

BTW I just ordered some copper DTR 12mm modules. My next 445 build is gonna be nasty.

 

[ApexProxy]

Ok, I realize I worded that wrongly. Edited.

DTR's modules are very well on spec by the way so they usually get very close to the perfect size to fit in the 12mm drilled hole and are even sometimes a little snug which is a very good thing. Lots of contact is definitely good.

I too started out on lasers basing my thermal knowledge on CPU heatsinks. For the most part lasers have drastically lower contact forces so thermal grease is much less effective as it would be on a CPU cooler but its never a bad idea.

 

[TheDukeAnumber1]

Not to break up the conversation but I'm pretty sure the aixiz modules are nickel coated brass.

You may continue :yh:

 

[Things]

Yes, they are nickel plated, not zinc.