The Micro Driver Module (MDM) concept (picture)

[Kryczeck]

I'll have to take a look at that rendering program, I've been using Keyshot mainly because it's fast. I've got a workstation though so hopefully renders will still run semi-fast!

I've got a few boost drivers on the way to measure, I was planning on having two plates that sandwitch over the top and bottom of the driver. These plates would be milled to match the driver so it would all fit together like a perfect puzzle piece. One thing I am worried about though is shorting something out so I was considering anodizing the diode heatsinks, possibly a hard anodizing to make sure it doesn't rub off and make a mess.

I'm still waiting for info on the V6 flex, I hope the dimensions are at least roughly the same size!

Hey, do you mind if I take a stab at a rendering using your design? Maybe a little mod here or there...but still your design ...you'll have full credit, not trying to step on toes...just want to see how it looks.

Thanks.

 

[Pontiacg5]

Sure, PM me an e-mail address and I'll send it to you. What file format works best for you?

 

[bbshamsa]

Hmmm. I like the idea but heatsinking isn't just about mass. I see you have some finned designs there that would increase surface area but without active cooling they won't have any air flow to effectively remove the radiated heat.

I'm glad someone mentioned this issue of fins on heat-sinks.

I've seen this error before, where a heat-sink is finned and then housed inside an external case!

That not only defeats the object of the fins, but compared to a non-finned HS with complete surface contact to the host, it's actually much worse.

You end up with heat pockets where conduction is impaired, as heat conducts much better through metal than air.

Also, with the fact that the air in those fins is effectively trapped, convection can't effectively take place either.

The Aixiz "Stage" Kit one looks much more effective as a heat-sink, and the ideal solution would be a host that included the heat-sink as an integral part of its design, in a solid single piece. That would eliminate the final impairment of conduction from one material of higher thermal conductivity, to a lower one; being the external host.

Basically, if the host was also the heat-sink, in a single, uninterrupted piece, that would be perfect.

I love the idea of a tailored host for these higher power 445's though.

Great images too, I had to double-take to make sure they weren't real

 

[Kryczeck]

Sure, PM me an e-mail address and I'll send it to you. What file format works best for you?

Doh! I never thought to actually ask you for the physical geometry/file... That would have been too easy :crackup:

I went ahead an threw this together...since I had the other objects and the materials already set, this actually was probably quicker.

Note to everyone: I take zero credit for this design, I had nothing to do with it...it's all based on Pontiac's vision and expertise.

 

[Kryczeck]

The Aixiz "Stage" Kit one looks much more effective as a heat-sink, and the ideal solution would be a host that included the heat-sink as an integral part of its design, in a solid single piece. That would eliminate the final impairment of conduction from one material of higher thermal conductivity, to a lower one; being the external host.

Totally agree...which is why I'll leave this up to those who actually know what they're doing. I'm just glad someone is already thinking about this, at it looks like Pontiac isn't too far off from having these.

 

[Morgan]

I'll have to take a look at that rendering program, I've been using Keyshot mainly because it's fast. I've got a workstation though so hopefully renders will still run semi-fast!

I've got a few boost drivers on the way to measure, I was planning on having two plates that sandwitch over the top and bottom of the driver. These plates would be milled to match the driver so it would all fit together like a perfect puzzle piece. One thing I am worried about though is shorting something out so I was considering anodizing the diode heatsinks, possibly a hard anodizing to make sure it doesn't rub off and make a mess.

I'm still waiting for info on the V6 flex, I hope the dimensions are at least roughly the same size!

Do you mean something like this but not fixed to the Aixiz module?

Dr.Lava suggested the chip with the minisink was the one to try and protect. It sits on the driver manifold, (just in the background), which is a semicylindrical bar that bolts to the wall of the heatsink. It's not as narrow as an Aixiz, as you can see the drivers sit across it in the assembled pic, but I think it's the same principle and seems to work well. I shouldn't think you will have to further insulate your two plates. I haven't had to as the resistors and any other, 'live', contacts are not exposed as the inductor and the pot both stand proud. I've only sinked one side and the minisink is all that is needed to give a flat bonding surface.

I still wonder how effective a heatsink will be if it's not bonded to the outer part of the module though. It will no doubt prolong the duty cycle but by how much? It may be better to dump the back part of the module entirely and get surface to surface contact between driver sink and host. It'd be much more efficient. [EDIT: Some pics were posted whilst I was writing. The ones just above are much better!]

M

 

[Kryczeck]

Do you mean something like this but not fixed to the Aixiz module?

...

Dr.Lava suggested the chip with the minisink was the one to try and protect.

Right...in fact it's that quote from Dr. Lava that actually got me thinking about something like this... The idea is that the part of the driver that needs heatsinking, would lie flush with the portion of the module's heatsink. I think Pontiac's design (see above) is much better for this, and he said he would make them specifically so that the driver would fit snug within the slotted portion. Having a module designed specifically with the driver in mind would eliminate everyone from having to re-invent the wheel when it comes to this. Your pics are great demonstration on how you could do it, but if these modules existed, you wouldn't have had to do it

 

[erdabyz]

Heatsinking an IC from the plastic back is not very efficient... Most swithcing IC's are packed in thermally enhaced leadless packages which get rid of most of its heat via a thermal pad in the back of the IC. You are supposed to make a lot of vias under that pad that connect it to a ground-heatsinking plane in the PCB. That's where most of the heat generated by the IC goes.

The drivers themselfs should have an uncovered copper area in the thermal plane to attach a heatsink if actually neccesary.

Know that we are achieving effciencies of arond 70-80% average with switching drivers and that means they don't actually dissipate too much heat. What feels hot for you is a comfortable temperature for the IC, and most of them have thermal protection. You can have your drivers working at 100ºC without any problem and the IC is fine. They've been designed for that.

 

[Morgan]

^This may be the case erdabyz but we are encountering thermal shutdown on a fairly regular basis driving these 445s at full bore. If the drivers were in free space it might not be such a problem but enclosing them, like we do, may be the reason temperatures are rising above the threshold. I went to the source for advice and although it's possibly not ideal, it was the advice from Dr.Lava that lead to the placement of the heatsink. I did a 3 minute run last night for the dual driver and although the power did drop, (to the eye at least), it didn't appear to go into thermal shutdown. Some minimal pulsing was seen, (not the complete shutoff usually observed when thermal shutdown happens so I suspect this might be the battery in this case as voltage and current were depleted after the test), so one or other of the drivers may have been stressed but from the PSU measurements I did, when one driver shutdown, the voltage to both was limited to ~2.8V. This may not be indicative of the same shutdown procedure due to heat as it was a lack of available current that showed this voltage drop on the bench but other tests should show up just what is going on in this particular build.

M

 

[HIMNL9]

I already do this with some different systems, and the system depend from the housing.

One is the same used from Morgan (when the space is a lot, using pieces of aluminium or copper, shaped for fill the more possible space from driver to host, then filling the rest with silicone thermal glue) ..... one, when the space is very tight, is filling the driver with thermal glue, do a wrap of kapton tape for insulate it (this ofcourse "squeeze out" the excess of glue from under the tape), then fill with other glue the remaining space from driver and host, this ensure insulation and still left good thermal conductivity ..... also, if the space is large and i can shape the inside, i try to build the driver board with the shape and measure that can grant the better thermal coupling ..... and where i cannot use glue, for the need to dismantle the assembly for prototypes or tests, i use silpads (are those soft silicone pads that you can find sometimes in old cd/dvd burners, where they couple thermally some chips with the bottom cover) ..... they are heat conductives, soft enough so they can be squeezed decently, and also insulating.

Also, considered the opportunity to place a PTC self-rearming "fuse" in serie to the LD, for thermal protection, but the commercial ones available actually are or too crap, or too big sized ..... still searching, anyway, cause this can made a great protection system, if i found some decent component

Never tried mineral oil bath (not yet :eg, but if you can obtain a sealed module container, it must be good also this (mineral oil is insulant, and is used for electrical transformers for cooling, too)

Also, not yet tried any handheld with a fan (made some draws, but not have access to the needed machines ) ..... also had some ideas and made draws for improve the existing ones (adding finned parts), but this need extra work from the builders, sometimes a lot of work, and sometimes re-anodization of some details, and is not easy .....

 

[Kryczeck]

Ok, some mentioned "fins" on the heatsink, however that would defeat the whole purpose of trying to do this within the current module size limits/requirements (which is why I used "inverted" fins in the OP).

However, just for kicks I went ahead and modeled these... Just looking at them cracks me up. :crackup:

Not sure how practical these would even be, or if they'd fit anywhere other than a labtop...which would then make this a labby...at which point why bother.

Anyways, I had fun coming up with the design so I thought I'd share it.

Note: This was just for kicks, so please go light on the bashing

 

[udanis]

Those are so cool!!! The only down fall i see for them is making them no matter what method you use to make them it will be a fairly complex multi tool operation..........but they are cool none the less :kewlpics:

I know I'd order a few if a GB ever came about I'm thinking they would be sweet for an actively cool high power 445nm

 

[Eudaimonium]

Gentlemen, behold, huge amounts of epicness is taking place right here.

Good job, gread idea!

Perhaps only the diode part of module is finned and it's being threaded into flat back end, then back end (and some part of front end) are pressed into heatsink, leaving fins out of the heatsink to dissipate heat, that would be epic.

 

[Kryczeck]

Perhaps only the diode part of module is finned and it's being threaded into flat back end, then back end (and some part of front end) are pressed into heatsink, leaving fins out of the heatsink to dissipate heat...

Hmmm, sounds interesting, but I'm having a hard time visualizing it... What would the "flat back end" look like exactly? I'm assuming that my finned module head (ie. "the diode part of the module") is already close to what you're talking about...it's just the back portion I'm not getting.

I'll be glad to throw something together to help with the visual aspect, but just not sure what it would look like yet.

Thanks for the feedback.

 

[Pontiacg5]

That would be near impossble to make unless someone can find extruded aluminum and copper in that shape. It would be possible with a 4th axis and a tiny, tiny endmill but it would take for ever and a day. The "pockets" of the fins would never be perfectly round either. Then you have the problem of lining the fins up when you thread the two pieces together...

But, it doesn't have to be pretty, you never see it. You could run a fan in a portable this way easily, some holes on the outside of the host for the intake then have the fan blow out the front. It would have the added benefit of forcing smoke and crap away from the lens too!

Time to get to work in solidworks! A host like that would be epic!

 

[Kryczeck]

That would be near impossble to make ...

I figured as much...but it was still fun to make "virtually"

Hey btw, did you get the pics I sent you?