MAGIC and IRON4D by rhd - Info and Guide

[rhd]

You CAN desolder individual ICs to lower the current, and no additional modification or bridging is required.

I would suggest against stacking additional ICs.

Stacking additional drivers should be fine.

 

[ApexProxy]

I was thinking IF one had only the option to stack ICs, that encasing the driver in Arctic Alumina could help counter the heat concentration. Why do you suggest against stacking ICs otherwise?

 

[rhd]

I was thinking IF one had only the option to stack ICs, that encasing the driver in Arctic Alumina could help counter the heat concentration. Why do you suggest against stacking ICs otherwise?

Encasing anything in Arctic Alumina doesn't help heat dissipation. That's not how Arctic Alumina (or Arctic Silver, etc) works. It's not a magic heat-seeking vampire that turns Florida into northern Canada.

Thermal epoxies/grease/paste/etc work by filling all the little tiny gaps between two heat conducting surfaces that you want to attach. In effect, it creates a greater overlap of the two surfaces than you would otherwise have if they were just in contact, but without the substance in between. However, thermal goops are not good heat conductors when used to fill large gaps. "Encasing the driver in Arctic Alumina" is likely to be a lot worse, and cause a lot more heat buildup, than simply leaving it open to the air. It would all depend on your host, but this strikes me as a loosing strategy 99% of the time.

In terms of stacking the ICs (as opposed to the drivers), my concern is with heat. You're now dissipating most of the top IC's heat through the bottom IC, meaning that the bottom IC has to do double-duty. Not only that, but it is doing double duty, without the benefit of it's own top-surface to dissipate heat through (either to the air, to the Aixiz module, or to the underside of another PCB, as the case may be). If the ICs get hot enough, they un-solder themselves.

More to the point, even if you're only partially concerned about the above, why take that approach when it confers absolutely no benefit? I guess you save 1.6mm of thickness? (the thickness of a PCB) That's a really silly reason to adopt a sloppy approach to building.

 

[IsaacT]

Since you seem to be knowledgable about Arctic Alumina I have a question for you: is it the same as two part epoxy? I need to affix a metal tail clicky to the switch beneath it and was told to fill the metal part up a bit with two part epoxy.

On topic though, I agree it sounds like a heat nightmare for the sensitive bits on the driver.

 

[arcaneinsomniac]

That's a great point you make, rhd! two IRON4Ds stacked is surely the answer.

wannaburnstuff: FWIW the Arctic Alumina *does* come in two parts. (two tubes of stuff you gotta mix)

 

[rhd]

Since you seem to be knowledgable about Arctic Alumina I have a question for you: is it the same as two part epoxy? I need to affix a metal tail clicky to the switch beneath it and was told to fill the metal part up a bit with two part epoxy.

Yes/No/Sort of/Sometimes

Arctic Alumina / Arctic Silver make single-tube thermal pastes that are nothing like two-part epoxy. If that's the stuff you have, then the answer is a clear no. It's not the same.

Arctic Alumina / Arctic Silver also make a two-part (two-tube) product. You mix the two parts, and in 5 to 10 minutes, it cures reasonably solid. In that sense, it is probably accurate to call it a two-part epoxy. It holds stuff together somewhat well, and it shares most of the physical properties with standard two-part epoxy, with the addition property of being a good heat conductor.

However, if someone tells you to use two-part epoxy, they're probably not envisioning that you would use Arctic Silver / Alumina. They're probably referring to typical (and cheap) "5-minute epoxy" (or something of similar name). Two-part epoxy is generally cheap. For a package containing 5x what you'd get in a pack of Arctic Alumina / Silver, you'd probably pay $1 to $2. You can find two-part epoxy at almost any dollar store.

Using Arctic Alumina / Arctic silver as a two-part epoxy to secure a tailcap, would be kind of like installing the plumbing for a new sink, and using pure silver as the solder.

 

[IsaacT]

Hahaha I will have to check my dollar general! Anything specific to look for?

 

[rhd]

Hahaha I will have to check my dollar general! Anything specific to look for?

sure is:

... They're probably referring to typical (and cheap) "5-minute epoxy" (or something of similar name) ...

 

[ApexProxy]

Encasing anything in Arctic Alumina doesn't help heat dissipation. That's not how Arctic Alumina (or Arctic Silver, etc) works. It's not a magic heat-seeking vampire that turns Florida into northern Canada.

Thermal epoxies/grease/paste/etc work by filling all the little tiny gaps between two heat conducting surfaces that you want to attach. In effect, it creates a greater overlap of the two surfaces than you would otherwise have if they were just in contact, but without the substance in between. However, thermal goops are not good heat conductors when used to fill large gaps. "Encasing the driver in Arctic Alumina" is likely to be a lot worse, and cause a lot more heat buildup, than simply leaving it open to the air. It would all depend on your host, but this strikes me as a loosing strategy 99% of the time.

I know Arctic Alumina isn't some magical heat sucking substance lol. My point was that in this use, since you are going to be soldering the driver directly to the pins of the diode module, using quite liberal amounts of the stuff will give the ICs the ability to pass heat into the module as well. Also, I would assume that for most cases these drivers will be stuck inside heatsinks along with the diode module there really isn't air around the driver even for micro amounts of convection. I was assuming alumina had better heat capacity and thermal conduction over air and if so, this would mean that bridging gaps with it would still be better than open air.

Basically, Arctic Alumina has a worse heat capacity and thermal conductivity than air?

In terms of stacking the ICs (as opposed to the drivers), my concern is with heat. You're now dissipating most of the top IC's heat through the bottom IC, meaning that the bottom IC has to do double-duty. Not only that, but it is doing double duty, without the benefit of it's own top-surface to dissipate heat through (either to the air, to the Aixiz module, or to the underside of another PCB, as the case may be). If the ICs get hot enough, they un-solder themselves.

More to the point, even if you're only partially concerned about the above, why take that approach when it confers absolutely no benefit? I guess you save 1.6mm of thickness? (the thickness of a PCB) That's a really silly reason to adopt a sloppy approach to building.

In my double IRON4D Mitsu 500 build I had to use tiny solder bridges to get the connection to my diode since the pins weren't long enough right from DTR. Stacking ICs would have saved enough space that would allow them to fit better on the pins. AFAIK, if you are building with one of these drivers I wouldn't imagine your goal would be infinite duty cycle since the entire point of these is to allow for extremely small builds right? This means that likely the heat pumped out by the diode is going to heat the main heatsink pretty quick as is which also means overall duty cycles will be short and the driver would have time to cool down, especially if thermal epoxy is used to bridge the ICs to the diode module or the heatsink itself. I know that I can't run my 1W Mitsu 500 in a WF-602C long since the heatsink heats up damn quick.