Dave! Thanks mate! you've taught me more in 22 minutes than any book. I'm an absolute 'noob', when it comes to surface - mounted tech. seeing that I still 'piddle about' with valves and through - hole tech, the wish to take that step and move 'over' to surface mounted, was hindered by the thermal issue. watching this video has made it all crystal clear! happy days! on a side note:- over this week, I finished watching your psu project vids. The insight on how it's done by a professional - from start to (near) finish, isn't just enlightening but something us hacker-types don't get a chance to see often. put bluntly; "fookin' marvelous!" is the easiest way to express it. Dave, you're a legend mate! cheers for posting and taking the time and effort. new project idea for you: design/build a cranial plug (ala "The Matrix") so we could download all that knowledge. don't forget to add the "kung foo program" though ;)
jonskunator He did it in the time it took for me to go pick up Sagan from pre-school. With 17 minutes to spare in fact as he proudly proclaimed when I got back!
EEVblog Good job. Why not put his socal mediums into the description such as tweeker, fazedlook, instagran, linkedgrim, peenterest, meatup and of course µspace? I understand that they are very popular these days (I don't use them myself).
Nice to see the Usupply is still alive! Cheers for the videos Dave! Im a student hoping to study Electronics at university and all of this content excites and motivates me even more!
PC PSU manufacturers like Seasonic have taken to mounting their SMT 12V switchers on the bottom of the PCB, with a thermal pad between them and the case. Sometimes a heatsink is added on the top side of the PCB as well to use both the cooling from the case and from the airflow inside the PSU. Seems to work pretty well for them, even in 1k+ Watt units :)
Here's an example of such a Seasonic design: www.jonnyguru.com/modules.php?name=NDReviews&op=Story5&reid=385 They're using Fairchild FDMS015N04B MOSFETs, rated for 150C: www.fairchildsemi.com/datasheets/FD/FDMS015N04B.pdf It'd be interesting to see where the heat in such a design goes :) Seasonic was the first with their X-series of PSUs to put those switchers in such a location and I like to think that SS knows their business when it comes to designing a good/excellent SMPS.
EEVblog Yeah, but in this case, they aren't going to be dissipating 10's of watts, so they don't need it to be as good. Still, adding the extra thermal relief is a good idea, to help keep parts as cool as possible for a longer lifetime.
Maya Posch They also have heatsinks soldered through the board that connect to the FETs though the planes. These are cooled by the fan or convective airflow when in fanless mode. I bet that is where most of the heat goes. The case is also steel, so not very good at conducting heat.
megabytephreak Yes, I mentioned that in my second comment in this topic. Maybe Dave could figure out with his fancy Flir equipment where the heat goes in such a PSU MOSFET cooling arrangement? :)
Skip to 6:30 if you want to see the meat and potatoes (It's kinda a long intro of what Dave has already covered IIRC) - I do like those SMD heat sinks - I've never seen one before. thumbs up.
Hey Dave, great video. What I miss is a ballpark calculation of the actual temperature you'd have using this approach with the ucurrent. You show how to do it - and that's great, but I thought the actual numbers were missing.
It seems to me if you are using thermal vias leading to the case, it's best to do that in the middle of the board, so heat spreads out in two directions on the case. Putting it at one end distributes heat in one direction along the case (efficient) and the other direction into the air (good insulator).
Here's an idea for the case mount heatsink option with the to-220 as Dave started with: Include a hole in the board (or drill) to fit a machine screw through. You can then screw down your regulator via a tapped blind hole in the enclosure! Better heat transfer B-)
Really awesome series. Super helpful for me, but I would love to see it finished. It's been 4 years. Is Dave 2 still around? Thanks for what you've already done, Dave. Regardless of what you've done so far here, the thinking out loud, analysis, testing, mechanical concerns, etc., are so useful for everyone. That being said, we all hope you finish it up.
Dave, you know way more about design than I, but at 1:00 that TO-220 is going to need a screw to clamp it to the case. The leads can't provide enough force to provide good contact. At 1:20 that heat sink should be isolated to avoid short circuits.
Great video. I was hoping for a demo: how cool would it have been if Dave had done the calculations for the design, built it, run a few watts of power through it and then measured the temperature to see if it it stabilised to something close to the calculated value? That would would be great (particularly if the result actually matched the prediction !!) And congrats to Dave #2 on the animation...very nice.
You probably dont care but does anyone know a trick to log back into an Instagram account..? I stupidly lost the login password. I would appreciate any help you can offer me.
@Mathias Steven i really appreciate your reply. I found the site on google and im trying it out now. Seems to take a while so I will get back to you later with my results.
I use this exact technique with Class D amplifiers. But I fill in the vias, and make sure the heat-sinking surface is flat. I dont use silpads, I use kapton with thermal compound. It works ok, even up to a few hundred watts, but since Class D is efficient, not much is being dumped out as heat anyway. I do as much to the PCB as I can, and avoid machining the "valleys" into aluminum for other methods which requires extra tooling/machining costs.
yesterday i teared down optical drives donated by my school, and i've seen this technique on the h-bridge driver for the CD motor. What they did was the've made a cutout on the board under the thermal pad of the component, put the silpad and then on the metal casing they've pushed the metal inwards where the cutout in the pcb is, so it will make a thermal contact and disapate the heat there
Good stuff, Interesting how some subjects in physics are this similar (Electrical and Thermal Resistance). Love to see a video on the uSupply new electrical design, Like you used to do with the original uSupply videos.
The 3D drawing shows the bar screwed from the case from the outside and from the pcb side also. This would ensure good thermal contact around the screw holes. Everywhere else on the bar, however may not be touching for thermal contact. You'd have to be pretty close if not just a shade too thick for the metal bar height or it could crack the PCB screwing down to it. Maybe you could do the same thing with some metal cylinder standoffs threaded from both sides instead of the bar.
Dave I love your channel but rarely login to comment. Your videos are very informative. Would it be possible to get more teardowns? They are educational beyond a specific topic, and very entertaining. Cheers.
David 2, µDave, D2, AfroDave, Mini Dave and Dave 2. These are some of the names given to your asistant. Honnestly the simplest is the best: we all know you as Dave and his name is David.
My experiences with a typical design are these: - Spread the force evenly, don't put bolts only in the corners of the PCB but also put some in between and around the components. You can use smaller ones and use less force, it won't deform the PCB material as much as 2 big ones in a corner as Dave did in this example. I made this mistake and had the PCB arc up in the middle. It even created an visual airgap between the bottom of the PCB and the aluminium bar it should've rest upon. Nobody wants that :( - Adding a matrix of thermal via's underneath the large solderpad did wonders for transporting most of the heat... Only a small part was dissipated by plenary conduction into the via's surrounding the components. So plug and cap your via's if possible as that stops the Tin whicking effect so it won't hurt the soldering process. - Adding thermal paste between the aluminium bar and the housing didn't help me much, there was so much surface area between the two that it didn't make much of a difference...
I was wondering when instead of using vias, you used jumper links (soldered both sides, sanded flat on the other side), if that would give any significant improvement? I know some people that do it for currents and claim it works great, and since current and temp flow is always similar...
In my home-etched boards, I do this often. I will also use like a thru-hole resistor lead and solder it on both sides, since the resistor will be there anyway.
Hi Dave, I like your fundamentals videos when you are in front of a breadboard, 10^6 times thumbs up! One question, in which case do I want to isolate the ground/tab of a regulator from the ground plane of the circuit? Could you make a short video about it? Actually I have another video request, if possible? How can one succeed to lay out a single sided board avoiding the jumper wires? It challenges me as hell as a beginner...
Hi Dave, excellent video as usual. One question, what would you do with the vias if they increase the stray inductance and cause voltage overshoot in a switching circuits?
One disturbing trend is that some manufacturers are using thick (~5mm) soft pads to transfer heat from large SMD ICs to cases. The problem is that over time they lose their ability to spring back and they leak oil all over the PCB.
You gas fill guys - Argon. Only way to go. Safe as well. Used it on a major heat pipe for a cyclotron gizmo at UCB (the hill guys from Livermore). Fun.
xDR1TeK I'm generally pretty hopeless at remembering theory, especially if I haven't touched it since uni, but thermal stuff like this is pretty basic and something I use reasonably often.
xDR1TeK I don't quite know how to respond to this without sounding arrogant, but once you've seen this stuff once, and properly understood it, it's going to stick. Visual memories of various component placed in different ways, you can imagine and play around with these various arrangements in your head. Remembering that things get hot isn't all that challenging. And all of the thermal resistance stuff? Looks incredibly obtuse and complicated, but may I draw your attention to the units of thermal resistance: degrees C per watt. For every watt you dissipate, this is how many degrees C difference you get. That's all you need to know, all the rest just falls out from common sense. I'm not saying that being guided through all of this for the first time by an experienced expert isn't valuable, quite to the contrary, but it's not like it's made up, illogical, or actually even remotely obtuse in any way. Do chefs remember everything from their chef school days? It's just a weird question to ask, really.
TheHue's SciTech I think it's an interesting question and touches on a lot of issues. If it's something I haven't touched since uni, have never really used it in practice, and didn't particularly like it much at the time, then it's unlikely I would remember it. Like for example if you asked me to solve an integral or something I probably couldn't do it, I'd have to go look it up. But once I look up the stuff it usually back to me pretty quick if not instantly. The good thing about thermal design is it's an easy concept that's hard to forget once you know it.
EEVblog Agreed on all points. I guess I saw the question (probably incorrectly) as coming from a "how do you memorize so many completely disconnected facts" kind of direction; I just wanted to point out that the same question could be asked of just about any profession or career, and that far from being obtuse and disconnected, all these facts slot together in a lovely and rational way. To an outsider who doesn't really follow, it looks far harder than it really is. Having said all that, my engineering ability drops 10x without Google by my side.
College/university education is as much about learning how to learn as it is about the specific topics you study. Even if you don't remember a specific topic, a good engineer can teach themselves what they need to know without having to have it memorized. At least that's my experience.
***** Argh, yes, I was going to go back and shoot anther clip mentioning that, but forgot in haste to get the video out (locked out of the lab for the entire weekend due to building work)
EEVblog and to answer your question, yes quite significantly, but you can get issues with lumpiness in the contact surface which may not be good. You can also get the vias filled with solid copper, but that's a bit of an exotic technique.
Huh. This is the first time I found out it's spelled "vias". I never heard the term before watching your videos, and just assumed it was spelled "veers".
What if I want to get rid of 80W of heat from a SMD power transistor? Or do I have to use TO-247 or TO-3P packages? My USB-operated dummy load dissipates some 80W power on IRLP3034N transistor but that is a through hole TO-247 part. Can I go to SMD parts on that?
EEVblog So better stick with TO-247/TO-3P for that level of power dissipation? I am trying to design a dummy load but with a design constraint that the entire dummy load have to be powered from USB.=, and that means smaller sense resistors and a hell lot dissipation on the MOSFETs.
I guess you could use insulating heatsink plaster instead of a Pad and a screw for a case dissipation method... much cheaper and elegant than a pad and a screw
If you didn't feel buying a licence for Solidworks there is a free parametric modelling open source equivilant called FreeCAD. I haven't used it all that much, and I'm sure that it isn't as feature complete as Solidworks, with the animations and physics modelling, but it is free!
could i use paper or cling wrap as a substitute for a silpad, or would they thermally insulate the part from the heatsink? i'm not exactly made of money.
EEVblog actually, i think a layer of paint or lacquer could be used. if you make the insulating layer thin enough the thermal resistance goes down drastically.
Why don't you let the transistor stick out just enough over the edge of the PCB (which can be shortened accordingly) so that its metal part can touch a metal bar - coming from below - that is level with the PCB? Would make things much easier imho.
If plastic is so bad at conducting, how come some TO-220 packages of high power MOSFETs I order come 100% plastic (no metal back or tab)? Eg. NDF08N50ZG (500V 7.5A)
+Ryan O'Connor thats because they have low power disspation rating, and they use a thin plasic and because it right on the stuff that heats up its not that bad. but if you are trying to say that plasic is a good conductor, I have nothing to do here.
Either I missed it, or there was no mention of the following in this video. While much if not most of the heat to be dissipated is conducted away from the source through the copper islands and through the vias, to the thermal connector bar and finally, to the metal case and into the air external to the case, a portion of that heat does not take this course, but is instead dissipated into the air trapped inside the case via the surfaces of the copper islands and the "vias" that provide a thermal connection(s) between those islands. And while the total amount of this interior-dissipated heat is probably minimal and immaterial relative to the total heat to be dissipated, and while this is probably not an issue for many if not the majority of applications, in theory there may be designs and applications that are in fact significantly impacted by this interior-dissipated heat. That said, if the goal of this video was to undertake a proper examination of and to discuss and to characterize a thermal system, then using the argument above, it has fallen short of that goal in its failure to address an important aspect of that thermal system. Sure, I know. This is not intended to be a comprehensive scientific examination of thermal systems which could span an entire library of books and several scientific disciplines. But by my intuition, it would seem that this could prove to be a common shortfall of such a design and should therefore at least have been mentioned briefly in this video. I should add that I don't have a degree or any professional experience in EE or thermal design. I am not an engineer (any kind) and the closest thing I get to being officially termed a scientist is by merit of my BS in Accounting. So, I am speaking mostly from the perspective of a hobbyist and a layman here. Bearing that in mind, I hope you will take this comment for what it is worth. I have chosen to make a point of this only for the potential benefit of those concerned and not to discredit the author, who is doing an amazing job and providing a wonderful service to the public with the EEVBlog.
aerofart I thought I mentioned that in a text overlay somewhere? It's not just the copper, *every item* in the thermal resistance chain (bar the silpad) has surface area that will radiate heat into the surrounds. I mentioned it was a simplified but practical model, and there are more complicated ways to analyse it all if you want to.
aerofart He mentioned every component radiating heat to its surroundings; you could also talk about natural convection ocurring inside and outside the case, but modelling that gets pretty meaningless pretty fast and you would just resort to trying it out. What I believe you are missing from your argument is that you have a heat source and a heat sink in the system, the device pn junction and the case surroundings, and with this model you are giving a ballpark estimation of what temperature to expect on the junction relative to the ambient one. The air inside the case could be represented as a thermal resistance in parallel from the package to the case with a much bigger value, and its temperature you wouldn't care about as it would just follow what the heat source provides it. These considerations may be relevant when you have a heavily temperature dependent device performance and a relevant heat source sharing the same enclosure, but that was certainly out of the scope of this explanation.
I can't believe they've run the dataline right throught the power supply. Amateur hour!! NB. I don't want tot spread misinformation. So just in case you didn't get the reference: *it was a joke*.
EliteReacon Dave has the $50- and $100-multimeter "shootout" videos that you might enjoy. They are a few years old but I don't think much has changed. Sure there may be some newer models not discussed, but, except at the high end, the trend is not to improve quality, but to decrease manufacturing costs. Some important factors in a multimeter are safety, accuracy, update rate and quality of leads. As I recall, Dave seemed to place too much emphasis for my liking on checking ruggedness by twisting the meters, but to each his own.
I have a UNI-T ut61c, it is more than excellent. Precise with alot of cool functions, and is safe as long as you dont work with anything high energy. Messing with household mains electricity wont be a problem, but isnt recommend. But dont buy the ut61c if you need it higher resolution and true rms. Buy the UT61E. Which is true rms. This meter never failed me. I love it
EliteReacon I live in uae. Here if you are lucky to find flukes. They are 2x the price. Don't even think about any other brand. But UNI-T based itself pretty good here. Its everywhere. That is why my experience is limited to it.
This is a goldmine of heatsinking info. Thank you so much, this has helped a ton
Dave! Thanks mate! you've taught me more in 22 minutes than any book. I'm an absolute 'noob', when it comes to surface - mounted tech. seeing that I still 'piddle about' with valves and through - hole tech, the wish to take that step and move 'over' to surface mounted, was hindered by the thermal issue. watching this video has made it all crystal clear! happy days!
on a side note:- over this week, I finished watching your psu project vids. The insight on how it's done by a professional - from start to (near) finish, isn't just enlightening but something us hacker-types don't get a chance to see often.
put bluntly; "fookin' marvelous!" is the easiest way to express it.
Dave, you're a legend mate! cheers for posting and taking the time and effort.
new project idea for you: design/build a cranial plug (ala "The Matrix") so we could download all that knowledge.
don't forget to add the "kung foo program" though ;)
The animation by µDave looks quite professional.
jonskunator He did it in the time it took for me to go pick up Sagan from pre-school. With 17 minutes to spare in fact as he proudly proclaimed when I got back!
jonskunator µDave ftw!
EEVblog Good job. Why not put his socal mediums into the description such as tweeker, fazedlook, instagran, linkedgrim, peenterest, meatup and of course µspace? I understand that they are very popular these days (I don't use them myself).
I vote this as his official moniker!
pillsnotbills i think dave(the boss ;) mentioned somewhere udave not participating in any social media business.
Wow a bast from the past. I do miss these type of videos. More please Dave.
Nice work David & Dave!! Awesome to see different specific areas of expertise come together to make an awesome end result!
I'd like to see more of these, Dave. Really educational.
Thank you!
Nice to see the Usupply is still alive! Cheers for the videos Dave! Im a student hoping to study Electronics at university and all of this content excites and motivates me even more!
Thanks for pointing out this video Dave. I've been in the dark on this aspect of product design for years and this is really great.
PC PSU manufacturers like Seasonic have taken to mounting their SMT 12V switchers on the bottom of the PCB, with a thermal pad between them and the case. Sometimes a heatsink is added on the top side of the PCB as well to use both the cooling from the case and from the airflow inside the PSU.
Seems to work pretty well for them, even in 1k+ Watt units :)
Maya Posch The loss through the plastic package to the thermal pad would be massive. Most likely getting away with mostly power plane dissipation.
Here's an example of such a Seasonic design: www.jonnyguru.com/modules.php?name=NDReviews&op=Story5&reid=385
They're using Fairchild FDMS015N04B MOSFETs, rated for 150C: www.fairchildsemi.com/datasheets/FD/FDMS015N04B.pdf
It'd be interesting to see where the heat in such a design goes :) Seasonic was the first with their X-series of PSUs to put those switchers in such a location and I like to think that SS knows their business when it comes to designing a good/excellent SMPS.
EEVblog Yeah, but in this case, they aren't going to be dissipating 10's of watts, so they don't need it to be as good. Still, adding the extra thermal relief is a good idea, to help keep parts as cool as possible for a longer lifetime.
Maya Posch
They also have heatsinks soldered through the board that connect to the FETs though the planes. These are cooled by the fan or convective airflow when in fanless mode. I bet that is where most of the heat goes. The case is also steel, so not very good at conducting heat.
megabytephreak Yes, I mentioned that in my second comment in this topic. Maybe Dave could figure out with his fancy Flir equipment where the heat goes in such a PSU MOSFET cooling arrangement? :)
We would love to see the finished product. I hope you are planning to allocate some time for this product.
I really like the "Fundamentals Friday" episodes. Much more useful than mail bag segments or even the tear downs (though I like those too!)
Skip to 6:30 if you want to see the meat and potatoes (It's kinda a long intro of what Dave has already covered IIRC) - I do like those SMD heat sinks - I've never seen one before. thumbs up.
Would love to see more design videos similar to this. Very handy. Thanks
Hey Dave, great video. What I miss is a ballpark calculation of the actual temperature you'd have using this approach with the ucurrent. You show how to do it - and that's great, but I thought the actual numbers were missing.
thank you dave, this good news for the µSupply psu project
It seems to me if you are using thermal vias leading to the case, it's best to do that in the middle of the board, so heat spreads out in two directions on the case. Putting it at one end distributes heat in one direction along the case (efficient) and the other direction into the air (good insulator).
Here's an idea for the case mount heatsink option with the to-220 as Dave started with: Include a hole in the board (or drill) to fit a machine screw through. You can then screw down your regulator via a tapped blind hole in the enclosure! Better heat transfer B-)
nice video. got something valuable concerning heat dissipation. especially for SMD. thanks big time.
Really awesome series. Super helpful for me, but I would love to see it finished. It's been 4 years. Is Dave 2 still around? Thanks for what you've already done, Dave. Regardless of what you've done so far here, the thinking out loud, analysis, testing, mechanical concerns, etc., are so useful for everyone. That being said, we all hope you finish it up.
Dave, you know way more about design than I, but at 1:00 that TO-220 is going to need a screw to clamp it to the case. The leads can't provide enough force to provide good contact.
At 1:20 that heat sink should be isolated to avoid short circuits.
Chris W I mentioned that.
Great stuff. Thanks.
I knew there was a reason my mom drilled all those holes in my head. No more "hot head" stuff out of me!
Great video. I was hoping for a demo: how cool would it have been if Dave had done the calculations for the design, built it, run a few watts of power through it and then measured the temperature to see if it it stabilised to something close to the calculated value? That would would be great (particularly if the result actually matched the prediction !!)
And congrats to Dave #2 on the animation...very nice.
You probably dont care but does anyone know a trick to log back into an Instagram account..?
I stupidly lost the login password. I would appreciate any help you can offer me.
@Royce Johnny instablaster ;)
@Mathias Steven i really appreciate your reply. I found the site on google and im trying it out now.
Seems to take a while so I will get back to you later with my results.
@Mathias Steven it worked and I now got access to my account again. Im so happy:D
Thanks so much, you saved my ass !
@Royce Johnny happy to help =)
I use this exact technique with Class D amplifiers. But I fill in the vias, and make sure the heat-sinking surface is flat. I dont use silpads, I use kapton with thermal compound. It works ok, even up to a few hundred watts, but since Class D is efficient, not much is being dumped out as heat anyway. I do as much to the PCB as I can, and avoid machining the "valleys" into aluminum for other methods which requires extra tooling/machining costs.
yesterday i teared down optical drives donated by my school, and i've seen this technique on the h-bridge driver for the CD motor. What they did was the've made a cutout on the board under the thermal pad of the component, put the silpad and then on the metal casing they've pushed the metal inwards where the cutout in the pcb is, so it will make a thermal contact and disapate the heat there
Good stuff,
Interesting how some subjects in physics are this similar (Electrical and Thermal Resistance). Love to see a video on the uSupply new electrical design, Like you used to do with the original uSupply videos.
The 3D drawing shows the bar screwed from the case from the outside and from the pcb side also. This would ensure good thermal contact around the screw holes. Everywhere else on the bar, however may not be touching for thermal contact. You'd have to be pretty close if not just a shade too thick for the metal bar height or it could crack the PCB screwing down to it. Maybe you could do the same thing with some metal cylinder standoffs threaded from both sides instead of the bar.
Dave I love your channel but rarely login to comment. Your videos are very informative. Would it be possible to get more teardowns? They are educational beyond a specific topic, and very entertaining. Cheers.
David 2, µDave, D2, AfroDave, Mini Dave and Dave 2. These are some of the names given to your asistant. Honnestly the simplest is the best: we all know you as Dave and his name is David.
Any word at all when the PSU is going to be available? In any form!
never, he has not updated this in years
Right! WTF, I just watched 15 parts or so for what nothing! Ahhhh! How anti-climactic! I'm pissed, show us the baby!
Matthew Holevinski
you can grab the schematics from his website and make it yourself, if you want
My experiences with a typical design are these:
- Spread the force evenly, don't put bolts only in the corners of the PCB but also put some in between and around the components. You can use smaller ones and use less force, it won't deform the PCB material as much as 2 big ones in a corner as Dave did in this example. I made this mistake and had the PCB arc up in the middle. It even created an visual airgap between the bottom of the PCB and the aluminium bar it should've rest upon. Nobody wants that :(
- Adding a matrix of thermal via's underneath the large solderpad did wonders for transporting most of the heat... Only a small part was dissipated by plenary conduction into the via's surrounding the components. So plug and cap your via's if possible as that stops the Tin whicking effect so it won't hurt the soldering process.
- Adding thermal paste between the aluminium bar and the housing didn't help me much, there was so much surface area between the two that it didn't make much of a difference...
you can get thermal pads up to 5mm thick to avoid using a bunch of machined parts and fasteners.
digikey 926-1148-ND
Today, i learn something new ... SIL-PADS .... thanks
I was wondering when instead of using vias, you used jumper links (soldered both sides, sanded flat on the other side), if that would give any significant improvement? I know some people that do it for currents and claim it works great, and since current and temp flow is always similar...
In my home-etched boards, I do this often. I will also use like a thru-hole resistor lead and solder it on both sides, since the resistor will be there anyway.
Interesting!! Nice video dave!
Very nicely done. And I like that t-shirt !
Or may be possible to anodise the metal bar for electrical insulation. Risk is scratches or damage can break the insulation layer however...
Nice work in solidworks D2
Have you finisched the powersupply?
Hi Dave,
I like your fundamentals videos when you are in front of a breadboard, 10^6 times thumbs up!
One question, in which case do I want to isolate the ground/tab of a regulator from the ground plane of the circuit? Could you make a short video about it? Actually I have another video request, if possible? How can one succeed to lay out a single sided board avoiding the jumper wires? It challenges me as hell as a beginner...
Great video!
Hi Dave, excellent video as usual. One question, what would you do with the vias if they increase the stray inductance and cause voltage overshoot in a switching circuits?
One disturbing trend is that some manufacturers are using thick (~5mm) soft pads to transfer heat from large SMD ICs to cases. The problem is that over time they lose their ability to spring back and they leak oil all over the PCB.
I'm excited to see you resume work in the uSupply! Is there any place to get the gerber files for the PCB you made?
Pretty much how all vehicle (12v here in Amurika) amplifiers are regarding the heat sinks designs.
Hey Dave, do you have a part number for the proto board that you showed in the beginning of the video?
sparkplug1018 It's an Altium Nanoboard prototype board.
I wonder what the thermal resistance of the solder between the transistor and the copper pad is, and what sort of temperature drop you get across it.
SnPb (63/37) is 50 W/mK @ 25°C according to en.wikipedia.org/wiki/List_of_thermal_conductivities
(I googled)
You gas fill guys - Argon. Only way to go. Safe as well. Used it on a major heat pipe for a cyclotron gizmo at UCB (the hill guys from Livermore). Fun.
How do you remember all this? What do you do when you need info that are from University days? Or do you remember everything?
xDR1TeK I'm generally pretty hopeless at remembering theory, especially if I haven't touched it since uni, but thermal stuff like this is pretty basic and something I use reasonably often.
xDR1TeK I don't quite know how to respond to this without sounding arrogant, but once you've seen this stuff once, and properly understood it, it's going to stick. Visual memories of various component placed in different ways, you can imagine and play around with these various arrangements in your head. Remembering that things get hot isn't all that challenging. And all of the thermal resistance stuff? Looks incredibly obtuse and complicated, but may I draw your attention to the units of thermal resistance: degrees C per watt. For every watt you dissipate, this is how many degrees C difference you get. That's all you need to know, all the rest just falls out from common sense. I'm not saying that being guided through all of this for the first time by an experienced expert isn't valuable, quite to the contrary, but it's not like it's made up, illogical, or actually even remotely obtuse in any way. Do chefs remember everything from their chef school days? It's just a weird question to ask, really.
TheHue's SciTech I think it's an interesting question and touches on a lot of issues. If it's something I haven't touched since uni, have never really used it in practice, and didn't particularly like it much at the time, then it's unlikely I would remember it. Like for example if you asked me to solve an integral or something I probably couldn't do it, I'd have to go look it up. But once I look up the stuff it usually back to me pretty quick if not instantly. The good thing about thermal design is it's an easy concept that's hard to forget once you know it.
EEVblog Agreed on all points. I guess I saw the question (probably incorrectly) as coming from a "how do you memorize so many completely disconnected facts" kind of direction; I just wanted to point out that the same question could be asked of just about any profession or career, and that far from being obtuse and disconnected, all these facts slot together in a lovely and rational way. To an outsider who doesn't really follow, it looks far harder than it really is. Having said all that, my engineering ability drops 10x without Google by my side.
College/university education is as much about learning how to learn as it is about the specific topics you study. Even if you don't remember a specific topic, a good engineer can teach themselves what they need to know without having to have it memorized. At least that's my experience.
If the Vias filled up with solder would that improve the thermal transfer? significant?
***** Argh, yes, I was going to go back and shoot anther clip mentioning that, but forgot in haste to get the video out (locked out of the lab for the entire weekend due to building work)
EEVblog and to answer your question, yes quite significantly, but you can get issues with lumpiness in the contact surface which may not be good. You can also get the vias filled with solid copper, but that's a bit of an exotic technique.
Huh. This is the first time I found out it's spelled "vias".
I never heard the term before watching your videos, and just assumed it was spelled "veers".
What if I want to get rid of 80W of heat from a SMD power transistor? Or do I have to use TO-247 or TO-3P packages? My USB-operated dummy load dissipates some 80W power on IRLP3034N transistor but that is a through hole TO-247 part. Can I go to SMD parts on that?
陈北宗 80W is a ton, generally you would not do that with SMD.
EEVblog So better stick with TO-247/TO-3P for that level of power dissipation? I am trying to design a dummy load but with a design constraint that the entire dummy load have to be powered from USB.=, and that means smaller sense resistors and a hell lot dissipation on the MOSFETs.
Project finished by Thomas van den Dries: www.instructables.com/id/Digital-Battery-Operated-Powersupply/ (thanks to Luc De Meester for the link).
analog for temp - neat. Talk about magnetic circuit analogs some day...
I guess you could use insulating heatsink plaster instead of a Pad and a screw for a case dissipation method... much cheaper and elegant than a pad and a screw
If you didn't feel buying a licence for Solidworks there is a free parametric modelling open source equivilant called FreeCAD. I haven't used it all that much, and I'm sure that it isn't as feature complete as Solidworks, with the animations and physics modelling, but it is free!
could i use paper or cling wrap as a substitute for a silpad, or would they thermally insulate the part from the heatsink? i'm not exactly made of money.
ravener Not a chance. Silpads have an extremely low thermal resistance, and a high tear/puncture and voltage isolation rating.
EEVblog actually, i think a layer of paint or lacquer could be used. if you make the insulating layer thin enough the thermal resistance goes down drastically.
ravener Kapton Tape
Most effective heat transfer with isolation at the same time is made by using thin plates made of diamond. Only problem is to actually buy them.
Why don't you let the transistor stick out just enough over the edge of the PCB (which can be shortened accordingly) so that its metal part can touch a metal bar - coming from below - that is level with the PCB? Would make things much easier imho.
If plastic is so bad at conducting, how come some TO-220 packages of high power MOSFETs I order come 100% plastic (no metal back or tab)? Eg. NDF08N50ZG (500V 7.5A)
+Ryan O'Connor
thats because they have low power disspation rating, and they use a thin plasic and because it right on the stuff that heats up its not that bad.
but if you are trying to say that plasic is a good conductor, I have nothing to do here.
isolation, see the fact it's rated for 500v ( answering for future people going through the comments )
Either I missed it, or there was no mention of the following in this video.
While much if not most of the heat to be dissipated is conducted away from the source through the copper islands and through the vias, to the thermal connector bar and finally, to the metal case and into the air external to the case, a portion of that heat does not take this course, but is instead dissipated into the air trapped inside the case via the surfaces of the copper islands and the "vias" that provide a thermal connection(s) between those islands. And while the total amount of this interior-dissipated heat is probably minimal and immaterial relative to the total heat to be dissipated, and while this is probably not an issue for many if not the majority of applications, in theory there may be designs and applications that are in fact significantly impacted by this interior-dissipated heat.
That said, if the goal of this video was to undertake a proper examination of and to discuss and to characterize a thermal system, then using the argument above, it has fallen short of that goal in its failure to address an important aspect of that thermal system. Sure, I know. This is not intended to be a comprehensive scientific examination of thermal systems which could span an entire library of books and several scientific disciplines. But by my intuition, it would seem that this could prove to be a common shortfall of such a design and should therefore at least have been mentioned briefly in this video.
I should add that I don't have a degree or any professional experience in EE or thermal design. I am not an engineer (any kind) and the closest thing I get to being officially termed a scientist is by merit of my BS in Accounting. So, I am speaking mostly from the perspective of a hobbyist and a layman here. Bearing that in mind, I hope you will take this comment for what it is worth. I have chosen to make a point of this only for the potential benefit of those concerned and not to discredit the author, who is doing an amazing job and providing a wonderful service to the public with the EEVBlog.
aerofart I thought I mentioned that in a text overlay somewhere? It's not just the copper, *every item* in the thermal resistance chain (bar the silpad) has surface area that will radiate heat into the surrounds. I mentioned it was a simplified but practical model, and there are more complicated ways to analyse it all if you want to.
EEVblog It's all good, Dave. It's just exciting when you miss something because it happens so rarely. ;)
aerofart He mentioned every component radiating heat to its surroundings; you could also talk about natural convection ocurring inside and outside the case, but modelling that gets pretty meaningless pretty fast and you would just resort to trying it out.
What I believe you are missing from your argument is that you have a heat source and a heat sink in the system, the device pn junction and the case surroundings, and with this model you are giving a ballpark estimation of what temperature to expect on the junction relative to the ambient one. The air inside the case could be represented as a thermal resistance in parallel from the package to the case with a much bigger value, and its temperature you wouldn't care about as it would just follow what the heat source provides it.
These considerations may be relevant when you have a heavily temperature dependent device performance and a relevant heat source sharing the same enclosure, but that was certainly out of the scope of this explanation.
jossesphere Yeah, yeah. I know. I was nitpicking . . .
What scares me is the price of a custom alloy thermal bar and the additional assembly cost.
are silpads better than mica? - what about kapton tape?
no one?
Now try attaching a heatsink to a lidless BGA.
Isn't a SILPAD made of silicone (hence the name)?
Around 06:00, i keep anticipating him poking himself with that pointy end.
Finish This! Please!
I have 2
Scosche HD4004 HD Speakers
4 oms 30 wat RMS 120 Peek
will the Lepai LP-2020TI drive both?
See also Mike's Electric Stuff "Making nice-looking front panels with no setup costs" ruclips.net/video/Yj0Bv4UEFSs/видео.html
what happened to dumpster diving?
I can't believe they've run the dataline right throught the power supply. Amateur hour!!
NB. I don't want tot spread misinformation. So just in case you didn't get the reference: *it was a joke*.
50 FPS looks weird.
[*] µSupply
dave i need a good multimeter but as an student i dont have enough money for a fluke :(. and i dont want some crappy shit
EliteReacon Dave has the $50- and $100-multimeter "shootout" videos that you might enjoy. They are a few years old but I don't think much has changed. Sure there may be some newer models not discussed, but, except at the high end, the trend is not to improve quality, but to decrease manufacturing costs.
Some important factors in a multimeter are safety, accuracy, update rate and quality of leads. As I recall, Dave seemed to place too much emphasis for my liking on checking ruggedness by twisting the meters, but to each his own.
I have a UNI-T ut61c, it is more than excellent. Precise with alot of cool functions, and is safe as long as you dont work with anything high energy. Messing with household mains electricity wont be a problem, but isnt recommend.
But dont buy the ut61c if you need it higher resolution and true rms. Buy the UT61E. Which is true rms.
This meter never failed me. I love it
Anas Malas yes i look for the ut61d or so but i would buy the agilent on ebay if it would aviable in my country.
EliteReacon I live in uae. Here if you are lucky to find flukes. They are 2x the price. Don't even think about any other brand. But UNI-T based itself pretty good here. Its everywhere. That is why my experience is limited to it.
I thought David doesn't like it when people call him "David 2"? Or you are doing it on purpose lol?
WoahWoah378 He doesn't like being called Dave.
Gosh, you do yap on a lot........anyhoo, where is 16-20?
Too much talk, not enough thermal design.