--
John Larkin Highland Technology, Inc picosecond timing precision measurement
jlarkin att highlandtechnology dott com
--
John Larkin Highland Technology, Inc picosecond timing precision measurement
jlarkin att highlandtechnology dott com
Yup, the resistor, ITSELF, may handle 11 W, but after the solder melts and it falls off the board, or the board catches fire, it will stop working. That's totally CRAZY to rate it at 11 W. I have some snubber resistors that are in an 1812 package, I think, at 2 W rating, and that is really pushing what the board can dissipate.
Jon
Data sheet says 2.4W @ 25C with standard PCB mounting. 11W with active temperature control, whatever that is-- water, maybe?
Your post subject is a bit misleading, John.
I posted a link to a data sheet that specs an 11 watt resistor. That's not misleading. Everybody is welcome to read the data sheet.
I have two resistors in parallel that are getting too hot.
and they are going to get a lot hotter as I crank the pulse rate up.
The end caps are well heat sunk, but the heat isn't being conducted well from the hot spot. The AlN resistors should be a lot better. I'll know on Monday.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
What do these look like through a VNA? Can you check? 50 ohm suggests (to me) low power RF dummy load - and an extremely compact one!
-- This message may be freely reproduced without limit or charge only via the Usenet protocol. Reproduction in whole or part through other protocols, whether for profit or not, is conditional upon a charge of GBP10.00 per reproduction. Publication in this manner via non-Usenet protocols constitutes acceptance of this condition.
I don't have a VNA, but I could TDR them, when they come in. I'd expect them to be about the same as ordinary 1206's, good to maybe 5 or 10 GHz.
There are tiny flange-mount AlN and BeO RF terminations, up into the hundreds of watts. Of course the heat has to go somewhere, so they need heat sinks.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
/me searches digikey for 3mm fans and heasinks...
MELFs would be better IME.
Cheers
-- Clive
Probably. There would be a lot more alumina to conduct heat to the end caps. I don't think convection cooling from the body itself will be a big help. My main cooling will be through the copper pours, to the ground plane and to the bottom of the board. I have gap-pads and an aluminum baseplate below.
I have a schottky diode that is getting too hot too. It's going to be an ugly kluge to fix that.
Damn, there's going to be a rev C.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
If you don't mind an ignorant question, since there is real estate occupied by the copper pour and other heat sinking, isn't it more reliable or just easier to use the space for a bigger resistor or more than one?
Seems like MELFs are overrated by old people.
No one likes them today. They roll off before reflow. They're only useful with glue and wave. They don't have specs comparable to, say, high power or pulse rated chips.
Right, the problem is distance from element to nearby metal. For a given aspect ratio (i.e., traditional chips), power goes more or less as sqrt(size), because of thermal diffusion.
So you have to change the aspect ratio to get more power. Wide format resistors are great for this, though it also helps to get ones that aren't cut in the usual way (i.e., single element with an 'L' shaped laser trim cut that concentrates heat at the end of the cut).
Probably, thermal pad right on top would do a lot for power, too.
I've used SOD-123F diodes good for an amp or two. Sure feels weird, but they don't run too hot. At least, not at any point that I can see, but who knows what the junction temp inside is.
Last time I did that, it was for a DC application; the heat just helped. You might not want to run one that small, at ratings, for switching...
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: https://www.seventransistorlabs.com/
No, you're thinking of MILFs.
I have two 1206s now, and there's not much room for bigger resistors. But I still think that the heat has to be conducted, through copper pours and vias, down to the baseplate, not removed by convection.
The resistors have nasty hot spots, but the copper pours aren't very hot. So what I can use is not bigger resistors, but resistors with better thermal conductivity.
This board has high power density, and especially hot spots of high power dissipation in small parts. That's a big problem in high-speed electronics.
The GaN fets each dissipate over a watt and have a surface area of
1.7e-6 square meters. That is about a megawatt per square meter. The heat flux through the solder balls is several times that. I think I have them under control.-- John Larkin Highland Technology, Inc lunatic fringe electronics
almost:
NT
Pah, real men use real resistors with coloured bands!
They're getting a little harder to find, sometimes a particular value may be out of stock, but they're still used a lot for high temperature stuff. Personally, I find them easier to handle with tweezers, and I really don't see any issues with reflowed PCBs. Lots of paste helps.
Cheers
-- Clive
541-2659-2-ND/5481873
But they do have to read the data sheet to realise that the 11W is conditio nal on keeping the body of the resistor below 150C. The AlN substrate will conduct heat from the resistive element to the - extra-large - mounting pad s, but you presumably have to put a thick copper pour on the inner layer of the board directly under the resistor and it's mounting pads to conduct 11 Watts away without eating too far into your 150C temperature rise at the r esistive film.
That's digging down quite a way.
The end caps are bigger on the 11W parts - they fill most of the area under the resistor.
Your printed circuit layout will have to be changed to exploit this.
With standard resistors the heat path is along the resistor to the end caps .
With these parts it is mostly straight down through the AlN substrate to th e bigger solder pads, and you've got to put enough copper on an adjacent in ner layer to allow the heat flux to spread out without running into too muc h in the way of extra thermal resistance.
-- Bill Sloman, Sydney
Here's the gadget, with one of the two paralleled resistors the standard 1206, and one replaced with the Vishay.
The AlN resistor isn't much hotter than the copper pours, and the old one is about 15C hotter. What I really need is more copper.
This is pulsing at 1 MHz. Two of the Vishays in parallel should survive at 4 MHz.
There are no fine-pitch parts on this board, so I could go to 2 oz copper, at least on the outer layers.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
how much does a forest of header pins gain you thermally?
NT
Without a lot of forced-air flow, almost nothing. Pin-fin heat sinks are about worthless in still air. My heat path is topside copper pours to vias to bottom copper pours to gap-pads to a water cooled aluminum baseplate under the board. Overall, the board is pushing a kilowatt per square foot.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
+1
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.