Win could use a heat sink below the board, with a gap-pad between to conduct heat from the board to the sink. The gap-pad would conform to the features of the board and the heat sink, neither of which will be very flat.
His board has high voltages and probably needs low capacitances, so a hard or thin connection to a heat sink might not work.
I keep fighting that problem, the tradeoff between thermal conduction and capacitance.
formatting link
formatting link
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
150oC???? And that's just at the surface...whatever the failure mechanism of those resistors, you're heading for it at breakneck speed. Looks like those tapes are just used to glue heatsinks to components.
Well yes, but the CRCW1210-HP are meant for use at high power levels. Vishay claims a 0.75-watt rating up to Tambient = 75C, then they derate linearly to 155C. They do not specify large copper ares or other tricks. So it may be this resistor is meant to work hot.
Okay, so you're right up against the maximum. They did they happen to mention what goes along with 155oC operation, like permanent change in resistance and MTBF?
John Larkin wrote in news: snipped-for-privacy@4ax.com:
NO! PERIOD.
Whenever possible, a hard connection to a sinking element should be made.
Gap pads are for conductively cooled designs where one wants the entire assembly to homogenize (settle) out at a single temp.
Want proof? Look at how they attach chips to their sinking die hats.
It ain't with some sponge pad. It is a high conductivity matrix. And yes I am referring to thermal conductivity and shouldn't have to say that either except for dopes like you and keith childishly nit picking everything.
A technicality is that as a mediun 'conducts', the heat in the target device is 'dissipated'. So even your jab at me there is in err.
John Larkin wrote in news: snipped-for-privacy@4ax.com:
Hard anodized Al is a non-conductive surface treatment.
Three levels are (technically still) available. All three are sufficient.
The cost (of hard anodizing) prohibits most Mfgr's designs to incorporate it, so most off the shelf sinks are chem etch or simple anodize treatment, which is not insulative. AND if it is floating, it is no different than the exposed nodes of the cicuitry as far as voltage goes. And IF hard anodized it would not carry any at all, AND it could also get tied to ground if one is worried about that.
Winfield Hill wrote in news:q37hrg01sn0 @drn.newsguy.com:
We mixed fine fiber glass strands into our simple RTV potting mix and of course, since I was doing HV and this was HV, we used vacuum to fill all the voids. We did not have HV SMD sections to worry about until my smoke inhaler driver, but many of those multiplier section were hand soldered parts on top of solder bumps to purposefully raise the part off the board. We even had slots under our HV diodes that end up becoming RTV walls.
But the point is that the fiberglass filler made the potting conduct heat better.
Were you to go to the daughterboard schema, you could increase the form factor chosen for certain of those resistors, or simply go with tiny radial thru hole devices. let them radiate and put the sink on the larger device alone or the resistors might actually soak that part up to a higher temp on a shared sink. Maybe change the three SOT-23 parts for a larger form factor, higher rated part with more surface area on top to interface with a sink, which the SOT-23 is not really made for.
snipped-for-privacy@decadence.org wrote in news:q386r7$87p$ snipped-for-privacy@gioia.aioe.org:
The gap pads we used then conducted the heat to the case through a thinner pad cross section. A critical factor in getting a good theta number. Gap pads work (well) in some applications. Carrying off the heat from discreet parts running at their limits... not so much. I just wanted to clarify that, because I refered to both the pedestals and the gap pad as "pads".
But... that's not all that's important. The rest of the system, including the PCB, is also at elevated temperatures, and maybe NOT as insensitive as the resistor parts.
I've seen electrolytics with melt/scorch marks from nearby components, and PCBs that turned black. Good news: you can bead-blast a blackened bit of fiberglass-renforced-plastic, then reimpregnate the fibers with Q-dope, and put the electronic ignition right back into the Buick. I sold that car years ago...
It was easy with the Flir to separately measure the resistor and SOT-23 temperatures. One sense resistor value* favored the resistors, they cooled and the transistors overheated. After watching that scene at 1200-volts too long, with Pd at 3x past rating, there was a snap (small-scale explosion). After I calmed down, all the semiconductors were destroyed. One went first and took out the rest.
apportions current between the resistor and transistors. I won't be using that value again at 1200 volts..
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.