Those module were well known not fitted with original LM2956 and the common failure mode is short circuit the Vout to Vin
Yeah, well... how about THIS failure mode...? Is THIS ONE well-known?!? (fuck you, bd... fuck you very cordially, you jinxing bastard. )So I figure
"Okay... drop back & punt." and decide to reassemble it. I can tell the whole damned thing was populated BY HAND.
Well, maybe by FOOT, I dunno.
Fortunately I've got my big chisel tip and can get enough heat to solder the DPAK sink to the substrate, and then I go through each & every component, pad by pad and make sure it's all soldered correctly and no shorts except the ones I'm wearing... (no, I have NOT learned that lesson YET.
) and... bupkis. I got continuity through to both inductors, I got Vin at the right pin, but like 300mV output. Caps test good, nearest I can surmise the circuit based on the datasheet, everything is actually connected as it should be.
By this point, I've had my fill... hunger and lack of caffeine drive me to the kitchen, and after some fumbling there's a fresh pot percolating. I step outside on the balcony with half a cuppa in one hand, rolling the problem around in my mind while rolling the board around in the other hand. The sun just peeking over the trees provides clarity in more ways than one; I put the cup on the rail and just fling the damn module as far as I can into the woods behind my back yard. Instantly my stress & aggravation level drops by 20dB; I step back inside & while I'm buttering my morning bagel,
a simple plan for a simple man...
err, dwagon... emerges.
Engage
"getting the fuck on with my life" mode.
I have some little MP1584EN-based switchers on the way, but they'll take a week or six because China,
so all I REALLY care about at this point is powering the 80mm 12V/0.12A Arctic silent case fan I just got in; it replaces one of the screamin' mimi 40mm fans on the printer, the one that cools the mainboard. While I AM pretty certain there's a setting where I could dial down the PWM for that fan channel in the firmware so it only sees 12 of the 24V from the PSU, I'm already annoyed and decide to apply the KISS principle. 500mA L78M12, a 47uF/16V MMLCC cap strapped across the output for stability and a bit of 0.100" pin header.
"But wait..." you cry out
"There's no heat-sink, and that's still ~1.5W or so dropped across the poor thing." But there is: 200mm of 18ga copper soldered right to the heat-sink tab. And another similar mass of copper on the Vout. I gotcher Thermal Design Considerations right heah, bd. Yes, it MATTERS. Testing
in situ reveals it gets just warm to the touch; maybe 45-50°C after 20 minutes. I sleeve it with some heat-shrink and plop the fan in place; put the bottom cover on the printer and set it upright. All told I've spent maybe an hour; that includes making a baffle and a filler panel for where the 40mm fan used to be.
I really didn't expect it to make much difference yet; the hotend still has another 40x10mm & 30x10mm screamers on it. But it IS noticeable. Now to print the Petsfang layer-cooling duct so I can install these much quieter 50x15mm & 40x20mm fans before the wife kicks us both to the curb.
mnem
*hammer-hammer... chisel-chisel...! braaap-braaaapppp...!! griiiiiiiiiiiind....!!!*