*This post was about an engineering sample of REV10, production boards have the BAT40 and don’t leak
REV10 engineering sample boards arrived. New buffer works, but the cheap Schottky diodes from my favorite component manufacturer are LEAKY. Bought the Diodes INC version. Installed one but the problem disappeared. Why? Cold morning.
I roasted the board on the radiator a bit to enhance the leakiness. The genuine part on io6/7 matches the datasheet and has half the leakage of the generic part.
It also looks like the generic part datasheet has screenshots of the graphs from Diodes INC datasheet.
I like this manufacturer a lot and have no qualms about using their parts, but stuff like this does happen. I try to make sure to only use them where the tolerances barely matter.
In this case I’m going to test one more Diodes INC part and use the real deal at 10 x the cost.
Replaced five BAT54 Schottky diodes with BAS40. The datasheet says they’re 10x less leaky. Tossed it on the radiator to enhance leakiness. Indeed, even at abnormally high temps there’s way less leakage.
I managed to damage the LCD flex connector trying to “protect” it from heat though, need to fix that after lunch.
This will go to FCC testing tomorrow, and we’ll order production PCBs now because that also takes a week. I’m hopeful testing goes well because we’ve passed twice with approximately the same board.
I received BAS40 samples from CBI (cheap jellybean parts) to replace the expensive Diodes INC version. It is way less leaky and will work just fine.
Theres 5 of these on the board. Diodes INC is .32rmb, CBI is going to be .14 or less (maybe 0.07). That saves one RMB per board. But the major advantage is that a reel is 200rmb instead of 1000rmb.
@ian so if I got rev10 is it recommended to replace those diodes as well? I saw them locally I can get those BAS40 but there are different kinds and I dont know what exacly to look for