I feel like the V BE would be insufficient for the drive to see it as the mitigation signal (max 225mV), though.īefore I go designing this backplane without LED driver transistors, does anyone have a better explanation as to what the transistor is doing in this case that makes it necessary? It's some form of mitigation for staggered spin-up (which also implemented on pin 11, but only if the pin is externally pulled to 0V on drive initialization). Original design was intended to use the SATA 3.3V rail, but someone realized they could source 3.3V right from pin 11 (and that SATA 3.3V is optional anyways), and then for some reason (?) never bothered to remove the transistor. Wouldn't this make the transistor redundant? And even the most efficient LED would be drawing 10-100x the current specification of the pin, so that specification must not matter at all? (I am aware that it would turn off the LED on activity instead of the opposite, which is fine by me.) When the signal is pulled low, it removes power from the LED and deactivates the transistor when in steady-state, 3.3V would power the LED and activate the transistor. Below is the schematic I worked out.īetween my PhD EE friend and me, we don't see a need for the transistor at all. I happen to have an IcyDock 5.25" to 3.5" adapter that uses this drive activity pin for it's own LED, so I decided to poke around and copy their homework, so to speak. I've found the SATA spec references, which list pretty paltry current ratings for this pin (on the order of 100uA), so most examples I've seen of others hacking their own end up using a BJT to control an LED. The signal is active low, with 3.3V provided when not driven low. Part of the SATA standard includes provisions for drive-controlled open-collector activity status routed through pin 11 of the SATA power connector. I'm working on a personal project to add a SATA backplane to my NAS (shhh, it'll be it's own post when it's done, no spoilers ). Though I may swap out the SATA connectors for a single SFF-8643 connector. I don't know why the people writing the docs couldn't have put in a transistor like they did on the left drawing.Īnyways, time to modify my design and respin it (kinda sucks that it's ~$50 a run). V+ = 5V, R1 = 1k (horizontal), R2 = 330, and the "LED Driver" is the transistor + the 1k (vertical) on the base. See below for updated schematic:Īnd convergently, thanks to cycon at who has demonstrations of exactly this circuit and partially made me want to go back and recheck my reference.Īnd the above circuit is actually in the SATA documentation, just somewhat obfuscated (see the right one): and to my chagrin found a trace to 5V that snuck underneath a bunch of silk-screen where I could barely see it. So I finally got out my reference part and took and even closer look. I was disappointed because "I literally copied the example, why doesn't mine work?" I implemented my original design into my SATA backplane project, but the activity LEDs were. Backed by StarTech.Okay, one year on follow up! Because I've realized that a) I made a mistake in the original drawing and b) this post is apparently the top reference for on Google for "SATA pin 11".Eliminates the need to upgrade your existing SATA power supply in order to use a PCI Express video card.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |