Jump to content


  • Posts

  • Joined

  • Days Won


Charixfox last won the day on October 12 2017

Charixfox had the most liked content!


19 Good
  1. This may be able to be fixed by (re)moving the contents of the plugins directory under iCue. I recall something about the Aura plugin being broken somehow, so I shut down iCue, moved the contents of plugins out, and restarted iCue. It would previously freeze up multiple times a day, and now it's been good for 72 hours. I recall needing to do this in the past, and things going bad again after upgrading iCue, so it may need to be done after every upgrade. Somebody may also want to investigate whether upgrading Aura stuff or other related lighting software corrects it.
  2. I'm officially sunsetting this project. Thank you to everybody who contributed and everybody who enjoyed it! The code is still available and people may improve on it if they like. I will still accept pull requests when appropriate, and I'm still happy to answer questions. What's next? Hmmmm... Maybe a controller based on an ESP33 instead. Bluetooth and WiFi anyone? Nah. Probably not. But you never know. ;) So long and thanks for all the fish!
  3. The lights are fed by the SATA connection, so honestly it sounds like you have a case of the computer not really shutting down. The PSU should need to still be providing power to SATA, which implies that the mobo still has it fully live even if little power is being used. That's much more than standby there. This is the kind of thing that can happen if the system works to shut down but fails to trigger the power transition. On Windows, that's usually an indication of something bad going on with the kernel or a driver.
  4. Yes. Think of it as "Everything goes -into- ground", so there's just two places from the controller going into ground. You can (and do) have dozens of things in the computer connected to ground intentionally and they all can and should be. I would recommend sticking to your above image to reduce complexity. While it -might- be fine, I do not know the specs of the controller, but just because it "can" does not mean it "should". Reducing voltage generates extra heat and strain on various components and increases the possibility of problems.
  5. As I described in my original reply, not having it connected won't hurt anything. The worst that can happen is that if it needs it, the electrical differential will not be enough to signal to the LEDs. This would only happen if there is some wonky pull circuitry in the controller. The hub is mostly just wires. Inside, it's a bunch of "wires" on a circuit board and one decoupling capacitor, so it's very hard to hurt it. The LEDs and the controller are much easier to hurt with things done wrong.
  6. Maybe. Probably not (98% chance you don't need to), but it depends on how the controller is wired inside
  7. Then: Yes, the controllers for the SP and HD fans are different, and the hub is the same. The controllers do different data signals. The SP fans take UCS1903 signals. The HD fans take WS2812 signals. Yes, the hub is exactly the same for all three kinds of fans: LL, HD, SP. The hub is mostly just a wiring system to get data signal from fan to fan. It has no brains in it. The three-button controller does not provide any power. The power (5V) goes from the hub to the three-button controller to power the controller. If the "Corsair" object in your bottom image is the HUB, then yes, it is correct. If it is a controller, it is not. You might need to go from the output ground (GND) on the SP110E to the hub ground also (both to the power and the output from the hub). Try it without. The LEDs will act wonky (or not at all) if it needs both and doesn't have them both, but it won't hurt anything. If you are using six SP fans, the SP110E should be in UCS1903 mode and set to a strip of 6 LEDs. If you are using HD fans, it should be in WS2812 mode and set to a strip of 12 times the number of fans. (24 LEDs for two fans, 36 for three, and so on). If you are using LL fans, it should be in WS2812 mode and a strip of 16 times the number of fans.
  8. I'll provide my input as well, since I posted this bit of fun... * "Looks good" is subjective, and I'd say that 1 LL at the back and three HD at the front would look good. * Yes, the project works with both and allows you to configure the fan types before you compile it. I have 3 LL and 6 HD for example running off the same controller. * The project requires the ATMEGA 32u4 chip. The linked Leonardo would technically work, but most folks prefer to prototype on bigger boards (Make something they can take apart and change around) and then build on more compact, purpose-built boards. So by that idea, a person would generally plan and fiddle on the full Leonardo board and then make the end thing on a Pro Micro so it doesn't take as much space. * Difference between my answer and the prior: Yes, the SATA power connector must be connected to the hub, and ONLY two wires should go from the Arduino to the hub. The Arduino does not pass enough power through its fuse to handle all the LEDs. If you know very well what you are doing and make use of the barrel jack at 5V and RAW on the Arduino, you can get away with it, but even I wouldn't because of other considerations. The Arduino will most happily power itself off the USB connection and then the LEDs are powered from the high-amperage SATA power connector attached to the hub. Extra information (You shouldn't need to use this, but it will answer your questions): VCC is a voltage-regulated 5V on the "regulated" side of the voltage regulator. RAW is on the UNREGULATED side of the voltage regulator. You can technically put, for example, 6 or 7 or even 9 volts in on RAW and it will work (and get hot on the regulator at 9-12 and possibly burn out at normal "12" in a car which is often 12.4 or more, not 12). The most efficient way to connect this project is the SATA power to the hub for the LEDs and the USB to the controller for 5V and control of the controller. * Given the shape of a female JST SM connector, you probably would not be able to get it into the hub. Also not that you only want 2 pins in the default build: Data and Ground. The 5V to the data end of the hub should not be connected since the controller in default should have a USB connection for controlling the controller and most motherboards these days still send power when the computer is off, which would cause an overloaded controller to try to power the LEDs through that USB power and the small thermal fuse. * Cautious: Do not connect anything backwards. Beyond that, the LEDs are relatively resilient. You're more likely to damage the Arduino than the LEDs. * Yes, and you should. Without a USB connection, the LEDs would look very boring. At this point, this project is an excellent learning and fun and less-expensive project. Corsair hardware and software has gotten a lot more capability and has convenience over this by far. This still has more capability overall but lacks some features (like easy plug and play and fancy software), and also my free time to develop it further has been cut into. But anybody can develop it further, which makes it a good thing too. So both have benefits.
  9. Thank you for the update! I didn't recall that the library lacked CRGBSets in that version, so the reminder is helpful. Have fun with the project!
  10. That impies that FastLED is missing as a library. Mea culpa for not documenting the need for the HID-Project library in the readme file, but it definitely mentioned the need for the FastLED library. Use the same process for FastLED. Let me know if the library was already installed. I suppose if all else fails, I can package the libraries in the code, though that has upsides and downsides.
  11. Hi Kino! In the Arduino IDE: Sketch Menu -> Include Library -> Manage Libraries Search for "HID" and one of the results will be HID-Project. Install that library in the IDE (Click on the library and an Install button should appear. Click that button when it does and follow any directions presented.)
  12. That could work. Pro Micro is a bit less-expensive. Another option is a DFRobot Beetle or equivalent: https://www.amazon.com/dp/B01B0IQFU4/ The BLE version uses an ATMega328, which loses HID capability and 20% of its SRAM, so at that point I find an ESP12-F with a power regulator and a Logic Level Shifter to talk to the LEDs a better idea. The ESP8266 can run Wifi with an access point or connect to other things, and thus there's no special program needed, just a web browser.
  13. Pretty much any version of a "Leonardo", such as the Pro Micro or the Beetle, as well as several variations on that with an ATMega 32U4. The 32U4 has built in USB handling which allows it to work with the HID. Hmmm... An ESP8266 has a few considerations... It runs at a 3.3v logic level, which is a little bit of a problem. Signalling to the LEDs is run by interrupts so some things can cause unexpected and unwanted behavior from the LEDs. I don't think the HID library would work on the ESP at all, so that would have to be torn out and all that benefit removed. The 3.3v logic level is not sufficient to signal the LEDs running at 5v. The LEDs detect high signal at 70% of VCC, which is 3.5v. So you'd have to have a logic level shifter or you'd need to drop the voltage driving the LEDs, which is non-trivial. You can pull SOME 3.3V off SATA power connectors, but the hub is hard-wired to snag the 5v. There are some other cheats like the sacrificial LED, but that has more complex wiring. The fact that LED signalling I think uses bit banging makes communications with the system a potential challenge without having LED updates collide with comms. At the same time, I can see a potential benefit to using an ESP8266: Web-based control from its own wifi network. Phone app anybody? <.< ^.^;
  14. The time has come to ask whether this project still has value to others, so in the thread at http://forum.corsair.com/forums/showthread.php?t=176109 I have done so. Feedback appreciated!
  15. The fact that TT (or anybody else) is using analog LEDs is sooooo annoying in this day and age. Either way, you are far ahead in knowledge compared to many folks, so you have an excellent head start. And there can never be enough prototype boards. ;) I tested my theories on a Mega before moving to the Pro Micro (Leonardo) for size reasons and control reasons. Other than the mosfet being a thing if the current would otherwise be too much for the thermal fuse on the controller, it looks like you primarily have things under control. Test all the things and see what you can get working successfully.
  • Create New...