• 17.09.2019, 23:20
  • Register
  • Login
  • You are not logged in.

 

Dear visitor, welcome to Aqua Computer Forum. If this is your first visit here, please read the Help. It explains how this page works. You must be registered before you can use all the page's features. Please use the registration form, to register here or read more information about the registration process. If you are already registered, please login here.

Thermaltake Fan / RGB Control With QUADRO and RGBpx SPLITTY 4

Monday, November 12th 2018, 10:36pm

Hello to everyone- this is my first post and after lurking for a while I decided I would start off by trying to give something back to the community. I discovered the Aquaero and Aquacomputer for the first time as I went through my recent rebuild, and I have to say I love the products.

Anyway, the reason I'm posting is because I have finally finished my efforts to reverse-engineer my Thermaltake fans to get them to work with Aquasuite's RGBpx control. I went with a Pacific W4 Plus waterblock and a 3-pack of Pure Plus 12 fans for my radiator as I liked the look of Thermaltake's RGB. Problem is though, just about everything else was lacking on the control end of things. The software is barely a shadow of the functionality offered by just about all other products, and having to have a separate control box using up an internal USB, power connection, and precious real estate inside my case was irritating.I spent a lot of time scouring the internet to see if anyone had gotten Thermaltake's fans to play nice with other products, but the information seems scarce. They don't want to post the pinouts on their stuff, and I couldn't find much that anyone else had posted. I got the fan control to connect to the Aquaero / QUADRO easily enough, as it turns out 4 of the pins are the standard fan connections and just needed to be broken out and connected accordingly. I wanted RGB though, and the remaining four pins remained a mystery. My breakthrough moment was finding this post-60LED/Meter WS2812 Strips, where it appeared an enterprising individual had managed to connect some Corsair HD120s and bog-standard WS2812 addressable LED strips to the Thermaltake controller. I reasoned that if these could work with the Thermaltake controller, then the RGBpx that can control Corsair HD120 fans would work with Thermaltake fans. The fact that it was essential to plug in fan #1 to connector #1, fan #2 to connector #2, etc on the Thermaltake controller also suggested to me I was dealing with WS2812 RGB LEDs, or at least a functional equivalent.

So, when my RGBpx SPLITTY arrived, I did some probing and came up with the following diagram for the Thermaltake fan connector (1st attachment). Note that the filled in / plugged / key on the fan connector is in the top left here, and that all fan related connections are on the top, with RGB connections on the bottom. It may also be worth noting that the RGB pins on the fans weren't color-coded- only the W4 waterblock had color coding on the RGB wires, so I opted to use those colors.



At this point I already had the fans control setup, but the SPLITTY let me break out the connections in a much easier fashion than trying to connect the fans together from the RGBpx connection on the QUADRO. With my breath held, I used some Dupont jumper wires and connected the RGB side of things to the SPLITTY, and was thrilled to see I had a fan I could control speed and RGB, all from Aquasuite. Without hesitation, I wired up the other two fans and the waterblock RGB (minus the fan connections as its RGB only) based on the following diagram (2nd attachment). Note that you need to connect things in order on the RGB side- i.e. start at port one and go from there. Since the WS2812 LEDs are really just one long addressed bus, the "Data Return" connection gets passed along to the "Data Out" of the following connector.



Complete success. I have all 39 LEDs (9 per fan and 12 for the waterblock) individually addressable and can now ditch the garbage Thermaltake software and controller. The difference is night and day. Now I just need my D5 Next and reservoir to get in to replace my old pump and I'll have the setup I've been chasing since I started this whole project.It bears mention that Thermaltake has something like 4 different types of fan control schemes now, so your mileage may vary depending on your hardware. I'd wager that if you're using the newest generation with the 9-pin connection you should be good, but either way I'll leave links to the actual products I used in my relevant links below. I hope this helps someone else save some time rather than having to do all the research I went through, and if anyone has any questions or thinks I left something out, please let me know!


Thermaltake Hardware:
Pure Plus 12 (Premium Edition)
Pacific W4 Plus

Posts I used as reference with pertinent info:
60LED/Meter WS2812 Strips (post on Corsair forums describing connecting Corsair HD120 fans / WS2812 LED strips to Thermaltake controller)

Floe Riing RGB Premium Edition: No PWM Fan Control Outside of Windows (post on Thermaltake forums describing fan connections on a 9-pin connector)
Edit: fixed attached images so they show up correctly

This post has been edited 1 times, last edit by "watchound" (Nov 12th 2018, 10:39pm)

DRF

Junior Member

Friday, July 19th 2019, 12:28am

Thank you very much

Friday, July 19th 2019, 2:27am

No problem at all! I'm thrilled to think it helped someone and saved a bit of effort having to reinvent the wheel and reverse-engineer things.

Also, in case it matters to anyone- the setup I've documented above has been running very well since I set it up. I had some oddities with my RGB lights flickering and strobing intermittently after I also added an Aquacomputer reservoir RGB light ring (I forget the exact model; they're generally the same beyond LED count anyway), but that has since gone away as an issue. I suspect it may be my soldered extension cables to reach the ring (it is on my reservoir that is external to my case; made due with what was available at the time) and my moving it up onto a table may have helped jiggle whatever connection, but I suspect the new Aquasuite version also helped as it seems there have been many improvements in the RGBpx realm in regards to stability and the like. It may also be that adding the LED ring on the end of the RGBpx circuit introduced more noise or a timing issue to the system, but I didn't try moving it forward in the RGB "lineup", so I can't be 100% sure.

Either way, I'm glad if this helped anyone else. If anyone ever has any questions then please feel free to ask on this thread and I'm happy to do what I can. I make no promises insofar as prompt responses- I sorta missed at least one PM for like a week; idiot Gmail spam filter- but please don't anyone think I'd ignore them. I'm more than happy to share the knowledge and try to help folks integrate their setup into what has easily been the best PC cooling management system I've ever had the pleasure of using.

DRF

Junior Member

Friday, July 19th 2019, 12:45pm

Actually i'm trying to reproduce what you've done but with 6x Riing Trio 140mm Product on TT, so i could have one PWM regulation system included with one RGB control system inside the case without worrying about cheap MB/TT software control or external controller.

• I figured out what was the Trio diagram
1st attachment

• I made a schema of the setup
2nd attachment

So i'm willing to use the relatively new FABWERK360 which is basically a "QUADRO" but only for RGBpx control with 4 channels of 90 LEDs each with 5V x 8A = 40W max power.

- Do you think there is any problem trying to do connect Trios to the SPLITTY4 ? Since on This TT RGB guide, he does say that Trios are not compatible with other TT controller than the Trio one, is it just about the order of the 4 RGB pin ?

- Do you think there is any problem on the schema at all ?
DRF has attached the following image:
  • SETUP2019(2).png
DRF has attached the following file:

This post has been edited 2 times, last edit by "DRF" (Jul 19th 2019, 1:01pm)

Tuesday, July 30th 2019, 12:07am

Sorry once again for the delayed response- work has been shipping me to remote places and I've been tied up holding that particular set of disasters at bay.

I'm a little hazy on the differences with the Trio vs the generation 3 fans I used in my setup, but I would suspect it should work. From what I can tell, the should both have the same pinout and same RGB LED chip controlling them based on the fact that the "TT Sync" controller is compatible with both Gen3 and Trio fans. However, what makes me a little worried there is the note they have about the Sync controller and the max 5 vs 4 Trios, as well as the big caps warning you referenced: "DO NOT PLUG IN RIING TRIO FANS IN THE OLDER GENERATION 3/PLUS GENERATION CONTROLLER." I would suspect that may be because the Trios draw more current than the previous Gen 3s given that the Sync controller is apparently pin-compatible with both, but its also possible Thermaltake did some actual engineering in an unusual breaking of character and have some kind of switching hardware to detect the fan model inside the Sync controller that isn't in the controller supplied with either fan type.
As far as connecting the Trios to the SPLITTY4- unless they have some crazy connector then you should be good. That's what I did with my Gen3 fans to get them on the Quadro, and beyond the ugly connections using Dupont jumpers it works fine.

Your schematic looks sane to me, but a fair warning- I haven't used a FABWERK360 so I'm not 100% familiar, but from what I've read I do agree on your assessment in regards to it essentially being an RGBpx-only Quadro. Using the Quadro to drive the fan side of things makes sense and I don't see an issue there unless it exceeds power consumption (I don't remember the Quadro max values off the top of my head). I would advise you to be extremely careful in how you hook that up when you're rewiring things and make sure to check for any shorts with a multimeter before you power it up though. If anything connects the fan and RGB sides of the setup (stray wire, short on the bottom of the SPLITTY from touching the case, etc), it might make some pretty expensive magic smoke. I made sure to put a patch of duck tape on the bottom of my SPLITTY just for this reason.

Your picture of the wiring to the fans is particularly interesting in regards to the compatibility question also. Going from the PCB it looks like you have [ 5V | DO | DI | GND ]- which, assuming those line up directly to the RGB pins on the connector- would make for a near-match on my initial diagram: [ 5V | Data Return / DI | DO | GND ]. I'd say that the important part is that the 5V and GND pins appear to match. That leaves DI and DO- the serial pins used to control the actual RGBs individually- to be sorted out. It could be the pins are swapped on the Trios vs Gen3s, or the wires themselves are swapped in their positions on the way to the actual connector. You can work around the former by swapping them in whatever you end up using for a wiring harness, and if they're the latter then they would match the Gen3 connections.

I think you have alot of it figured out at this point, but my best advice would be to have a go at testing incrementally and use a technique similar to what I did- get a spare PSU to test with, open up the Thermaltake controller that came with them, hook up a single fan for testing and while the testing harness is powered up- I can't stress the carefully part enough here- very carefully test the pins on the RGB (i.e. the bottom set of pins) side of things to confirm the pinout. These Thermaltake control boxes have big 90 degree pin connectors inside to connect to the actual port, so a multimeter with some smaller alligator-clip probes worked well. From what I've seen you should only have to be 100% certain on which pin is 5V and which pin is GND. The DO and DI pins can't easily be sorted from each other without an oscilloscope or digital logic analyzer (or maybe a BusPirate in a pinch), so I was iffy on those, and my wife didn't really go for the idea of dropping $1k+ on an oscilloscope. If you are absolutely certain you have 5V and GND correct though, I think you've largely minimized the potential for damage if you swap DO and DI by accident. From what I remember in my research the WS2812-style chips that drive most of these RGB widgets run on 5V TTL logic for the actual RGB control (which is why the 5V and GND are the most important to get right), and thus like many serial-style buses have a decent tolerance for swapping DO / DI (Tx / Rx in serial parlance, respectively). That said, if you go for the test rig approach and power it up but don't see the LEDs react or behave as you'd expect, don't leave the test rig powered up. Though 5V TTL can usually tolerate having their input / outputs mismatched (i.e. DO -> DO and DI ->DI instead of DO -> DI and DI -> DO), that doesn't mean they like it and it could burn up something if left for too long. If you want to go down this road and have any questions then let me know; I think I still have one of by disassembled controllers from my Gen3 fans around and I can post some pictures to illustrate.

For what its worth also, I seem to recall seeing an Arduino project somewhere that would control these kinds of RGB setups. I mention it only because the Arduino generally has good protection for bad wiring so it might be a safer alternative for testing, assuming you really want to be certain. The 5V and GND lines of the fan would unconditionally have to be powered via an external supply as the Arduino can't put out that much forward current to actually light the LEDs without burning out, but it would at least allow for a test harness to test the digital control lines on the LED ICs themselves.

DRF

Junior Member

Wednesday, July 31st 2019, 7:55pm

Man... thanks !
I effectively think that i will test it with your advices. I finished my waterloop yesterday (with only 6xTrio and 3x NB e-Loop running - need a SPLITTY9 lol) and thanks god everything is running well.
Just for the anecdote i lost all sensitivity in my fingerprints because i am on a Primoflex tubing / Bitspower barbs 3/8 5/8 and it's a big pain to screw this (not the same manufacturing norms/ or the metric system... i don't really know) i had to go with a kettle at 90°C with the tubes soaking inside before screwing barbs... phiew that's done.

I will order the pieces and re-read your advice before testing. I'll update after that.

I can't wait to finish this !

EDIT: I added something i got in my head. I will try to integrate this Y-PWR to start the full cooling system whatever the system is powered on (it's a dual system cooling). As aquacomputer pieces can be autonomous (i'll try to drive them with custom sensors inside case) it might be good. I just have to figure out how to convert the mini Fit Jr into FAT "molex". Problem with 5V i think quadro and RGBPX pulls some 5V from FAT molex and i don't feed the 5V with the Y-PWR

Have an easier way. I will just put another low power semi passive PSU inside case with a button plugged into ATX24 to power up the cooling system.

This post has been edited 3 times, last edit by "DRF" (Aug 2nd 2019, 12:43pm)

Wednesday, August 7th 2019, 9:34pm

Congrats on getting the loop going- that's always a big relief. No matter how many times I recheck my lines I'm always paranoid about leaks; I burned up a graphics card years ago from a leak I didn't catch in time, and I suspect its ghost will haunt me forever.

I hear you on the Primoflex tubing too. I use it as well along with- if I recall correctly- XSPC fittings and its a pain to get them tamped down all the way. I usually end up resorting to vice-grip / locking pliers and/or an adjustable wrench, but I'd never considered your idea of soaking the tubes in hot water. I'll have to try that. Its far too late for feeling in my fingers these days though- I lost that (and probably some of my fingerprints) due to numerous repeated soldering iron burns years ago. lol :rolleyes:

I'm not sure I understand your objective with the Y-PWR, but assuming I am, the concern is power consumption? It seems like you have some good ideas, but I would offer you may be able to get around the whole thing by checking consumption manually to see if it will be an issue in the first place. Your PSU should give you maximum watts you can pull on the 5V / 12V Molex rails, and coupled with the Quadro / Fabwerk docs you have the ceiling wattage consumption. In this case, I'd assume the restraining factor would be the amount of wattage the Quadro / Faberk can push, so PSU max may be irrelevant. Either way, most multimeters can measure current- you just need to set the mode correctly and connect it in serial with the load rather than the parallel connection used for voltage (in between the 5V supply and the 5V connection to the RGB, for example). You can test with one fan and see what the current value (probably in the mA range for RGB) is, and from there just plug it into the standard wattage equation: Volts * Amps = Watts. Watch the units though- for example 500mA == .5A, so you'd need to convert to get the right data- 5V * .5A = 2.5 Watts. From there you have an idea of one fan, so just multiply by number of fans and maybe assume you'd need 5-10% more than the total due to the nature of napkin-calculations. If you go down that path, feel free to let me know if any of that needs clarification.

Regardless, good luck man. I hope if turns out well for you, and assuming it does it sounds like you'll have one hell of a killer setup. I'd love to see what it looks like when you finish if you want to post some pictures. I've got to flush my loop really badly, so I could use some inspiration for new upgrades while I'm in there anyway lol

Edit: spacing

This post has been edited 1 times, last edit by "watchound" (Aug 7th 2019, 9:37pm)