Raijintek triton Liquid cooler. Where to plug fans and water pump?

[Wort]

Hey guys. So 2 months ago I bought Raijintek TRITON cooler.
In the package along with the cooler I got two Molex to 3pin connectors, so you could plug waterpump and radiator fans directly to psu. After a week or so I RMA the cooler because it leaked, and today I got a new one (same model). This time, there were no molex to 3pin adapters included. It looks like there's new User manual, and there's not a single word about those 2 connectors. It looks like Raijintek took a different approach to plugging fans and waterpump.

Now I'm wondering, since there's no instructions about plugging the water pump, where on motherboard is it safe to plug those?

I connected fans to case's speed controller along with the other case fans (got Fractal Design define r4), but I don't know where to plug waterpump. It's currently plugged to PWR_FAN, but I have no idea whether this is safe or not.

Help would be greatly appreciated.

 

[BonzaiDuck]

Hey guys. So 2 months ago I bought Raijintek TRITON cooler.
In the package along with the cooler I got two Molex to 3pin connectors, so you could plug waterpump and radiator fans directly to psu. After a week or so I RMA the cooler because it leaked, and today I got a new one (same model). This time, there were no molex to 3pin adapters included. It looks like there's new User manual, and there's not a single word about those 2 connectors. It looks like Raijintek took a different approach to plugging fans and waterpump.

Now I'm wondering, since there's no instructions about plugging the water pump, where on motherboard is it safe to plug those?

I connected fans to case's speed controller along with the other case fans (got Fractal Design define r4), but I don't know where to plug waterpump. It's currently plugged to PWR_FAN, but I have no idea whether this is safe or not.

Help would be greatly appreciated.

Usually, PWR_FAN isn't a thermally-controlled fan option, but only runs the fan at its default RPMs, amperage and voltage.

There should be a PWM CPU_FAN port, possibly at least (at least!) possibly a second CPU_FAN_(OPT), one or more CHA_FAN ports on the motherboard, but this depends on the motherboard. Even so, if the board is less than 5 years old, I would expect that one or more of the CHA_FAN ports are also PWM 4-pin plugs.

With only motherboard connections, you can divide fans into two groups, if the mobo software allows thermal-control of CPU and CHA_FANs separately. You could then use something like a Swiftech 8W-PWM-SPL-ST splitter:

http://www.frozencpu.com/products/2...able_Splitter_-_SATA_Power_8W-PWM-SPL-ST.html

And you can control the two groups of devices from only two PWM ports on the motherboard, without drawing power from the motherboard. They would be powered by the PSU.

You can use those splitters for a mix of fans and pumps, so that anything connected to such a splitter would increase or decrease its duty-cycle % as temperature for the CPU rises and falls.

However, barring all these possibilities, to answer your question I would connect the waterpump to the CPU_FAN port. Better -- connect both the radiator fan and the pump to that port with a Swiftech splitter.

I'll give a disclaimer: I don't use watercooling at the moment, and someone who does may offer a different -- or better -- opinion about it.

I did take a look at that cooler as an item at Newegg. I'd suppose it performs fairly well, for what it's designed to do. On reviewer had an i7-4790K Devils Canyon processor: at its rated 4.4 "Turbo" mode, his IBT "Maximum" burn test readings were about 60C. Clock the processor a couple hundred Mhz higher than that, and the temperatures would rise into the 70s at least.

Based on the comparison with the Hyper 212 EVO air-cooling heatpipe tower, I'd wager I can beat the Raijintek with my own air-cooled setup by about 1C degree for similar processors and tests. Just a guess, but it looks like that.

I'd say it's every bit as good as an H110i or similar cooler. Hopefully, the replacement won't leak for you.

 

[Wort]

Usually, PWR_FAN isn't a thermally-controlled fan option, but only runs the fan at its default RPMs, amperage and voltage.

There should be a PWM CPU_FAN port, possibly at least (at least!) possibly a second CPU_FAN_(OPT), one or more CHA_FAN ports on the motherboard, but this depends on the motherboard. Even so, if the board is less than 5 years old, I would expect that one or more of the CHA_FAN ports are also PWM 4-pin plugs.

With only motherboard connections, you can divide fans into two groups, if the mobo software allows thermal-control of CPU and CHA_FANs separately. You could then use something like a Swiftech 8W-PWM-SPL-ST splitter:

http://www.frozencpu.com/products/2...able_Splitter_-_SATA_Power_8W-PWM-SPL-ST.html

And you can control the two groups of devices from only two PWM ports on the motherboard, without drawing power from the motherboard. They would be powered by the PSU.

You can use those splitters for a mix of fans and pumps, so that anything connected to such a splitter would increase or decrease its duty-cycle % as temperature for the CPU rises and falls.

However, barring all these possibilities, to answer your question I would connect the waterpump to the CPU_FAN port. Better -- connect both the radiator fan and the pump to that port with a Swiftech splitter.

I'll give a disclaimer: I don't use watercooling at the moment, and someone who does may offer a different -- or better -- opinion about it.

I did take a look at that cooler as an item at Newegg. I'd suppose it performs fairly well, for what it's designed to do. On reviewer had an i7-4790K Devils Canyon processor: at its rated 4.4 "Turbo" mode, his IBT "Maximum" burn test readings were about 60C. Clock the processor a couple hundred Mhz higher than that, and the temperatures would rise into the 70s at least.

Based on the comparison with the Hyper 212 EVO air-cooling heatpipe tower, I'd wager I can beat the Raijintek with my own air-cooled setup by about 1C degree for similar processors and tests. Just a guess, but it looks like that.

I'd say it's every bit as good as an H110i or similar cooler. Hopefully, the replacement won't leak for you.

wow, this was a really good answer and I really appreciate your effort to write this

I do have some question though. I always thought CHA-FAN and PWR-FAN are the same ports, with 1 exception, that you can control CHA-FAN and PWR-FAN works constantly at full power(?) That was my understanding of these 2 ports.

Well, Raijintek used to include the splitters, but with this new package it just wasn't included. They even deleted it from the user manual. you can see here: http://www.raijintek.com/en/products_detail.php?ProductID=18 in "UNBOXED" tab.

I have AsRock Z87 Extreme4 motherboard, so it would be expected that it does have all the required ports. (http://www.newegg.com/Product/Produ...69&cm_re=z87_extreme_4-_-13-157-369-_-Product)

 

[BonzaiDuck]

wow, this was a really good answer and I really appreciate your effort to write this

I do have some question though. I always thought CHA-FAN and PWR-FAN are the same ports, with 1 exception, that you can control CHA-FAN and PWR-FAN works constantly at full power(?) That was my understanding of these 2 ports.

Well, Raijintek used to include the splitters, but with this new package it just wasn't included. They even deleted it from the user manual. you can see here: http://www.raijintek.com/en/products_detail.php?ProductID=18 in "UNBOXED" tab.

I have AsRock Z87 Extreme4 motherboard, so it would be expected that it does have all the required ports. (http://www.newegg.com/Product/Produ...69&cm_re=z87_extreme_4-_-13-157-369-_-Product)

That board has exactly two PWM ports that I can see: one CPU_FAN1, and the other is CHA_FAN1 at the bottom of the board. It is possible and even likely that you also have traditional BIOS or software-assisted thermal-fan-control on the 3-pin CHA_FAN2 and 3 ports. You'll have to find out. It is also possible that the mobo proprietary software could group the three CHA_FAN ports under a "Chassis fan" fan-curve.

But at bare minimum, you could do just fine with $20-worth of Swiftech PWM splitters and the two PWM ports. And obviously, the fewer fans and pumps in your cooling strategy, the bigger the possibility you might only want to use a single splitter or none at all.

You'll just have to find out the particulars, but you have plenty of options.

The only drawback to these Swiftech splitters: You can only monitor one of the connected devices (fans or pumps) among the eight possible connections -- for RPM. There's even a way get around that problem, with enough mobo fan ports and a limited number of items on the splitter.

There is also a 2-pin fan port labeled "CPU_FAN2" or similar. You'll have to investigate how that works. It may allow for only a static fan speed, or it may allow variable speed per the PWM signal of the first CPU_FAN port -- I couldn't say. But a decent fan strategy might allow for this or that fan to run at constant speed while the others are thermally controlled.

If you use a Swiftech splitter, just be sure that all the devices connected to it are PWM 4-pin devices -- pumps or fans. 3-pin devices will only run at their top-end if connected to the splitter.

Oh. just to explain again for the general membership, who know of my lengthy posts: I've got fast fingers. Sometimes it's just too easy . . . to "run off at the mouth . . . er . . . fingers, that is . . "

 

[Wort]

That board has exactly two PWM ports that I can see: one CPU_FAN1, and the other is CHA_FAN1 at the bottom of the board. It is possible and even likely that you also have traditional BIOS or software-assisted thermal-fan-control on the 3-pin CHA_FAN2 and 3 ports. You'll have to find out. It is also possible that the mobo proprietary software could group the three CHA_FAN ports under a "Chassis fan" fan-curve.

But at bare minimum, you could do just fine with $20-worth of Swiftech PWM splitters and the two PWM ports. And obviously, the fewer fans and pumps in your cooling strategy, the bigger the possibility you might only want to use a single splitter or none at all.

You'll just have to find out the particulars, but you have plenty of options.

The only drawback to these Swiftech splitters: You can only monitor one of the connected devices (fans or pumps) among the eight possible connections -- for RPM. There's even a way get around that problem, with enough mobo fan ports and a limited number of items on the splitter.

There is also a 2-pin fan port labeled "CPU_FAN2" or similar. You'll have to investigate how that works. It may allow for only a static fan speed, or it may allow variable speed per the PWM signal of the first CPU_FAN port -- I couldn't say. But a decent fan strategy might allow for this or that fan to run at constant speed while the others are thermally controlled.

If you use a Swiftech splitter, just be sure that all the devices connected to it are PWM 4-pin devices -- pumps or fans. 3-pin devices will only run at their top-end if connected to the splitter.

Oh. just to explain again for the general membership, who know of my lengthy posts: I've got fast fingers. Sometimes it's just too easy . . . to "run off at the mouth . . . er . . . fingers, that is . . "

I really appreciate your effort to help. I thought that all CPU_FAN ports and all CHA_FAN ports are the same and that there's no real difference between CHA_FAN1 and CHA_FAN2. I already plugged radiator fans to a controller which is plugged directly to PSU, so that is not my concern.

My concern is the pump. I know you tried to explain it already, but I'm not very familiar with this... I thought pump MUST run on full power. For example, when you plug it to PSU with molex-to-3pin connector, it runs with full power. So, what is the difference if it's plugged into motherboard on CHA_FAN1 or PWR_FAN, don't they all run on full power (12V)?

So what exactly is the difference if I plug water pump to CHA_FAN1, CHA_FAN2 or even PWR_FAN?

I'm really sorry for the inconveniences, but I think we got a misunderstanding here, since I asked where to plug my pump, not fans, as you mentioned

 

[BonzaiDuck]

I really appreciate your effort to help. I thought that all CPU_FAN ports and all CHA_FAN ports are the same and that there's no real difference between CHA_FAN1 and CHA_FAN2. I already plugged radiator fans to a controller which is plugged directly to PSU, so that is not my concern.

My concern is the pump. I know you tried to explain it already, but I'm not very familiar with this... I thought pump MUST run on full power. For example, when you plug it to PSU with molex-to-3pin connector, it runs with full power. So, what is the difference if it's plugged into motherboard on CHA_FAN1 or PWR_FAN, don't they all run on full power (12V)?

So what exactly is the difference if I plug water pump to CHA_FAN1, CHA_FAN2 or even PWR_FAN?

I'm really sorry for the inconveniences, but I think we got a misunderstanding here, since I asked where to plug my pump, not fans, as you mentioned

As far as I know from being a forum member and learning much from other members, pump-action is monitored as "RPM" like fans. It is possible to control pump-speed thermally, just as with the fans. Since pumps are less of a noise problem, you could run the pump(s) full bore without thermal control. But since motherboard needs to show a device connected to CPU_FAN so that computer will start, I'd connect either the radiator fans, the pump -- or all aspects of your AiO cooler -- to the CPU_FAN. I'd also investigate further how that CPU_FAN2 plug works with the main one.

Thermal control is most advantageous for "noise" control and lower power consumption. It may even allow fans/pumps to last longer. Moreover, lower RPMs for intake fans when the CPU is at idle and lowest temperature also means less of a dust-accumulation problem, and less of a need to filter the intakes.

It just all depends on the strategy you choose together with the constraints/limitations of ports and software. If fans and pumps are on the same splitter and PWM port, you'd have to balance the minimum pump speed you need against the minimum fan-speed, but that should be easy.

As far as I've known for either ASUS or Gigabyte boards -- possibly older EVGA models, the BIOS and proprietary software will -- at most -- allow for two different fan/pump profiles -- one for CPU and one for CHA. It might be more elaborate for a high-end, $400 motherboard, but someone else would have to confirm it. If you need or want a more refined set of control options, then you'd buy an add-on controller like the Aquaero. But that adds complexity to both your software configuration and your hardware. Personally, I'm more inclined to live with the simplicity and limitation of two fan/pump profiles.

As to the CPU and CHA profiles, again. . . Also, as far as I know, the controlling temperature is the CPU temperature and nothing else. I STILL find this completely adequate, but if you want to control case fans to the temperature of some other "hot" component like a graphics card, you'd have to use the bundled thermistors (sensors) that come with an add-on controller kit. Then, you'd have to calibrate them. Otherwise, decent or high-end graphics cards now provide their own fan-control features to work with software for "fan-curve-definition" -- like AfterBurner or EVGA Precision.

Some motherboards had been shipped with thermistor/sensor-wire accessories, and they provided ports for those sensors. My only experience with that goes back to an ASUS Striker Extreme 680i board. Usually, that feature is only provided for your top-end expensive $300+ motherboards.

 

[Aikouka]

Isn't the issue with trying to regulate an AIO's pump just that they (usually) aren't that powerful to begin with? I looked up that specific unit, and according to the KitGuru review, the pump is a 4W pump. That means it uses .33A, which isn't too high -- it's about 3x a 120mm fan. To give you an idea of the difference compared to custom loop pumps, a Laing DDC pump (EKWB variant) may use between 15-20W. The linked DDC pump pushes 350Lph where the Raijintek's pump pushes 120Lph -- nearly 1/3 of the EKWB.

 

[BonzaiDuck]

Isn't the issue with trying to regulate an AIO's pump just that they (usually) aren't that powerful to begin with? I looked up that specific unit, and according to the KitGuru review, the pump is a 4W pump. That means it uses .33A, which isn't too high -- it's about 3x a 120mm fan. To give you an idea of the difference compared to custom loop pumps, a Laing DDC pump (EKWB variant) may use between 15-20W. The linked DDC pump pushes 350Lph where the Raijintek's pump pushes 120Lph -- nearly 1/3 of the EKWB.

I knew someone would come through with a valid caveat.

Now that I think of it, that issue about AiO's came up some months ago -- not much earlier, anyway. I think it was a problem with a Corsair H100 or 110. I think you're only implying the best option, so I'll say what it is.

Under those conditions, it would be better to power the pump at full-bore and directly from the PSU.

It would then remain to be seen how using the three traditional pins exclusively for the pump on something like the 8W-PWM-SPL-SAT device. It was either written in specs, shown in a review or discussed on a forum: a three-pin device would run at full-bore on the Swiftech splitter, whether monitored or not. Best to look for evidence or lack of it that this in any way affects how any other device (with/without monitoring) would be controlled by the PWM signal-wire.

So in such a case as you (Aikouka) seem to imply, it would all work out in a shake without denying any other options -- like controlling the bundled AiO fans to the CPU temperature. And just by itself, it offers an opportunity to at least try beefier fans for the cooler with some good noise-level results.

Just my opinion . . .

Now that I think about it, a 3-pin fan wouldn't affect how the other fans behave under PWM control on the 8W-PWM-SPL. And I would also explore that 2-pin (looks like it anyway . . ) CPU_FAN2 or _FAN_OPT port on the OP's motherboard. But if it's a 2-pin connector, you wouldn't be able to monitor the pump.

And to be honest -- I don't know what sort of wiring the OP"s AiO pump has. But it's bound to accommodate at least one or more of those options -- whatever it is. With the sort of problem as Aikouka suggests, it may not have a PWM plug at all. The OP would know.

MORE: I just looked at whatever I could find out about the Triton at the Egg. It is clear that the manufacturer meant those devices to run directly off the PSU. For the fans, that's still good -- IF they're PWM fans and you can run them off the splitter; it's also no problem if you replace them with 4-pin PWMs that are beefier and do the same thing. But the implication is that you could run the pump directly from the PSU, or no less -- "directly from the PSU" through an 8W-PWM-SPL.

PS Too many people receive those Triton coolers with the box soaked because the assembly leaks.

 

[Wort]

I knew someone would come through with a valid caveat.

Now that I think of it, that issue about AiO's came up some months ago -- not much earlier, anyway. I think it was a problem with a Corsair H100 or 110. I think you're only implying the best option, so I'll say what it is.

Under those conditions, it would be better to power the pump at full-bore and directly from the PSU.

It would then remain to be seen how using the three traditional pins exclusively for the pump on something like the 8W-PWM-SPL-SAT device. It was either written in specs, shown in a review or discussed on a forum: a three-pin device would run at full-bore on the Swiftech splitter, whether monitored or not. Best to look for evidence or lack of it that this in any way affects how any other device (with/without monitoring) would be controlled by the PWM signal-wire.

So in such a case as you (Aikouka) seem to imply, it would all work out in a shake without denying any other options -- like controlling the bundled AiO fans to the CPU temperature. And just by itself, it offers an opportunity to at least try beefier fans for the cooler with some good noise-level results.

Just my opinion . . .

Now that I think about it, a 3-pin fan wouldn't affect how the other fans behave under PWM control on the 8W-PWM-SPL. And I would also explore that 2-pin (looks like it anyway . . ) CPU_FAN2 or _FAN_OPT port on the OP's motherboard. But if it's a 2-pin connector, you wouldn't be able to monitor the pump.

And to be honest -- I don't know what sort of wiring the OP"s AiO pump has. But it's bound to accommodate at least one or more of those options -- whatever it is. With the sort of problem as Aikouka suggests, it may not have a PWM plug at all. The OP would know.

MORE: I just looked at whatever I could find out about the Triton at the Egg. It is clear that the manufacturer meant those devices to run directly off the PSU. For the fans, that's still good -- IF they're PWM fans and you can run them off the splitter; it's also no problem if you replace them with 4-pin PWMs that are beefier and do the same thing. But the implication is that you could run the pump directly from the PSU, or no less -- "directly from the PSU" through an 8W-PWM-SPL.

PS Too many people receive those Triton coolers with the box soaked because the assembly leaks.

Mine leaked as well. It seems like water block usually cracks but this one was alright (thank god).

I see you had quite a debate with Aikouka. I did read everything you guys said, but I don't understand everything.

I want my pump to be running at its default "capacity" without getting "burnt" or broken from plugging it to the wrong header. I do realise that this AIO was meant to be ran directly from PSU, but they removed 3pin-to-molex connector from the Triton bundle, so I'm left with no 3pin connectors. I'm planning on buying one myself, but until then I want to safely plug water pump into whatever motherboard's header, without burning pump due to "too high current" (if that's even possible)

 

[Aikouka]

I'm planning on buying one myself, but until then I want to safely plug water pump into whatever motherboard's header, without burning pump due to "too high current" (if that's even possible)

It should be fine. If I remember correctly, the usual advice is to never hook more than 1 amp total to all fan headers. It uses .33A, so it will eat up 1/3 of your amperage budget, but most case fans only use ~.1A at max speed anyway, so you still have room for about six fans at max speed.

 

[BonzaiDuck]

It should be fine. If I remember correctly, the usual advice is to never hook more than 1 amp total to all fan headers. It uses .33A, so it will eat up 1/3 of your amperage budget, but most case fans only use ~.1A at max speed anyway, so you still have room for about six fans at max speed.

Actually, I think this deserves further "research" of motherboard user guides.

An ASUS board I bought in 2007 had been released in late 2006. It was a "flagship" board, with some five fan ports and pin-outs for thermal sensors as well. The user-guide had two alternative specs associated with fan amperage: one spec stated the "per-port" amperage, the other was a "cumulative" number. The "per-port" spec was 1 amp. I didn't find that level of detail in specs I'd seen for more recent boards.

On my Z68 and Z77 boards, I've used fan combinations that easily exceeded a cumulative 1 amp: for instance, an exhaust fan of maybe 0.7A, and two or three other fans of amperage ranging from 0.3 to 0.5. Without any problems, I would assume that the spec from the 2007 motherboard still applies generally to any number of more recent boards of different manufacture, but I couldn't be sure.

This was one good reason to deploy a device like the Swiftech 8W-PWM-SPL splitter, since all the power comes from the PSU directly. Having such a device handy provides an "out" at the first sign of any trouble. Frankly, I'd just as soon power all my fans from the PSU, if I could control them from the motherboard.

AFTERTHOUGHT: My latest example leaning toward a 1A/port spec is my Z68 sig-rig board. The (well-muffled) exhaust fan is a GT AP-30 with 0.56A and starting current of 1.35A. The CPU fan is an Akasa Viper 140R with amperage ~ 0.30+A. A 200mm NZXT fan is spec'd around 0.7A, and a second 200mm CM is about 0.30A. Of course, nothing is connected to the PWR_FAN port, but the total is 1.86A. This all worked fine without the Swiftech splitter -- which was added some time later.

 

[Aikouka]

Actually, I think this deserves further "research" of motherboard user guides.

I remembered that there was a 1A limit, but I couldn't recall whether it was split or combined. So, I did a bit of research, and the results showed a 1A total. However, those were also older discussions. So, I filtered out based on date, and most people are talking about each plug (usually) having a 1A limit.

 

[BonzaiDuck]

I remembered that there was a 1A limit, but I couldn't recall whether it was split or combined. So, I did a bit of research, and the results showed a 1A total. However, those were also older discussions. So, I filtered out based on date, and most people are talking about each plug (usually) having a 1A limit.

It certainly pays to get it right.

I had a ~0.90A Delta Tri-Blade on my CPU cooler with an old Springdale motherboard, and I think I had other fans on different ports that would push the total to 2A.

I just don't think the newer motherboards would be offered with a lower, more severe amperage limit. I'd even guess at this point, the expectations would be the same between boards of different manufacture, and between models varying in the number of ports.

But like I said -- that makes these $10 PWM splitters very attractive. They can be used in combination with fans powered from other fan ports. The only drawback is that you can only monitor a single fan/pump/device connected to the splitter. But you could also run the tach wires to the right pins on any free ports, if I'm not mistaken . . .