Production Modultra LOBO: Direct die kit gauging interest.

[stanleyguan]

I'm totally aware that a dual 240 setup is safe and tested, but i'm not into the thought of having a radiator dangling freely in there.
I'm also a lot into tinkering, and a "proof of concept" build like this was right down my alley, so i wanted to try it.

You can mount the side 240 to the bottom panel by a set of standoffs and L-brackets, like here.

 

[jztreso]

HI guys, I'm gauging interest in a solid metal low profile DDC pump block with optimized flow paths for intel 115x and AM4 socket. The bare block height would be 22mm, I believe this would be the lowest height DDC pump block available. The block would use a 5mm thick copper cold plate with 3d fin arrangement for pull though functionality. The block uses a 6 degree pump skew to help minimize flow restrictions inside the block. The Block would be made from either brass, or stainless steel.

The brass could be either nickel pated entirely or just nickel plated internally. Leaving a natural brass finish on the outside and nickel plated coolant channels.

Pros for raw brass with nickel plated coolant channels: It looks awesome when polished. Kind of like a gold bar sitting on your CPU!

Cons for raw brass: Brass will tarnish over time; luster can be restored with hand polishing. Also more expensive as machining would be required after plating. Not scratch resistant.



Pros for nickel plated brass: shiny, corrosion and scratch resistant.

Cons for nickel plated brass: kind of looks like everything else, shiny silver. Finish can corrode over time.



Cons for stainless steel: Expensive, machining stainless has a high attrition rate on tooling. Polished stainless is not highly scratch resistant.

Pros for stainless: Little possibility of corrosion ever. Raw stainless can have Shiny or matte finish. A multitude of secondary finishes can be applied. Stone wash, vapor deposition coatings, paint, Cerakote, powder coat, electro plating. Lifetime warranty.

Will we be able to mount the block vertically on am4 socket, like the barrow ddc pump block?
This would definitly be the gamechanger for me, since the horisontal mount (like aqanaut) is really limiting in terms of both flow path, but also ram clearance.

 

[Alloy Craft]

Will we be able to mount the block vertically on am4 socket, like the barrow ddc pump block?
This would definitly be the gamechanger for me, since the horisontal mount (like aqanaut) is really limiting in terms of both flow path, but also ram clearance.

Unfortunately he AM4 and AM3 mounting holes wont allow the block to mount in both orientations.

 

[lulle2007200]

So I'm working on a way to sweeten the pre order deal. I'm thinking about doing the LOBO metal pump and heatsink combo for $180 shipped in the USA. Deal will include all hardware and thermal interface pad. Colors would be carbon grey and titanium. Sockets would be LGA 115x-1200-1700 or AMD AM4. I will post some more info soon, still working out some of the details.

Sounds like a great deal, if it's going to happen.
I'm very interested in the LOBO, sign me up for pre-order ^^

 

[lulle2007200]

So I'm working on a way to sweeten the pre order deal. I'm thinking about doing the LOBO metal pump and heatsink combo for $180 shipped in the USA. Deal will include all hardware and thermal interface pad. Colors would be carbon grey and titanium. Sockets would be LGA 115x-1200-1700 or AMD AM4. I will post some more info soon, still working out some of the details.

Interested in pre-ordering aswell, sign me up for that deal if it's going to happen.

 

[Nautilus]

This looks pretty awesome, can't wait to (pre)order one.

By the way, did any of the testers happen to test if it fits on an ASRock x570 ITX phantom gaming mainboard? For some reason they decided use Intel 1151 instead of AMD AM4 style coolers. I have a feeling the io/MOSFET heatsinks might be a bit obstructive.

 

[morj]

This looks pretty awesome, can't wait to (pre)order one.

By the way, did any of the testers happen to test if it fits on an ASRock x570 ITX phantom gaming mainboard? For some reason they decided use Intel 1151 instead of AMD AM4 style coolers. I have a feeling the io/MOSFET heatsinks might be a bit obstructive.

Nothing a bit of dremeling wouldn't solve, I assume

 

[Nautilus]

I guess... as a last resort. But I'd rather not

 

[Valantar]

This looks pretty awesome, can't wait to (pre)order one.

By the way, did any of the testers happen to test if it fits on an ASRock x570 ITX phantom gaming mainboard? For some reason they decided use Intel 1151 instead of AMD AM4 style coolers. I have a feeling the io/MOSFET heatsinks might be a bit obstructive.

From the looks of it the Intel socket mounting plate adheres to Intel's Keep-Out Zone restrictions, which that motherboard also does. The VRM heatsink might interfere, but it shouldn't if both are designed correctly. It will definitely limit your available mounting orientations though - there's no way you'd be able to have the ports oriented towards the rear I/O. But either upwards or downwards should be fine, and you might even get away with having them pointed towards the RAM slots.

 

[SFFMunkee]

So I’ve got my beta unit installed with an EK DDC 4.2.

I’ve got a few issues at the moment:
1) it’s rather loud - even when turned down - compared to either the DC-LT2 from the EisbaerLT solo or the D5 I had before,
2) the PWM range is really limited for some reason, anything over about 40% is 4000+ rpm, so I had to make separate control curves for pump vs fan which is fine.

I’m wondering how low I can go with the pump and not have significant performance loss, given it’s an 18W pump I’m sure I could run it at quarter or half speed and it should be fine, right? (I’ll test it anyway)

 

[Vareen]

So I’ve got my beta unit installed with an EK DDC 4.2.

I’ve got a few issues at the moment:
1) it’s rather loud - even when turned down - compared to either the DC-LT2 from the EisbaerLT solo or the D5 I had before,
2) the PWM range is really limited for some reason, anything over about 40% is 4000+ rpm, so I had to make separate control curves for pump vs fan which is fine.

I’m wondering how low I can go with the pump and not have significant performance loss, given it’s an 18W pump I’m sure I could run it at quarter or half speed and it should be fine, right? (I’ll test it anyway)

Is it loud because the motor is naturally loud anyway, or is it perhaps resonating with the block?
Just trying to figure out if this is a problem with the block (sadface), or just with your chosen pump (sadface for you, but not an issue for the creator).

 

[Valantar]

So I’ve got my beta unit installed with an EK DDC 4.2.

I’ve got a few issues at the moment:
1) it’s rather loud - even when turned down - compared to either the DC-LT2 from the EisbaerLT solo or the D5 I had before,
2) the PWM range is really limited for some reason, anything over about 40% is 4000+ rpm, so I had to make separate control curves for pump vs fan which is fine.

I’m wondering how low I can go with the pump and not have significant performance loss, given it’s an 18W pump I’m sure I could run it at quarter or half speed and it should be fine, right? (I’ll test it anyway)

Mine is a Laing DDC, and it's clearly not the same, but it shares some characteristics: 100% speed is around 60-65% PWM, among other things. I never run it even close to those speeds - at idle I run it at 20%/1200rpm, which is perfectly adequate for desktop use, and under load it ramps up to a peak of 45%/3400rpm, at which point it provides plenty of flow and is still essentially inaudible.

It's worth noting that noise does not increase linearly with pump speed with my DDC, and I had to test a bunch of speeds to find the most quiet ones, then tune its curve to hit those spots and not just ramp willy-nilly. For example, 2800-3000rpm are far noisier than 3400; 2000-2200 is pretty bad as well, while the mid-2000s are okay - but 3400 is no louder, and performs better.


Edit: it's also worth noting that with a noise source with a bit of mass, like a water pump, resonances are your worst enemy, and that's why running through the frequency spectrum to find the least resonant areas is worth doing. What resonates and how will vary depending not only on the pump and how it vibrates, but on pretty much all the hardware in your build, from the water block/pump mount to the case, AICs, motherboard, other components, the tightness of screws, production tolerances for various components, etc.

 

[stanleyguan]

So I’ve got my beta unit installed with an EK DDC 4.2.

I’ve got a few issues at the moment:
1) it’s rather loud - even when turned down - compared to either the DC-LT2 from the EisbaerLT solo or the D5 I had before,
2) the PWM range is really limited for some reason, anything over about 40% is 4000+ rpm, so I had to make separate control curves for pump vs fan which is fine.

I’m wondering how low I can go with the pump and not have significant performance loss, given it’s an 18W pump I’m sure I could run it at quarter or half speed and it should be fine, right? (I’ll test it anyway)

DDC pumps might not be a good choice if you are really after silence. I have EK DDC 4.2 as well and it's way louder than my EK SPC-60 pump, regardless of how low I run EK DDC 4.2 at. See also last plot of this review. With EK SPC-60 (similar to D5 in noise level according to that review) at 70%, I cannot hear the pump 40cm away, in a VERY quiet room. With EK DDC 4.2, I can CLEARLY hear it even when running it at the lowest speed. There was a buzzing sound to it regardless of the speed.

If anyone can suggest a DDC that is extremely quiet and PWM-controlled I'd love to try it out.

 

[SFFMunkee]

Is it loud because the motor is naturally loud anyway, or is it perhaps resonating with the block?
Just trying to figure out if this is a problem with the block (sadface), or just with your chosen pump (sadface for you, but not an issue for the creator).

I doubt it's a problem with the block, DDCs are by all accounts relatively loud pumps.
It could be quieter if there were a way to isolate the pump frame a bit more, but that's hardly feasible when you need to bolt the pump TO to the block!

Mine is a Laing DDC, and it's clearly not the same, but it shares some characteristics: 100% speed is around 60-65% PWM, among other things. I never run it even close to those speeds - at idle I run it at 20%/1200rpm, which is perfectly adequate for desktop use, and under load it ramps up to a peak of 45%/3400rpm, at which point it provides plenty of flow and is still essentially inaudible.

It's worth noting that noise does not increase linearly with pump speed with my DDC, and I had to test a bunch of speeds to find the most quiet ones, then tune its curve to hit those spots and not just ramp willy-nilly. For example, 2800-3000rpm are far noisier than 3400; 2000-2200 is pretty bad as well, while the mid-2000s are okay - but 3400 is no louder, and performs better.


Edit: it's also worth noting that with a noise source with a bit of mass, like a water pump, resonances are your worst enemy, and that's why running through the frequency spectrum to find the least resonant areas is worth doing. What resonates and how will vary depending not only on the pump and how it vibrates, but on pretty much all the hardware in your build, from the water block/pump mount to the case, AICs, motherboard, other components, the tightness of screws, production tolerances for various components, etc.

(Please correct me if I'm wrong, but) I think all DDCs are made by Laing (AKA Swiftech for international units back in the day) but usually with the branded top/frame/heatsink/etc.
Although there may be DDC-'compatible' units from other brands that have are similar to the DDC design (copied? IDK) these aren't actually DDCs I believe.

That said, there are definitely different models, I've seen mention of: DDC310, DDC-1T, DDC-1T Plus, DDC 3.2, DDC 3.25, DDC 4.2, DDC-1RT

DDC pumps might not be a good choice if you are really after silence. I have EK DDC 4.2 as well and it's way louder than my EK SPC-60 pump, regardless of how low I run EK DDC 4.2 at. See also last plot of this review. With EK SPC-60 (similar to D5 in noise level according to that review) at 70%, I cannot hear the pump 40cm away, in a VERY quiet room. With EK DDC 4.2, I can CLEARLY hear it even when running it at the lowest speed. There was a buzzing sound to it regardless of the speed.

If anyone can suggest a DDC that is extremely quiet and PWM-controlled I'd love to try it out.

Agreed, it looks like a lot of them are quite loud, but I wonder if any of the other models might be quieter, even if they don't offer the 4-7m head pressure of this one.

I assume the main reason the D5 is quieter is because they're ALWAYS isolated by mounting with a silicone/rubber-lined clamp. The pump itself has no mounting points!
The DDC, on the other hand, has screw holes in the pump body which means it can be bolted directly to a block/frame/etc so it's not necessarily going to be isolated as well.

I'm tempted to just go 100% external, have a D5 and my 420mm radiator mounted under the desk, have a teeny tiny PC on the top... but then I don't get to use my sexy AlloyCraft block

If anyone's interested I actually tested out the PWM curves here:

I noticed that my pump and fans have wildly different responses to PWM input signals, and thought that I'd test it out, have only checked the devices I have attached right now but here's the data anyway!

PWM input control: FanControl (https://github.com/Rem0o/FanControl.Releases)
RPM output view: HWInfo64 (https://www.hwinfo.com/download/)
Motherboard: ASRock B550 Phantom Gaming ITX/ax (w/ Ryzen 5600x + 2x8GB DDR4-4400 Patriot Vipers)
Fan(s): EK-Furious Vardar EVO 140 (6x fans connected to one header with PWM splitter, mounted in push/pull on a 420mm Alphacool Monsta radiator)
Pump: EK DDC 4.2 PWM (attached to AlloyCraft 'lobo' beta waterblock, powered via SATA from PSU, PWM from motherboard pump/fan header)

Method: All I did was manually set the PWM input in FanControl for each fan/pump, noted the highest/lowest speed output reading from HWInfo over about 30-60 seconds, and graphed the results.

EK Furious Vardar EVO 140
EK 140mm fans have this curved response, with each 5% increase from about 15% to 100% causing a noticeable change in speed (rpm).
Anything below 15% is approximately equivalent ~960rpm. No zero-rpm for these fans.

In terms of noise output, they're:

• effectively silent from 0% -30% (up to ~600rpm)
• barely audible from 30% - 45% (up to ~830rpm)
• quiet but clearly audible from 45% - 55% (up to ~1000rpm)
• becoming irritating from 60% - 70% (up to ~ 1100rpm)
• definitely loud from 70% (~1300rpm all the way to full speed ~2400rpm)

As there were no further changes to RPM output below about 12% I have truncated the graph below 10%.

EK DDC 4.2 PWM pump
The EK DDC 4.2 PWM pump has a very different curve, it's almost linear but is heavily truncated to only ~12% to ~42% PWM.
The same speed (950-1000rpm) is produced from anything below ~13%
Likewise full speed (~4440rpm) is produced from 44% all the way up to 100%.

In terms of noise output, it is:

• effectively silent from 0% - 19% (~960rpm to ~1600rpm)
• barely audible from 20% - 28% (~1650rpm to ~2660rpm)
• quiet but clearly audible from 29% - 32% (~2750rpm to ~3130rpm)
• becoming irritating from 33% - 35% (~3240rpm to ~ 3500rpm)
• definitely loud from 36% (~3550rpm all the way to full speed ~4440rpm)

Weird results, I reckon, do other pumps behave similarly?

::EDIT:: In case it helps anyone out, these are the curves/sensors I have setup in FanControl

Fans are controlled with:

1. Default CPU/GPU temperature sensors.
2. CPU/GPU curves for fan pwm settings based on temperature, with a 3C hysteresis and 6 sec response time.
3. Mix curve, combining the two curves to pass whichever is highest to the fan PWM.

Pump is controlled with:

1. Custom sensors for each CPU/GPU giving the average (30sec) temperature.
2. CPU/GPU curves for pump pwm settings based on temperature, with 5C hysteresis and 10 sec response time.
3. Mix curve, combining the two curves to pass whichever is highest to the pump's PWM.

 

[Valantar]

(Please correct me if I'm wrong, but) I think all DDCs are made by Laing (AKA Swiftech for international units back in the day) but usually with the branded top/frame/heatsink/etc.
Although there may be DDC-'compatible' units from other brands that have are similar to the DDC design (copied? IDK) these aren't actually DDCs I believe.

That said, there are definitely different models, I've seen mention of: DDC310, DDC-1T, DDC-1T Plus, DDC 3.2, DDC 3.25, DDC 4.2, DDC-1RT

I also think they're all made by Laing, but different brands spec them differently in various ways. Mine is a 1T Plus, which AFAIK is the same as the DDC 3.2 PWM - at least the names seem to be used interchangeably at times. There are also a bunch of clones, though as you say those aren't actually DDCs (the pumps Barrow bundles with their CPU pump mount are a good example of this), and some, like EK's VTX pumps, aren't actually DDC compatible at all even if the designs and form factors are similar.

I would still recommend that you go through your pump's range and identify the least intrusive speeds, and then tweak your pump curve to keep to these levels rather than ramp smoothly. In my case, that made the pump go from 'kind of annoying in most scenarios' to 'I never even notice it'. This requires a bit more setup, but other than convention there's no reason why the pump needs a smooth temperature-controlled curve. Mine sticks to 1200rpm until fluid temperatures exceed a certain threshold or system load exceeds another threshold, at which point it ramps to 3400rpm and stays there until things cool down again.

 

[SFFMunkee]

I also think they're all made by Laing, but different brands spec them differently in various ways. Mine is a 1T Plus, which AFAIK is the same as the DDC 3.2 PWM - at least the names seem to be used interchangeably at times. There are also a bunch of clones, though as you say those aren't actually DDCs (the pumps Barrow bundles with their CPU pump mount are a good example of this), and some, like EK's VTX pumps, aren't actually DDC compatible at all even if the designs and form factors are similar.

I would still recommend that you go through your pump's range and identify the least intrusive speeds, and then tweak your pump curve to keep to these levels rather than ramp smoothly. In my case, that made the pump go from 'kind of annoying in most scenarios' to 'I never even notice it'. This requires a bit more setup, but other than convention there's no reason why the pump needs a smooth temperature-controlled curve. Mine sticks to 1200rpm until fluid temperatures exceed a certain threshold or system load exceeds another threshold, at which point it ramps to 3400rpm and stays there until things cool down again.

Good plan, when I was checking the PWM response curve I actually noted down which speeds/frequencies caused harmonics/vibration through the case at each step, so I'll see if I can set FanControl curve to avoid the annoying ones, maybe focusing on 2 or 3 steps where the NVH isn't too bad!

Sorry for thread clutter everyone, but hopefully this is useful to the next person who thinks their Lobo is loud when really it's the DDC pump needs a bit of tuning

 

[Alloy Craft]

Good plan, when I was checking the PWM response curve I actually noted down which speeds/frequencies caused harmonics/vibration through the case at each step, so I'll see if I can set FanControl curve to avoid the annoying ones, maybe focusing on 2 or 3 steps where the NVH isn't too bad!

Sorry for thread clutter everyone, but hopefully this is useful to the next person who thinks their Lobo is loud when really it's the DDC pump needs a bit of tuning

Munkee, do you have a copy of the install manual? I think you got the LOBO second hand so I don't know if the original tester forwarded it to you. Anyway if the EK pump you are using is a spherical rotor type, it is possible that the rotor is hitting the frame at certain frequencies. The spherical rotor can spin off axis due to balance defects. The pump will work fine but it means that it will have higher off axis tendencies. The pump can be shifted on the body a little bit, you might try shifting the pump away from the water cut slightly to see if this helps with noise.

 

[ziggehh]

Hi,

Been following the project (somewhat) from the early days! I was wondering if it would be possible to get my hands on a LOBO - I am about to commence on a 5900x/3080 build in a Jimu D+ v2 and was really hoping to use the LOBO as my CPU/pump block. Happy to post testing numbers/provide feedback if I can purchase a unit.

 

[Alloy Craft]

Hi,

Been following the project (somewhat) from the early days! I was wondering if it would be possible to get my hands on a LOBO - I am about to commence on a 5900x/3080 build in a Jimu D+ v2 and was really hoping to use the LOBO as my CPU/pump block. Happy to post testing numbers/provide feedback if I can purchase a unit.

Hopefully I can get some done soon, everything is taking longer than expected. However I am working on the heatsinks today. here is a photo of the 4 available pattern options. After I get these painted I most likely will start a pre order.

 

[SFFMunkee]

Munkee, do you have a copy of the install manual? I think you got the LOBO second hand so I don't know if the original tester forwarded it to you. Anyway if the EK pump you are using is a spherical rotor type, it is possible that the rotor is hitting the frame at certain frequencies. The spherical rotor can spin off axis due to balance defects. The pump will work fine but it means that it will have higher off axis tendencies. The pump can be shifted on the body a little bit, you might try shifting the pump away from the water cut slightly to see if this helps with noise.

Didn’t have the manual, I don’t think. I don’t remember seeing one anyway. Do you have a PDF/digital version I can refer to? If I could make it quieter I’d be over the moon!