All MCP 355/DDC+ (and even DDC 3.2) Pump Owners Please see this !!!!! (With Pics)

[Jedda]

There seems to be a tendency to pull out the big screw-driver and spanner with WC in general. Mostly driven by fear of leaks, I guess.
We could do with some quality six inch torque wrenches. (semi-jesting)
The big spanner causes more trouble than it cures.

[C'DaleRider]

I thought the DDC dies when you run it with no water? Or was your DDC already with death symptoms? If not you may have killed it...

BTW, for inspecting the change in height in the impeller, I suggest using an acrylic top.

No, it was not a dying DDC-2, rather a less-than-week old brand new pump, Danger Den branded. I did lubricate the ceramic bearing with a drop or two of straight Pentosyn G11, but that's all I did. I did that to reduce the possibility of drag-induced movement of the impeller's turning.....the Pentosyn is great as a lubricant in this case. Straight, it made the ceramic bearing quite slick.

Interestingly, throughout all the running, the pump never became hotter than normal to touch on the bottom or inside. But I do have the pump elevated approx. 1/4" from sitting flat.....I'm using 3/4" x 3/4" x 1/4" self-stick closed cell foam blocks that are placed at each corner, covering the screws for the top. The rest of the bottom of the pump is exposed to air.....that's the way it sits in my case with a bit of air flow over it from my front case fan.

I actually have four blocks on the bottom and two on one side of the pump with patches of the fuzzy part of Velcro attached to those two. The side blocks mate to similarly fashioned foam blocks with the hook part of Velcro stuck on them where the pump "fixes" to the case. So the pump is isolated by foam blocks all around and has worked very well in damping any pump noise transferred to the case. I'd tried rubber blocks but they were too hard and solid and did almost nothing to dampen the noise.

I also wonder how many failures are coming from people using things like Petra's gel block or have lined the entire bottom with the neoprene pad. Seems to me that because the motor does generate heat while operating, lining the entire bottom of the pump is a good way to ensuring premature heat failure.

Originally, I was using a DDC pump and had tried the gel block form Petra. After using it a few days, and it did work in isolating noise, I lifted the pump and was very surprised by how hot the bottom of the pump had gotten and how hot the gel pad was. That's when I changed to the foam block setup I'm using now.....does let a bit of air circulate underneath the pump. Figured couldn't hurt and had to help at least a little.

BTW....the pump is back in service, working as always, quiet as ever. Whether this little bit of testing has shortened its life, time will tell. But as of now, it appears to not be a problem.....but I tend to run my computer 24/7.

[Anemone]

CD, think I'm going to adopt that method. The closed cell foam was it any particular foam, or just packing foam? I'm assuming it wasn't cut up neoprene squares, but curious even still.

thank you

[MotF Bane]

I also wonder how many failures are coming from people using things like Petra's gel block or have lined the entire bottom with the neoprene pad. Seems to me that because the motor does generate heat while operating, lining the entire bottom of the pump is a good way to ensuring premature heat failure.

Originally, I was using a DDC pump and had tried the gel block form Petra. After using it a few days, and it did work in isolating noise, I lifted the pump and was very surprised by how hot the bottom of the pump had gotten and how hot the gel pad was. That's when I changed to the foam block setup I'm using now.....does let a bit of air circulate underneath the pump. Figured couldn't hurt and had to help at least a little.

I'm still convinced heat killed my first one after about 10 minutes, it was too hot to even touch the plastic shell. Stuck it on some sticky foam all over the bottom. My friend's is warm to the touch, sitting on Petra's gel stuff. My new one is cool to the touch, and it is currently suspended midair with fans behind it and the radiator fans ahead of it.

[Sparky]

Mine is mounted to the bottom of my case by the two mounting tabs. I have a bolt attached to the case with a rubber grommet, a nut on the bolt to hold it in place, then the pump has the rubber grommets it came with attached to the ears and another nut on that. Then on the opposite corners of the ears I have some soft rubber bumpers attached to stabilize the pump. A good 1/4" or so of airspace under the pump and it runs cool.

[C'DaleRider]

CD, think I'm going to adopt that method. The closed cell foam was it any particular foam, or just packing foam? I'm assuming it wasn't cut up neoprene squares, but curious even still.

thank you

Actually, they were the foam feet removed from a Corsair Nautilus 500 case. Quite dense foam, but I'd imagine that dense closed cell foam that you'd use around doors would approximate it pretty well.

[Xeon th MG Pony]

Wait a minute. You say that Petra's top is just as deeper as the MCP355 stock top ? Which would mean that neither Petra or Laing knew that the impeller would rise a lil more than an MCP350 since it's spinning faster ?

That would also means that running the stock MCP355 top would cause this problem ? Not enough space between the impeller and the top ?

I seriously doubt that this could have been done. I mean, these guys are professionals. Things have probably been done with this problem in mind, no ? You know what i mean ?

Seriously, i would like to have someone with more knowledge or someone from Swiftech\Laing to see this thread and give us answers about that.

EDIT: BTW : Recently before i build my latest loop, i had that exact problem once. Someday that i was idle, temps went to 83c exactly. So, i decided to restart the comp. So, after a clean reboot, i had a look into the bios and temps were back to normal. I booted into windows, and temps were still fine. I thought maybe it was just a bug about Asus Probe... But then, you make me think that it might have been the pump!

A 50 year experienced engineer blew out a quarter of a million dollar chiller simply because the manufacture gave him the wrong data and a part was engineered wrong, moral: Just because their a big company does not mitigate them from screwing up. NASA Blew up a rocket due to mistranslating a single digit from metric to imperial. So why do you find it so hard to believe they could have screwed up a simple design by the point they may have neglected the slight change in the impellers hydrodynamic film that it rides on?

[Xeon th MG Pony]

^^ Well, this theory is better than the others atleast.

After reading it, I do kind of agree, with the first post. It could just be an oversight from Laing. This is the only explanation I agree with.

Maybe with the 3.2 Laing reduced the size of the Impeller? Or put something to stop it from rising above the clearance space?

Looking at the pic of the top notice the slight grooves cut into the top from some thing landing on it?

[Xeon th MG Pony]

Well, curiosity got the best of me so I snuck into work (Torrington Bearings plant) and used the optical comparator we use in the machine shop. Right now, it was only set up with the 10X power lens. We use Dorsey Metrology comparators.....this one is the "quickie" check unit....a smaller 24" unit.

Our others are 36" and larger.....

Anyway, I carried with me a 5A portable power supply AC to DC converter power supply. Hooked it up to the pump. Focused in the comparator on the impeller and set the reference line. As the comparator was set, it was in a 0.001" resolution, but I think that's close enough to get to 0.1mm that was mentioned.

Even under this magnification and measuring scale, still didn't see any lifting of the impeller.....not one 0.001" of lift at all. Repeated the turn off/on sequence over a dozen times and never saw lifting.

I did see, on the other hand, very minor imbalance and wobble within the impeller's spin, most noticeable at low rpm's, or when it was starting its spinup. Once the unit achieved full speed, the wobble from imbalance disappeared completely, and this may be what is the real problem.....a balance problem from manufacturing intolerances. Quite possibly, the tolerances of manufacturing are not as tight as should be for the impeller, being it is a cast piece of plastic or resin, and very possibly as molds age and change shape minutely, QC fails to catch the out-of-round impellers, and this causes those impellers to wobble more than they should on spinup leading to them jamming against the sidewalls of the top.

Just a theory.

Problem is you don't have a fluid film that it will normaly ride on with your tests, to get a usable result you'd need a clear top and actualy be pumping water, how ever the "out of round" issue certainly will not help with things!

[nikhsub1]

Hmm, a couple of things would discredit this theory, although I suppose it has some merrit. FWIW I have 2 DDC 18w with alphacool tops and not a single issue. The reason why I don't think your theory is correct is that if it were, this would happen from the get go and not over time. Furthermore, pumps that have been 'stuck' or had a startup problem were found to run just FINE when sent in for RMA to DD, Swiftech and Petra. I am still thinking this is PSU/electrical related in some way.

[jayhall0315]

I'm still convinced heat killed my first one after about 10 minutes, it was too hot to even touch the plastic shell.

Please remember guys that if the wire looped around the stator base to form the electromagnet is not laid out precisely (especially considering the power for such a small form factor), then when the impeller 'wobbles', it WILL contact the ceramic bearing with more friction and transfer much more heat, leading perhaps to some of the premature failures that are being reported.


Jay

[jayhall0315]

Hmm, a couple of things would discredit this theory, although I suppose it has some merrit. FWIW I have 2 DDC 18w with alphacool tops and not a single issue. The reason why I don't think your theory is correct is that if it were, this would happen from the get go and not over time. Furthermore, pumps that have been 'stuck' or had a startup problem were found to run just FINE when sent in for RMA to DD, Swiftech and Petra. I am still thinking this is PSU/electrical related in some way.

Hey Nikhsub,
Good to have your input on this problem. So far though, with respect, I have to disagree. For a quality PSU, the ripple is just not usually enough to cause these types of problems.
As for the nature of the problems, please see point #7 in the original post. According to what I can gather, for some MCP 355 owners, this is an IMMEDIATE problem, for others it is time delayed and for others, it never arises, and for still others it is a cyclical issue. This extreme variability that has made it so difficult for the people back at Swiftech, or Liang or Petra, to say, "Ahh, this is the exact problem", is because there is a small degree of variability in how the wire was looped for the electromagnet at the Liang factory (which is not as critical for the lower amperage DDC but IS critical for the higher amperage MCP 355), and in combination with how tightly each owner has cranked down their top, makes these two variables which are actually coupled, hard to separate.

Jay

[Pete]

Jay i got a DDC2 here, PC+P 510, Alphacool and OClab plexi top. Got a decent camera and tripod. I can video it if it was to help.

I can do it if it will help at all!

[kiikkuja]

@ Jayhall0315

It's Laing not Liang.

The misspelled word bothered me so much i had to post...I respect your efforts greatly on this subject!! No disrespect meant!!

[jayhall0315]

Jay i got a DDC2 here, PC+P 510, Alphacool and OClab plexi top. Got a decent camera and tripod. I can video it if it was to help.

I can do it if it will help at all!

Pete, you are welcome to try this if you believe it will help. However, even with slowed down footage, I believe you will have a very hard time discerning a possible 0.1 mm wobble. Without some more precise equipment (like an infared laser - which I may get around in the next few days) then we are going to need to wait for Laing's or Swiftech's troubleshooters to get involved.

Jay

[jayhall0315]

@ Jayhall0315

It's Laing not Liang.

The misspelled word bothered me so much i had to post...I respect your efforts greatly on this subject!! No disrespect meant!!

Duhh !!!

Thanks, I got it. So, it is Liang, not Laing, correct ?

Jay

(okay, okay, just joking)

[Pete]

Pete, you are welcome to try this if you believe it will help. However, even with slowed down footage, I believe you will have a very hard time discerning a possible 0.1 mm wobble. Without some more precise equipment (like an infared laser - which I may get around in the next few days) then we are going to need to wait for Laing's or Swiftech's troubleshooters to get involved.

Jay

Good luck with hearing off them. Gabe just hurls legal abuse i found!

Laing are boring and told me to stop with my investigations to the model info i was sorting. Kinda sad really but that the US/EU divide i guess doing it's thing!

HD camera!!! Well providing i can find the battery for it! It will be very clear with the macro lens and i can slow it down as much as you like. I'll try and sort it out

[Petra]

Wow, I take a weekend off and look at what happens...

Well, I have a lot of catching up to do today but, that said... the stock top's impeller housing ceiling is somewhat conical and, AFAIK, all of the custom tops are not. According to my Solidworks models, the DDCT-01 and DDCT-01s are exactly 8mm deep (samples from our last run came out to be about 8.02-8.03mm deep, according to my calipers).

Here's a cutaway comparison of the DDCT-01 impeller housing ceiling and the stock DDC housing (due to the inlet lips, you can tell that the stock top is from a DDC-1/DDC-2 and the custom top (in red) is one of my earlier ones).



Now, with the DDC-2's being discontinued, and us needing to run another batch of tops very, very soon... it wouldn't be difficult to make a quick modification; however, I'm not certain what sort of a performance change would result. Granted, I have been considering taking an extra 0.25mm-0.5mm out of the center (in a 18mm dia. circle), but that's something that I was planning on tinkering with for the DDCT-02 design.

One last tidbit to toss out... Many of the DDC-2's that came back as either dead/defective or as having starting issues had malformed/warped impellers (in other words, there were high spots along the outer edge of the impeller). I wouldn't really write that off as heat related, though, because I found several that arrived in said condition from Swiftech/Laing. These warped impellers usually became more of an issue in operating conditions that caused increased impeller deflection (since the DDC's impeller seems to tilt depending on the operating pressure, thus, reducing clearance on one side).

[Pete]

The impler is weighted, twist it apart you will see the ball bearings which are the weights Alex!

Also the high and rised bits are due to bad mass moulding!

[virtualrain]

Good thread Jay... It's good to see some analysis of why these pumps are failing occasionally.

IMHO, based on years of experience with solving problems with nearly anything... is that issues that prove difficult to replicate are most likely a combination of things and NOT a single simple oversight... slight differences in manufacturing tolerances, coolant viscosity variations, the cooling provided to the pump, PSU behavior at system startup, and all the other things you've touched on, are likely contributors to a greater or lesser extent (even from one setup to the next) in determinging if a pump will fail.

However, if there are certain things we can do to ensure our pumps live a long happy life, that would be great.

For example, I think this thread has already provided several good tips to DDC owners...

DDC TIPS (from this thread):

- Where possible, orient your pump horizontal (right side up).
- Use a lubricant in your coolant (Pentosin or something similar).
- Do not sit your pump fully on a gel pad... allow air to cool the underside.
- Do not over-tighten the top... use only enough clamping force to prevent a leak... no more.
- Inspect your pump impeller for warpage before you use it
- Clean your loop thoroughly before running your pump in it to ensure there are no damaging particles in your loop

[Maviryk]

Nice pictures Petra! BTW my services are still available if required.

[migueld]

Originally, I was using a DDC pump and had tried the gel block form Petra. After using it a few days, and it did work in isolating noise, I lifted the pump and was very surprised by how hot the bottom of the pump had gotten and how hot the gel pad was. That's when I changed to the foam block setup I'm using now.....does let a bit of air circulate underneath the pump. Figured couldn't hurt and had to help at least a little.

That's definitely important, I've had my DDC ran very hot (and it's a 9w) whenever there's poor ventilation, and I can see what you are saying about the gel pad. I haven't seen that in surveys though, maybe it's being overlooked? In the recent DDC survey there was an item for DDC positioning but not ventilation. Maybe this should be asked as well.

[jayhall0315]

Wow, I take a weekend off and look at what happens...

Well, I have a lot of catching up to do today but, that said... the stock top's impeller housing ceiling is somewhat conical and, AFAIK, all of the custom tops are not. According to my Solidworks models, the DDCT-01 and DDCT-01s are exactly 8mm deep (samples from our last run came out to be about 8.02-8.03mm deep, according to my calipers).

Here's a cutaway comparison of the DDCT-01 impeller housing ceiling and the stock DDC housing (due to the inlet lips, you can tell that the stock top is from a DDC-1/DDC-2 and the custom top (in red) is one of my earlier ones).



Now, with the DDC-2's being discontinued, and us needing to run another batch of tops very, very soon... it wouldn't be difficult to make a quick modification; however, I'm not certain what sort of a performance change would result. Granted, I have been considering taking an extra 0.25mm-0.5mm out of the center (in a 18mm dia. circle), but that's something that I was planning on tinkering with for the DDCT-02 design.

One last tidbit to toss out... Many of the DDC-2's that came back as either dead/defective or as having starting issues had malformed/warped impellers (in other words, there were high spots along the outer edge of the impeller). I wouldn't really write that off as heat related, though, because I found several that arrived in said condition from Swiftech/Laing. These warped impellers usually became more of an issue in operating conditions that caused increased impeller deflection (since the DDC's impeller seems to tilt depending on the operating pressure, thus, reducing clearance on one side).

Hi Alex,
Yes, as soon as you have time, I would welcome your investigation of these findings so far. Also, if you can, please pass this along to the folks at Laing that actually design these units (and not some flunkies in the PR dept) for their investigation as well. I am conducting testing as we speak and will have full details later today.

Jay

[Senater_Cache]

Hi Alex,
Yes, as soon as you have time, I would welcome your investigation of these findings so far. Also, if you can, please pass this along to the folks at Laing that actually design these units (and not some flunkies in the PR dept) for their investigation as well. I am conducting testing as we speak and will have full details later today.

Jay

great looking forward to this...

PS, Alex, nikhsub, Gabe, or jay, I am fluent in German, and if ther are any potential language barriers preventing this from getting through ( I have never dealt with Laing PR so I dont know), I would be glad to translate a letter for you guys into formal German.

looking forward to your testing results Jay.

[nealh]

my only thought is..shouldnt all the pumps have starting issues if it was due to the pump impeller rising and clearence is the same for the stock and Petrs tops