New pump big enough?

Dave · 08-09-2005, 10:09 AM

3.5"x3.5"x1.75", 25-30W..... 14+ PSI head.
Max noise of 24dB

Just been asked to design tooling for this pump, to determine costs.

Understand it is an OEM request to AVT.

Wondering if there is also any interest from you guys as well?

JamesAvery22 · 08-09-2005, 10:38 AM

Quote:

Originally Posted by Dave

3.5"x3.5"x1.75", 25-30W..... 14+ PSI head.
Max noise of 24dB

Just been asked to design tooling for this pump, to determine costs.

Understand it is an OEM request to AVT.

Wondering if there is also any interest from you guys as well?

Definitely wouldn't say it was too big. 14psi? Isnt that close to like 10m head?? How much?

Etacovda · 08-09-2005, 10:54 AM

9.86m head

Looks like Cathars wet dream (hohoho)

Ls7corvete · 08-09-2005, 10:59 AM

How do you think prices would work out?

Sence you already have established retailers how much time would it take to get a new pump to the market?

Dave · 08-09-2005, 11:48 AM

Think your getting ahead, I was just asked to judge interest.

AVT or C-Systems will reduce tooling costs to our OEM client, if there is another market for the pump, and they agree to co-marketing.

The pump is for liquid cooling, just NOT for PC cooling at this time.

Personally I think it is too big, and uses to much power.

Happy with the current MAG, or in my case two of them, one a combo unit (block intergrated) and a second pump for GPU cooling, with it's own rad.

JamesAvery22 · 08-09-2005, 11:54 AM

I think its (spec wise) awesome for us who have large external setups. Past the mcp655 there really aren't many options other than getting multiple pumps or an Iwaki. And 30watts compared to 50watts makes me tingle, not to mention how much quieter it would be if it follows the current CSP-mag trend. Id ebay my iwaki and buy one right away if the price was good. But for the avg "single block single radiator" setup, I agree its too much. Im the large external setup though so Im VERY interested

Cathar · 08-09-2005, 12:21 PM

Quote:

Originally Posted by Etacovda

Looks like Cathars wet dream (hohoho)

Just to re-iterate my model specs for the ultimate water-cooling pump:

20' or 6.5mH2O (~9.5PSI) of peak pressure at dead-head
3.5gpm or 13.5lpm of peak flow
11.5' or 3.5mH2O of pressure head at 2gpm or 7.5lpm (**)
12vdc
25dBA noise level at most
20W peak power consumption at (**) flow/pressure noted above (preferably less)
Maximum of 10x10x10cm bounding box size, including barbs, preferably even smaller

What is the targetted peak flow on this pump Dave? If it's anything in the 12-20LPM range then it'd still make for a fantastically desirable pump, even if people have to under-volt it in some fashion.

The "ideal pump" specs I list is a very nice target point, but it's also nice for some people to have additional grunt on tap if required, such as those with dual CPU's and/or lots of radiator capacity to deal with the extra pump heat. Ideally that's achieved by people sticking 13-14v through their pump with an additional PSU over and above the specs I listed.

The specs you gave appear to match something like a stock Iwaki RD-30, and that's definitely too hot for most applications, but then an RD-30 at full-speed dumps more like 30-35W of heat into the loop. Need to know the peak flow rate on this planned pump first though.

Very interesting nonetheless, even if it slightly overshoots the mark. Much like radiator fans and/or my Iwaki RD-30, it's always better to have a little bit too much and undervolt to compensate, than to have too little and be left wishing for more.

No chance of simply doing a minor electronics mod and releasing a pump that runs at slightly lower RPM for the water-cooling world as opposed to your OEM group? I estimate that it'd need to run at 80% of the RPM that your OEM application is demanding. That'd be perfect then, 'cos AVT will know that the pump's motor is rated for the loading if people do things like stick an extra 20% voltage through it. Leverage off the OEM supplier, with a really minor electronics change, and provide the water-cooling world with a dream pump that we're all waiting for. Heck, it'd be even quieter at 80% RPM too.

Now I am drooling. First, need to know peak open flow planned for OEM.

Dave · 08-09-2005, 01:09 PM

Voltage is No issue Cather, even with current MAG, as it will reduce duty cycle and transistor / coil loading, not increased it.

In a fixed flow system, this effect is common.

What will to flow be for new pump?
Need to guess at 180-200 gph, as it was no designed by myself, and we where only given a targeted head.

It is odd you and this company have same spec's, as I can just about be 100% sure they have nothing to do with PC liquid cooling. It is a very large world wide company, and I guess it is possible there is some co-developement somewhere.

I would suggest one of there engineers is likely reading this board

JamesAvery22 · 08-09-2005, 01:20 PM

Quote:

Originally Posted by Cathar

...
If it's anything in the 12-20LPM range then it'd still make for a fantastically desirable pump, even if people have to under-volt it in some fashion.
...
Now I am drooling. First, need to know peak open flow planned for OEM.

+

Quote:

Originally Posted by Dave

...
What will to flow be for new pump?
Need to guess at 180-200 gph, as it was no designed by myself, and we where only given a targeted head.
...

=
Some math...

180GPH = 3GPM = 11.355LPM
200GPH = 3.3_GPM = 12.62LPM

Thats close enough, right?

If Cather is drooling Im lactating.

maxSaleen · 08-09-2005, 02:55 PM

I'd go for it. Have a good pump is like having a big engine (in your car). You might not always need it, but it's nice to have it. What would you expect the heat dump to be? What type of communication will be used between the impeller and the motor (magnetic or shaft)?

Dave · 08-09-2005, 03:09 PM

^ Take a current MAG, add 1.4" to it, plus two more poles and larger transistors.... there it is

billbartuska · 08-09-2005, 03:11 PM

Quote

No chance of simply doing a minor electronics mod and releasing a pump that runs at slightly lower RPM for the water-cooling world as opposed to your OEM group?

And with buildt in flow monitoring for us?

Dave · 08-09-2005, 04:12 PM

No problem, just add different magnetic material to produce more back EMF, and reduce RPM.

As to monitoring, I can not say at this time, sorry.

Cathar · 08-09-2005, 05:08 PM

Hmmm, if targetted peak flow is in the 180-200gph mark, then by reducing impeller speed by 20% we'd be talking about getting down to 144-160gph, which is getting too low for some of the higher flow focused blocks (e.g. Asetek Antarctica, White Water, DD RBX/TDX) once we factor in typical system resistances. Now by "too low" I really mean very marginally less than optimal and such a pump would still be clearly better than anything else flow/pressure/heat wise

So if that's the case, then a compromise at 90% of the OEM RPM speed then, which should also still keep Dave happy for purposes of motor loading. That would reduce power draw to approximately 20-24W, for a heat dump in the ~17-20W range, which is still very, very nice, if just a touch high. No more than 16W heat dump is what I was hoping for, but hey, beggars can't be too choosy. I'm just hunting for that last 0.1-0.2C as usual.

Dave, not sure why they'd be following my specs. My pump specs are derived from intensive analysis of parameters very particular to PC water-cooling only, and certainly are not generic for all scenarios. I'd say perhaps just a chance of lucky co-incidence.

Cathar · 08-09-2005, 05:47 PM

Wait a minute. Gotten my math wrong on the rpm vs power thing. Power is proportional to RPM^3. So at 90% RPM, power would be ~0.73 as at 100%.

So 0.73 of 25-30w = 18.25 - 22W. Factoring in that not all input power makes it into the loop as heat (~85%), so 15.5W - 18.7W.

So yeah, this then reaffirms that given the initial specs as stated that water-coolers would be wanting something that spun at 90% of OEM rpm speed.

kaotic504 · 08-09-2005, 07:44 PM

well given final specs of the pump, how much heat it would add to a loop, and how much it would cost, i'd seriously be interested in one. i've already bought 2 MAG's, but if this was as powerful as both my MAG's, i'd definitely sell one and keep the other as a backup.

Ls7corvete · 08-09-2005, 07:55 PM

Even if the OEM did change specs out of our desired range slightly you would be able to compensate by making water cooling specific impellor correct?

Dave · 08-10-2005, 06:59 AM

^ Sure, should have tooling costs figured by end of week.
I suspect Dan will go for a reduction given there is interest here.

Just be aware our client is a very big company, and as such, likely disfunctional, so it could be months before anyone gives a go ahead.

Cathar · 08-10-2005, 08:37 AM

Sure, there'd be big interest, the amount of which depends on the final cost. If it costs upwards of $100, then the market interest will be quite small. Get it into the $75-80 range and the demand will be competing directly with the other established DC pumps, despite the specs of this pump being better. Get it into the sub-$70 range and the entire mid-range and above enthusiast market is pretty much your oyster.

We're talking about something with slightly less heat dump than 2 of the AQX-50Z/MCP600's in series, but equalling effectively 3 x AQX-50Z/MCP600's in series in terms of pressure. Even in an approximated "wet dream" specs format with the 90% OEM RPM impeller speed, such a pump would stand head and shoulders above any other single pump on the market out there. Throw in the low-noise factor, make sure it's reliable, and coupled with its small size then people would be very hard pressed to find a reason NOT to buy it.

About the only reasons why people would choose something else would be:

Hard cost/budgetary constraints - not soft constraints. In this market I've seen that generally if you give people enough of a good reason to spend an extra $25, they will do so.
Ultra-low noise crowd (i.e. Eheim 1046/1048's would still be the favorites here). These would be the people who favor silence above all else, and as such a super-performance tweaked pump still means little to them if they can save 5dBA on pump noise.
Low heat load applications (<50W). For people using very cool running CPU's even a ~16W pump heat dump becomes a liability, but the size of this market who are also water-cooling is quite small, and generally cross-over somewhat with the ultra-low-noise crowd.
Low radiator capacity crowd. The pump as defined with its heat dump would "blur" somewhat performance wise against other pumps in very low radiator capacity scenarios. By that, meaning people with radiator C/W's of ~0.1 and higher. Again, this would represent quite a small section of the overall market as even a single 12cm fanned radiator with a low-moderate quiet fan will achieve better than 0.1 C/W's.

My feeling is that the market is very, very hungry for a pump that does it all. If AVT make this pump, and make it to a level of finish and quality that is comparable with the better alternative 12vdc pumps on the market, then AVT will stand to capture a huge amount of marketshare with such a pump if marketed correctly, and even futher, it will create a greater market for itself just because it exists.

How big would that market be? I'd be surprised if it were less than 10K units over time with a sub-$75US price-point.

Dave · 08-10-2005, 10:26 AM

Our designs are all very cost effective Cather, I would guess a 20-30% increase over current MAG pricing. Please remember we successfully compete in the automotive parts industry, where cost I admit is everything.

With an OEM picking up most of tooling bill, any other sales are just "gravy"

Long Haired Git · 08-10-2005, 09:38 PM

Only other criteria I'd offer are:
1) Ability to cope with wide range of temperatures (ie Mred at negative 20°C, or a passive rig with 50°C)
2) Ability to change the inlet/outlet barb sizes for 3/8", 7/16",1/2",5/8" tubing etc.
3) Ability to underclock/overclock by varying voltage (ie, good if it ran okay at 7v for my winter, and then I change it to run at 12v in summer).
4) Given you've got electronics, power, and easy access to the coolant, I've always thought the pump would be a good place to put a temperature probe and display. But then, I'm weird like that and I doubt the OEM would be interested.

Risky · 08-11-2005, 05:43 AM

I might suggest that you thread the part at 3/8 instead of 1/4 for the bigger pump (assuming this is physically possible. It may or may not make a material difference, but some people will be attacted by any tiny reduction in restriction.

Dave · 08-11-2005, 07:01 AM

Finished last night, and AVT agreed to kick in 20K, and C-Systems another 10K, so we should be presenting an attractive bid

Hopefully we will hear something soon.

Razor6 · 08-12-2005, 04:47 PM

I hope your competitors don’t read this board.

kaotic504 · 08-12-2005, 11:30 PM

damn, watercooling just gets sweeter by the day.

08-09-2005, 10:09 AM	#1
Dave Cooling Savant Join Date: Oct 2004 Location: Canada Posts: 260	New pump big enough? 3.5"x3.5"x1.75", 25-30W..... 14+ PSI head. Max noise of 24dB Just been asked to design tooling for this pump, to determine costs. Understand it is an OEM request to AVT. Wondering if there is also any interest from you guys as well?

08-09-2005, 10:54 AM	#3
Etacovda Cooling Savant Join Date: Feb 2004 Location: Dunedin NZ Posts: 735	9.86m head Looks like Cathars wet dream (hohoho) __________________ Hypocritical Signature I tried to delete: Procooling: where scientific principles are ignored because big corporations are immune to mistakes and oversights.

08-09-2005, 11:54 AM	#6
JamesAvery22 Cooling Savant Join Date: May 2004 Location: Northern VA Posts: 383	I think its (spec wise) awesome for us who have large external setups. Past the mcp655 there really aren't many options other than getting multiple pumps or an Iwaki. And 30watts compared to 50watts makes me tingle, not to mention how much quieter it would be if it follows the current CSP-mag trend. Id ebay my iwaki and buy one right away if the price was good. But for the avg "single block single radiator" setup, I agree its too much. Im the large external setup though so Im VERY interested __________________ Stabbing. My Heatware

08-10-2005, 08:37 AM	#19
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Sure, there'd be big interest, the amount of which depends on the final cost. If it costs upwards of $100, then the market interest will be quite small. Get it into the $75-80 range and the demand will be competing directly with the other established DC pumps, despite the specs of this pump being better. Get it into the sub-$70 range and the entire mid-range and above enthusiast market is pretty much your oyster. We're talking about something with slightly less heat dump than 2 of the AQX-50Z/MCP600's in series, but equalling effectively 3 x AQX-50Z/MCP600's in series in terms of pressure. Even in an approximated "wet dream" specs format with the 90% OEM RPM impeller speed, such a pump would stand head and shoulders above any other single pump on the market out there. Throw in the low-noise factor, make sure it's reliable, and coupled with its small size then people would be very hard pressed to find a reason NOT to buy it. About the only reasons why people would choose something else would be: Hard cost/budgetary constraints - not soft constraints. In this market I've seen that generally if you give people enough of a good reason to spend an extra $25, they will do so. Ultra-low noise crowd (i.e. Eheim 1046/1048's would still be the favorites here). These would be the people who favor silence above all else, and as such a super-performance tweaked pump still means little to them if they can save 5dBA on pump noise. Low heat load applications (<50W). For people using very cool running CPU's even a ~16W pump heat dump becomes a liability, but the size of this market who are also water-cooling is quite small, and generally cross-over somewhat with the ultra-low-noise crowd. Low radiator capacity crowd. The pump as defined with its heat dump would "blur" somewhat performance wise against other pumps in very low radiator capacity scenarios. By that, meaning people with radiator C/W's of ~0.1 and higher. Again, this would represent quite a small section of the overall market as even a single 12cm fanned radiator with a low-moderate quiet fan will achieve better than 0.1 C/W's. My feeling is that the market is very, very hungry for a pump that does it all. If AVT make this pump, and make it to a level of finish and quality that is comparable with the better alternative 12vdc pumps on the market, then AVT will stand to capture a huge amount of marketshare with such a pump if marketed correctly, and even futher, it will create a greater market for itself just because it exists. How big would that market be? I'd be surprised if it were less than 10K units over time with a sub-$75US price-point. Last edited by Cathar; 08-10-2005 at 08:44 AM.

08-10-2005, 09:38 PM	#21
Long Haired Git Cooling Savant Join Date: Jan 2003 Location: Sydney, Oz Posts: 336	Only other criteria I'd offer are: 1) Ability to cope with wide range of temperatures (ie Mred at negative 20°C, or a passive rig with 50°C) 2) Ability to change the inlet/outlet barb sizes for 3/8", 7/16",1/2",5/8" tubing etc. 3) Ability to underclock/overclock by varying voltage (ie, good if it ran okay at 7v for my winter, and then I change it to run at 12v in summer). 4) Given you've got electronics, power, and easy access to the coolant, I've always thought the pump would be a good place to put a temperature probe and display. But then, I'm weird like that and I doubt the OEM would be interested. __________________ Long Haired Git "Securing an environment of Windows platforms from abuse - external or internal - is akin to trying to install sprinklers in a fireworks factory where smoking on the job is permitted." (Prof. Gene Spafford) My Rig, in all its glory, can be seen best here AMD XP1600 @ 1530 Mhz \| Soyo Dragon + \| 256 Mb PC2700 DDRAM \| 2 x 40 Gb 7200rpm in Raid-0 \| Maze 2, eheim 1250, dual heater cores! \| Full specifications (PCDB)

08-09-2005, 10:59 AM	#4
Ls7corvete Cooling Savant Join Date: Oct 2004 Location: FL Posts: 164	How do you think prices would work out? Sence you already have established retailers how much time would it take to get a new pump to the market?

08-09-2005, 11:48 AM	#5
Dave Cooling Savant Join Date: Oct 2004 Location: Canada Posts: 260	Think your getting ahead, I was just asked to judge interest. AVT or C-Systems will reduce tooling costs to our OEM client, if there is another market for the pump, and they agree to co-marketing. The pump is for liquid cooling, just NOT for PC cooling at this time. Personally I think it is too big, and uses to much power. Happy with the current MAG, or in my case two of them, one a combo unit (block intergrated) and a second pump for GPU cooling, with it's own rad.

08-09-2005, 01:09 PM	#8
Dave Cooling Savant Join Date: Oct 2004 Location: Canada Posts: 260	Voltage is No issue Cather, even with current MAG, as it will reduce duty cycle and transistor / coil loading, not increased it. In a fixed flow system, this effect is common. What will to flow be for new pump? Need to guess at 180-200 gph, as it was no designed by myself, and we where only given a targeted head. It is odd you and this company have same spec's, as I can just about be 100% sure they have nothing to do with PC liquid cooling. It is a very large world wide company, and I guess it is possible there is some co-developement somewhere. I would suggest one of there engineers is likely reading this board

08-09-2005, 02:55 PM	#10
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	I'd go for it. Have a good pump is like having a big engine (in your car). You might not always need it, but it's nice to have it. What would you expect the heat dump to be? What type of communication will be used between the impeller and the motor (magnetic or shaft)?

08-09-2005, 03:09 PM	#11
Dave Cooling Savant Join Date: Oct 2004 Location: Canada Posts: 260	^ Take a current MAG, add 1.4" to it, plus two more poles and larger transistors.... there it is

08-09-2005, 03:11 PM	#12
billbartuska Cooling Savant Join Date: Jun 2005 Location: Skokie, Illinois Posts: 322	Quote No chance of simply doing a minor electronics mod and releasing a pump that runs at slightly lower RPM for the water-cooling world as opposed to your OEM group? And with buildt in flow monitoring for us?

08-09-2005, 04:12 PM	#13
Dave Cooling Savant Join Date: Oct 2004 Location: Canada Posts: 260	No problem, just add different magnetic material to produce more back EMF, and reduce RPM. As to monitoring, I can not say at this time, sorry.

08-09-2005, 05:08 PM	#14
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Hmmm, if targetted peak flow is in the 180-200gph mark, then by reducing impeller speed by 20% we'd be talking about getting down to 144-160gph, which is getting too low for some of the higher flow focused blocks (e.g. Asetek Antarctica, White Water, DD RBX/TDX) once we factor in typical system resistances. Now by "too low" I really mean very marginally less than optimal and such a pump would still be clearly better than anything else flow/pressure/heat wise So if that's the case, then a compromise at 90% of the OEM RPM speed then, which should also still keep Dave happy for purposes of motor loading. That would reduce power draw to approximately 20-24W, for a heat dump in the ~17-20W range, which is still very, very nice, if just a touch high. No more than 16W heat dump is what I was hoping for, but hey, beggars can't be too choosy. I'm just hunting for that last 0.1-0.2C as usual. Dave, not sure why they'd be following my specs. My pump specs are derived from intensive analysis of parameters very particular to PC water-cooling only, and certainly are not generic for all scenarios. I'd say perhaps just a chance of lucky co-incidence.

08-09-2005, 05:47 PM	#15
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Wait a minute. Gotten my math wrong on the rpm vs power thing. Power is proportional to RPM^3. So at 90% RPM, power would be ~0.73 as at 100%. So 0.73 of 25-30w = 18.25 - 22W. Factoring in that not all input power makes it into the loop as heat (~85%), so 15.5W - 18.7W. So yeah, this then reaffirms that given the initial specs as stated that water-coolers would be wanting something that spun at 90% of OEM rpm speed.

08-09-2005, 07:44 PM	#16
kaotic504 Cooling Neophyte Join Date: Mar 2004 Location: New Orleans Posts: 86	well given final specs of the pump, how much heat it would add to a loop, and how much it would cost, i'd seriously be interested in one. i've already bought 2 MAG's, but if this was as powerful as both my MAG's, i'd definitely sell one and keep the other as a backup.

08-09-2005, 07:55 PM	#17
Ls7corvete Cooling Savant Join Date: Oct 2004 Location: FL Posts: 164	Even if the OEM did change specs out of our desired range slightly you would be able to compensate by making water cooling specific impellor correct?

08-10-2005, 06:59 AM	#18
Dave Cooling Savant Join Date: Oct 2004 Location: Canada Posts: 260	^ Sure, should have tooling costs figured by end of week. I suspect Dan will go for a reduction given there is interest here. Just be aware our client is a very big company, and as such, likely disfunctional, so it could be months before anyone gives a go ahead.

08-10-2005, 10:26 AM	#20
Dave Cooling Savant Join Date: Oct 2004 Location: Canada Posts: 260	Our designs are all very cost effective Cather, I would guess a 20-30% increase over current MAG pricing. Please remember we successfully compete in the automotive parts industry, where cost I admit is everything. With an OEM picking up most of tooling bill, any other sales are just "gravy"

08-11-2005, 05:43 AM	#22
Risky Cooling Neophyte Join Date: Aug 2002 Location: London Posts: 96	I might suggest that you thread the part at 3/8 instead of 1/4 for the bigger pump (assuming this is physically possible. It may or may not make a material difference, but some people will be attacted by any tiny reduction in restriction.

08-11-2005, 07:01 AM	#23
Dave Cooling Savant Join Date: Oct 2004 Location: Canada Posts: 260	Finished last night, and AVT agreed to kick in 20K, and C-Systems another 10K, so we should be presenting an attractive bid Hopefully we will hear something soon.

08-12-2005, 04:47 PM	#24
Razor6 Cooling Savant Join Date: Apr 2002 Location: Houston, TX Posts: 192	I hope your competitors don’t read this board.

08-12-2005, 11:30 PM	#25
kaotic504 Cooling Neophyte Join Date: Mar 2004 Location: New Orleans Posts: 86	damn, watercooling just gets sweeter by the day.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)