100,000 hour pump!

[BRiT]

Dave/Cathar/Anyone of influence... What you must do now is clear. Get busy on convincing Swiftech they need such a pump!

[Cathar]

Originally Posted by BRiT

Dave/Cathar/Anyone of influence... What you must do now is clear. Get busy on convincing Swiftech they need such a pump!

Well, it's a bit of a chicken and the egg scenario at this level.

The water-cooling market isn't really large enough for a single water-cooling manufacturer to justify the investment, but the global enthusiast water-cooling market is plenty large enough (>50K units easily) to justify getting such a pump manufactured.

So therein lies the rub. A single manufacturer has trouble justifying getting the pump made on the basis of their own level of market share. A pump manufacturer can easily justify it on the basis of the global market size, but that pump manufacturer needs to understand and believe the size of the market that they'd be both selling into, and also creating for themselves by providing the pump. It's also somewhat of an "If you build it, they will certainly come" scenario. If such a pump were made, then additional market option avenues would likely also open up for it, just because it exists.

The european waterblock manufacturers are screaming out for a small, quiet and high pressure pump as suits their philosophy, just as the USA makers also desire something that's powerful and of sufficient flow rate capacity and efficiency. Such a pump model as described bridges all worlds, but sadly it's left in the hands of pump manufacturers who either really don't care a great deal about computer water-cooling market-place to realise that, or don't fully understand the water-cooling market-place, even though the market-place size and demands easily justifies it.

I mean, we're talking about making a pump that pretty much is what the single answer would be almost world-wide when people ask "What pump do I choose for my system?"

[flatline]

as noted in another thred wudnt useing bouth sides of the moter makeing a dual head pump for ppl useing dual loops its just an idea but whats the theoretical out put of shutch a creashon ?

tnx 4 time

[Dave]

^ This we can do, as long as tooling is under 10K, I have little problems getting funds. Can do some kind of dual motor in single block design, using the new drive cases.

Cather, in many ways your correct. The exception is what an OEM requires for a liquid cooling pump. AVT as I am aware of, has never had a request for a more powerfull pump. In most cases the MAG is overkill, and we have many request for a smaller pump, 1-3 LPM.

OEM's have no interest in spending considerable extra money (percentage wise) just to get another 1-2DegC.

We are also working with several OEM's that have new high performance systems that operate at lower flows then being suggested on this board.

Our retail market is about 1/50 to 1/100 of our industrial / OEM market. So you can understand where both companies spend there tooling dollars.

Personally as I have expressed before, I think high flow / pressure is not the way to go. Heat exchangers size, design and placement is far more important.

I am going to start testing your latest block, have one on order.
Will start with what I can do with standard drive, then go from there.

I have some idea's for offset coils, that will not require a new circuit board (pump will be longer, instead of wider) I want to pass by Dan and Kevin

[Cathar]

Originally Posted by Dave

Personally as I have expressed before, I think high flow / pressure is not the way to go. Heat exchangers size, design and placement is far more important.

Oh, for sure, bigger/better radiator (heat exchanger) design will always be an easier "win" in terms of temps.

The pump as specified isn't really high flow, more of a well balanced flow/pressure regime that spans the multitude of block designs on the market. We're not after insanely high-flow/pressure, we're after a balanced pump of sufficient pressure/flow that fits the PC water-cooling model about as well as it can possibly fit. We don't want a stronger pump than what's specified, that's stepping backwards due to pump heat dump.

OEM's will ask for a pump that suits their specific needs/wants, but OEM's as you say don't want to eke out the last possible 0.1C of water-cooling performance, and it is here where their philosophy differs from that of the people who frequent this forum.

We wish to explore every last facet of the watercooling technology, and we wish to get that last 0.1C if it's available.

Have already been pushing on getting better radiators/heat exchangers going, and believe me, there are some really good benefits in the works as far as PC water-coolers are concerned - radiators that are specifically tuned for maximum performance when coupled with 20-90cfm fans and in a compact form factor.

This is all we're doing. Examining, pushing and exploring every single last facet, and providing a list of specifications in the hope that a manufacturer one day will read and understand why the need exists, takes the chance, pick it up and run with it.

Sure, the retail water-cooling scene is a drop in the ocean of the OEM world, but it is a scene that is growing. One day a manufacturer will see the market potential, look to break into that market-place with a pump that is as close to all things to all people as it can possibly get, and reap the market penetration rewards from that.

I'll continue to wave the flag and say that this is what is needed and what will work, with detailed rational explanations as to why, and hope that someone will answer the call.

[Dave]

"One day a manufacturer will see the market potential, look to break into that market-place with a pump that is as close to all things to all people as it can possibly get, and reap the market penetration rewards from that."

And the next day, everyone else will copy what they did

As I said, maybe after things quite down a bit, I will see what can be done to get into the 6-7 psi range. We have a 4-5 psi MAG, just need to figure out how to make it "hacker" proof, and likely already have the answer.

Right after I get molds for flow meter, will design a dual case, and see how that works out.

But the REAL answer is for a OEM too come along and ask for a pump in your range. Often what happens in these cases, AVT cuts the OEM a real good price for co-marketing. In otherwords, OEM needs a pump for one purpose, but AVT has a client or a market for another purpose, and reduces tooling costs to customer.

The MAG drive for example, is being used by 11 different companies right now, with little true cross application.

[Arivaldo]

Originally Posted by Dave

The MAG drive for example, is being used by 11 different companies right now, with little true cross application.

Just curious, Dave.
I know you can not tell us companies name, but i would like to hear about MAG applications that is running nowadays.

[Dave]

^ Sure
Silent washer pumps for cars
Micro hydro activation pump
Liquid cooling of optic mount assemblies
Washing fluid pump (consumer product)
TEC liquid cooling system (consumer and industrial product)
Liquid cooling of electrical projection lights
Solar toilets (no joke)
Liquid cooling of servo motors- precision instruments

and of course electronic liquid cooling

[Razor6]

Dave, when you say "hacker proof", are you referring to hiding the internals from other manufacturers trying to duplicate the design, or something to do with end-user modifications?

[bigben2k]

This is going to be very simple; can we expect a longer longevity from the "sapphire on sapphire bearing system"? I'm just looking for a simple Yes/No here, not a quantitative answer.

The next Q will be; what's the expected cost increase?

I'm looking forward to seeing this thing work.

[flatline]

"Right after I get molds for flow meter, will design a dual case, and see how that works out"

1st id like to thankyou for bouth lissening to ppl on this forum and considering there ideas most of my comments used to get bling shouted at them (thow billa has i right to his opinion and shud post hear if he likes imo)

2nd i will be awateing results of bouth your upcomeing projects with baited breath

goodluck!

tnx 4 time

[Dave]

Razor, end user mod's or even worse, companies that produce aftermarket pump chambers that will overload motor, as they do with Laing DDC.

No idea Bigben, system is still in testing and awaiting patend.

If our OEM customer excepts results and proceeds, cost increase could be low do to the number of units produced.

[bobkoure]

Originally Posted by Dave

aftermarket pump chambers that will overload motor, as they do with Laing DDC.

OK - but the DDC is one of those pumps that isn't so much "hack"-able as it just cries out to have a ninety degree bend removed from the inlet.
AFAIK, with C-systems if you want to have a straight inlet, you just buy the straight-inlet model.

Then there's integral reservoirs (I'm personally a big fan of the "hybrid" in-line/submersed pump setup you get with a reservoir attached to the pump inlet). I'll bet there are ways to make it easy to attach a semi-integral reservoir to one of the C-systems pumps without disturbing the chamber - maybe just a thread-to-smooth-pipe adapter that someone could use to design a "push-on" reservoir around (the way Innovatek does for the little Eheims). Maybe just something like a thread-to-male-speedfit adapter

(which would likely need shorter thread depth, and as speedfit adapters tend to have relatively small IDs for their ODs, maybe get a "special" from JohnGuest or just put a larger diameter thread in the pump inlet housing. A reservoir designer/manufacturer could then either use a pair of o-rings in a machined-to-fit socket (like Innovatek) or a cartridge style speedfit. Just a thought, of course...
The other advantage with speedfits as they allow the connections to twist a bit without leaking. Very nice for the aftermarket.

Then there are modifications for bling. There's no predicting that'll get done there. UV reactive is popular, as is being able to see some sort of moving part, well, move, preferably with lighting - and then there are the guys who years ago would have had HD Electraglides with lost of chrome and bullet lights. Beats me - but I've no clue about bling.

Speaking of Innovatek, I'm surprised they aren't already at your door. You have a good replacement/upgrade for their current 12V 1046 - it might even be quieter, if such a thing was possible...

[Dave]

There is no problem with "bling" Mod's, the problem is changing the inlet or outlet diameter.

Increasing diameter, will increase flow, which increases pump load.... it does not decrease pump load as some people think.

Pump chambers in an intergrated pump/motor design, can not be modified without effecting load.

Best I know Bob, C-Systems will not work with Innovatek, and before you ask, I have no idea why. It is unlikely "upper types" will let me know, and if they did I dout I could post it.

Intergrated reserves, UV mods, all good
Just don't change chamber, and you must be careful when designing a reserve to avoid vortex propagation.

[Risky]

Innovatek aren't the only ones that did reservoir for the Eheims.

Of couse all you need is something about the right size with a 1/4 thread hole and then a double-male 1/4 pipefitting to attach. you could do this very nicely with watercool's reservoir which has 1/4threads

[Roscal]

Originally Posted by Dave

Increasing diameter, will increase flow, which increases pump load.... it does not decrease pump load as some people think.

Pump chambers in an intergrated pump/motor design, can not be modified without effecting load.

Not really a problem. A motor isn't designed to be good at 10W and burn at 11W.. There's a tolerance in electronics components about 2 times, power transistors can handle easily more current to coils. A "hack mod" will give you ~0.5-2 W more in total power drawn (real flow gain in a circuit is quite small), and electronics will handle that without any problems. Doing a hack, you'll approach the max consumption when there's no restriction in the loop in normal conditions, and power drawn differences between 0L/h and max flowrate is always small.

It's probably worst to boost a pump with 15 V in its face than make a more "optimized" volute. Laing allows several volutes for the D4/D5 with different diameters and without any changes in motor/electronics. Others manufacturers propose different rotors to choose our own flow characteristics without motor changes and however there could have very big differences in final flowrate.

[Dave]

^ Sorry no, in brushless DC motor design, you often run the transistors beyond max load, because duty cycle is low. When you increase flow, you decrease RPM, which in turn increases your duty cycle, and this can push coils and transistors beyond test conditions.

A 2W increase on a 10W pump is a 20% increase, and if you work with electronic life prediction software, the drop off is clear.

Increasing voltage will only decrease duty cycle, and therefore is often fine.
You need to watch a brushless motor on a scope to see this relationship.

As a manufacture, we do not care if someone mods there pump themselves, because it voids warranty. However if you buy a direct replacement chamber, we have no way of knowing if it has been modded.