What if you had a large tall collumn, ~ 5 feet tall that wat say 6in in diameter. You had an outlet on the bottom of the collumn which would feed your system. Then you would have a reservoir on at the end of the loop before the pumps. Then you would have a large pump or a few smaller pumps to move the water back up to the top of the collumn. [water tower->sytem->pump] The pumps would just have to move the water out of the reservoir at the same rate as it is being filled. This might be pointless because the pumps still have to put the same amount of energy into the water to lift it as the pumps would use just pumping in the usual way in wc systems. It might, however, be a better way to create pressure in a system.
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
Thinking out loud - Centrifugal vs. Positive Displacement.
- Thread starter Graystar
- Start date
- Joined
- Jul 15, 2002
- Location
- Toronto, Canada
You cannot cheat in this way, and I'll tell you why. This idea fixes the amount of head that the system has by using the water column to drive the flow. The flow in the system will stabilize at a flow rate that corresponds with a resistance equal to the difference between the column water surface and the lower reservoir water surface (i.e. 5ft in this case). Since this head difference that is causing the flow is the same as the head difference that the pump is working against when refilling the first cylinder you've not gained a single thing besides two large reservoirs and a bunch of extra tubing (and roughly double the system flow resistance).
Besides, if you're going to use a tall water column to drive flow, why take only 5ft? There are lots of pumps, mine included, that can do much better than that.
Besides, if you're going to use a tall water column to drive flow, why take only 5ft? There are lots of pumps, mine included, that can do much better than that.
- Joined
- Jun 22, 2002
- Location
- Asheville, NC
LOLZ!!!!
unless my chem teacher is a lier i do belive taht oil actually absorbs/transfer' heat better than water...
- Joined
- Jul 15, 2002
- Location
- Toronto, Canada
Sorry CrashOveride, but I do believe that your teacher is a liar.
- Joined
- Nov 12, 2002
- Location
- Rootstown, OH
first, I am not sure if the thermal transfer properties of oil are better than water, but I am pretty sure they aren't.
Viscosity has always been shown to be a VERY big issue in our applications, perhaps it could be overcome by doing things differently, however if you have to overcome more by using oil then what is it helping?
OP
- Joined
- Aug 20, 2002
- Location
- Brooklyn, NY
- Thread Starter
- #28
Here's a document on things to consider when deciding between a PD pump or a centrifugal pump.
http://www.pumpschool.com/intro/pd vs centrif.pdf
The document mentions a Best Efficiency Point (BEP) for cetrifugal pumps. I sure would like to know how to calculate that to see where on the curve our pumps are operating.
I was reading a review of the Swiftech pump (MCP-300) as compared to the Eheim 1048.
http://www.amdmb.com/article-display.php?ArticleID=216
The Eheim, with a 158 GPH rating, was only pushing 43 gallons through the system, and the new Swiftech, with a 340 GPH rating, was pushing only 60 gallons. Just seems to me like we're working on the wrong end of the BEP curve.
http://www.pumpschool.com/intro/pd vs centrif.pdf
The document mentions a Best Efficiency Point (BEP) for cetrifugal pumps. I sure would like to know how to calculate that to see where on the curve our pumps are operating.
I was reading a review of the Swiftech pump (MCP-300) as compared to the Eheim 1048.
http://www.amdmb.com/article-display.php?ArticleID=216
The Eheim, with a 158 GPH rating, was only pushing 43 gallons through the system, and the new Swiftech, with a 340 GPH rating, was pushing only 60 gallons. Just seems to me like we're working on the wrong end of the BEP curve.
Those results are not atypical of a lot of rigs (I personally don't use two CPU blocks, but I plan to use a heatercore, for a while). I'll also be using 3/4 tubing, not 3/8 (but that wouldn't have made a difference here)Graystar said:
This is an issue about the pump operating at peak energy efficiency, which we often ignore. It's nice to be able to look at a pump's curve, to determine what one really needs.
All in all, you need the most amount of "hydraulic" power, but you should pick a pump that'll run efficiently. Price is usually, and by far, the biggest selection factor: I picked up a Little Giant 2-MDQ-SC, because I found it for $40.
this has been discussed (lamented by me) on a somewhat more technically oriented forum (procooling) at some length
yes, the BEP will typically be around 70% of the max free flow rate
note that centrifugals hav also a minimum, below which the impeller is, in aerodynamic terms, 'stalled'
but this minimum is seldom described by the pump mfgrs, and is totally ignored by WCers
this is the Hydor pump data (off the box, 50~ data):
the 60~ data for the different US version sold by CoolTechnica (w/o the minimums !) is:
the penalty for this pump misapplication is significantly increased pump heat
be cool
yes, the BEP will typically be around 70% of the max free flow rate
note that centrifugals hav also a minimum, below which the impeller is, in aerodynamic terms, 'stalled'
but this minimum is seldom described by the pump mfgrs, and is totally ignored by WCers
this is the Hydor pump data (off the box, 50~ data):
the 60~ data for the different US version sold by CoolTechnica (w/o the minimums !) is:
the penalty for this pump misapplication is significantly increased pump heat
be cool
OP
- Joined
- Aug 20, 2002
- Location
- Brooklyn, NY
- Thread Starter
- #31
I have just realized something. If you replace your 300 GPH/6' Head pump with a 600 GPH/6' Head pump, the flow rate through your system will be exactly the same. All you've succeeded in doing was adding more heat to the system.
Centrifugal pumps move water by producing pressure. That pressure is described as the height of a column of water known as "Head". A taller head means more pressure.
You need pressure to push water through a watercooling loop in order to overcome the resistance of the loop. The speed of the water is determined by the pressure of the water. This means that two pumps with equal Head ratings will move water at the same speed. The volume of water that a pump can deliver has no meaning at all in watercooling systems for PCs.
Therefore, what we actually want is a pump that moves a small amount of water at a very high head.
So forget those GPH ratings and concentrate on head ratings!
Centrifugal pumps move water by producing pressure. That pressure is described as the height of a column of water known as "Head". A taller head means more pressure.
You need pressure to push water through a watercooling loop in order to overcome the resistance of the loop. The speed of the water is determined by the pressure of the water. This means that two pumps with equal Head ratings will move water at the same speed. The volume of water that a pump can deliver has no meaning at all in watercooling systems for PCs.
Therefore, what we actually want is a pump that moves a small amount of water at a very high head.
So forget those GPH ratings and concentrate on head ratings!
welcome to the party Graystar
but it may increase slightly with the larger pump if its deadhead pressure is greater, note 'may'
but, unfortunatly for us 'purists', this also about what works;
and what it costs
there are jillions of pumps made for aquarium users,
how many for WCers ?
be cool
but it may increase slightly with the larger pump if its deadhead pressure is greater, note 'may'
but, unfortunatly for us 'purists', this also about what works;
and what it costs
there are jillions of pumps made for aquarium users,
how many for WCers ?
be cool
- Joined
- Oct 12, 2001
- Location
- Los Angeles
Graystar, you may be interested in these curves from reading your last post... As BillA said, none made FOR WC'ers but these do a fine job if you are willing to pay the price or hunt ebay religiously.
- Joined
- Dec 1, 2002
Boy, those iwaki's keep looking nice for impingement every time I see those data sheets.
OP
- Joined
- Aug 20, 2002
- Location
- Brooklyn, NY
- Thread Starter
- #37
That MD-6Z looks pretty perfect! 115 bucks though...ouch :-(
- Joined
- Oct 12, 2001
- Location
- Los Angeles
Like I said, HUNT ebay, I got my MD-15R for $70 and it was brand new.Graystar said:
OP
- Joined
- Aug 20, 2002
- Location
- Brooklyn, NY
- Thread Starter
- #39
nikhsub1 said:
Found a 6N, no 6Z....damn! Close, but no cigar.
Guess I'll have to setup one of the automatic emails.
Thing is though that these are still centrifugal pumps. Would be nice to find a real PD pump that's really small.
- Joined
- Oct 12, 2001
- Location
- Los Angeles
Yes they are centrifugal... but they are one of the best (if not the best) for our use, provided you get the right model for the job. They are almost silent, and put almost no heat into the water even though the casing gets very hot. I had a 20RL, too big to fit in the case in a reasonable fashion so I sold it and opted for the 15R instead, about the same size as a 1250.Graystar said:
