Sounds very, very similar to the Pondmaster 4200 pump (that's the Australian Pondmaster brand which is different to the USA Pondmaster brand) that I had once (I sold it). It was able to push slightly more than 10lpm of flow through a White Water system plus radiator. It was rated for 4200lph, 4.4m of pressure head, and 90W of power.

The secret is in the peak head though. 15.5', or 4.7m of pressure head at shut-off (no flow).

I suggest that the problem with the flow measurement is for you to find. We can't see what your setup is, so we can't see what might be wrong, even if you describe it correctly, but I can say for an almost certainty that however you're measuring it is giving values about 3x higher than actual.

If it can help, to diameter normal intern of a connection of 1\2 " it is 3\8, all my connections have internal diameter of 7\16 and my tubes are in on measure with internal diameter of 9\16

That would help a little, but not a huge amount, and certainly not enough to triple the flow rate, maybe help by 10%, if that. The bulk of the restriction is within the waterblock and radiator.

I will re-do the tests and to photograph to try to find the mistake in the weekend, what continues me inconveniencing it is behind the flow that I obtained with a bomb from 30w to minutes, with the rad 6x6 and the tc4 in the circuit, I obtained 2 liters to every 11 seconds, that according to the calculations is also impossible, however it was done.


thank you very much cathar

Bil your water column x flow rate in the tests of the cpu die simulator of the lrww is simulating one of 1\4 of diameter?

sorry Ivo, I don't understand the question

the flow rate was measured with a Danfoss MAG 5000 flowmeter with MAG 1100 ¼in. magnetic flowtube (±1/2% of reading accuracy)
- magnetic flow meters read the velocity which must be within a certain range, had I wished to measure higher flow rates I would have used a larger flow tube (which I have up to 2" ID, and positive displacement pumps to match)

pressure drop with a Foxboro 823DP differential pressure transmitter and Digitek digital gauge (~±1 to 2% accuracy)

the pressure drop is accurate over the flow rates shown, and as Cathar has pointed out will increase exponentially beyond the flow rates tested

varying the diameter of the column to be supported if it varies the pressure to be submitted to reach the same level

a bomb pushing a column with diameter of 1\2 inch reaches a height

the same bomb pushing a column with diameter of 3\4 of inch reaches another height

the commercial bombs are hour measured with a column in your exit diameter, hour measured with a diameter considered pattern by the manufacturer, or just for the generated psi.

I wanted something simple, I wanted to know when you speak in 3mh2o, that pressure you apply to the fluid to simulate that, I can like this in a rustic way to compare with the supported maximum columns published by the manufacturers of the bombs with yours, since they don't have a defined pattern

did you understand my considerations?

nope

a pump with a 2m head capability will push a column 2m in height
1/8" diameter, 2" diameter - no difference at all
(just longer to pump the larger volume)

I must compliment you on your radiator design - it seems to be very efficient. Drool-inducing, too - it's been a while since i've seen something that polished.

However, I have to agree with cathar, unregistered and the rest of those who have been saying you can't possibly pump 20-30 lpm through your setup with anything short of a massive industrial pump. I'm looking forward to seeing the tests redone, if you do get a chance to do them this weekend. Something's amiss - and if not, you need to teach the rest of us to break the laws of physics too.

I will re-do the tests on the weekend, some on this day, in some point I wandered, and as the tests were made only one time that it is not impossible of happening.


If the results still go out of the normal or not, I will take the complete system for laboratory of the federal technical school or federal university, which I should have done before publishing rude numbers of flow rate.


living and learning

I am looking forward to the new / revised results. ;)

Good work.

Gentlemen, re-done all the tests of flow rate and the results are posted below.

I actually was publishing the data of entrance flow rate in the block and not exit, a member of the forum xtremeforum asked me the question of as my system obtained of flow rate of the bomb going by the rad, (referring the pressure drop caused by the rad) he asked me which the flow rate that I arrived in the block, I was with that in my head as being final flow rate. (that even so were not precisely correct )

it seems, and it is a stupid mistake, but i was made, and luckily having corrected on time for the bill ,cathar and other friends to which my sincere gratefulness.

I assure friends that the lesson here learned, it was much larger than you can imagine

Speak-king on them later

method

The tests are still made in the same rustic way and they can contain mistakes, they are not definitive in hypothesis none

I caught two reservoirs of 30 liters, I made 10 demarcations of 2 in 2 liters in one of them
4 passages for each presented measurement

data of the tested equipments

PUMP

quiet one model 40R175254V

MAX-FLOW (GPH) 1140
horse power 1/25 HP
WATTS 87w
SHUT-OFF PRESSURE 6.5 psi
Shut Off Heat Ft 15,5

observations as the connections of the pump and tubes used in the test of flow rate of the pump

the entrance connections and exit of the pump has internal diameter of 9/16 and the tubes that I use has internal diameter of 1\2 inch

with 1 meter of tubes in the entrance and 1 meter of tubes in the exit the flow rate presented by the bomb + tubes it went of 37 seconds to move 20 liters / 32 liters per minute


PUMP + RAD F2

with 1 meter of the twith 1 meter of the tubes in the entrance of the bomb and 50 cm in the exit ties y of entrance of the rad, more 50 cm until the exit , the flow rate presented by the bomb+rad + tubes it went of 46 seconds to move 20 liters / 26 liters per minute

PUMP + RAD F2 + LRWW

with 1 meter of the tubes in the entrance of the bomb + 50 cm in the exit ties y of entrance of the rad, + 50 cm until the exit + LRWW + 50cm tubes , the flow rate presented by the bomb+rad + lrww + tubes it went of 93 seconds to move 20 liters / 12,9 liters per minute

PUMP + RAD F2 + CASCADE

with 1 meter of the tubes in the entrance of the bomb + 50 cm in the exit ties y of entrance of the rad, + 50 cm until the exit + CASCADE + 50cm tubes , the flow rate presented by the bomb+rad + CASCADE + tubes it went of 140 seconds to move 20 liters / 8,5 liters per minute


observation, all the used connections, less the one of the block's that if you maintain original they are re-worked and your internal diameter is of 7\16 ( the one of the bomb were already mentioned and the tubes also)

for today it is alone that, I thank all, it is already late in Brazil, 4 hours of the dawn, tomorrow comment some things that I learned and another that I observed in the system and I place here for us to appreciate OR NOT:)

good night

today it is the first day with the cascade, him I think an exelente waterblock, as there was change of thermal paste of shin-etsumicrosi for as3 and modified some connections, although the results are very similar I cannot foresee it had only won or loss, in all the aspects him this very similar one in behavior with the lrww, tomorrow I will set up the lrww again with as3 to evaluate the differences

Cathar congratulations for the cascade, I think an exelent wb

A consideration only, who will migrate for a cascade, I advise the change of the fluid and a good one filtered in the same, because any dirt obstructed with easiness the pins injetors

The lrww was already easy to obstruct the micro chanel, with the cascade it is I criticize any particle in the fluid

A good way and to leave two filters done with woman's fine stockings in the tips of the hoses and arrested with elastic and inside of a recipient with a little more of fluid, some hours rotating and I think will already be ready to clamp the cascade.

new systems request attention, because every impurity type can have inside of the components, result of the production or storage of the components.

Did anybody with p4 make that same migration of the lrww for the cascade? could it happen your impressions on the change?

thank you very much friends

I forgot to mention...


New cores with dual fan soon, inside of some days new rads F2 in the sizes of it cores of
6,5"x 6 (48 flat tubes) and 5 " x 6 " (40 flat tubes) using the same technology and materials :cool:


Thank you very much to all;)


Ivo Guilhon (Copyman)

Yes. I have both the cascade and the WW and use a P4. I have seen no difference between either block in terms of temps, seems identical. I think, if one were using an AMD CPU, the small difference may be detected, but it seems not on my P4 anyway.

As Scott mentioned above, he saw no reported difference for P4's.

Two other P4 users have contacted me via email. One said he saw around a 1C difference between the White Water and the Cascade, and the other a 2C difference.

This hardly suprises me though when it comes to P4's. I've had one person say they saw no reported difference between a Maze 3 and a White Water as well, and then I've had people tell me they saw a 7C difference.

AMD system's temperature differences always seem to correlate pretty well based on CPU type/speed/voltage, but for P4's there appears to be a significant proportion of "stubborn" CPU's when it comes to temperature reporting, and I know this is something that has been noticed by water block designers other than myself.

that good to know of that, i had a light advantage for the lrww, but I have result better in the p4 using shin-etsu, it believed that they would be very similar or same, the blocks of the cathar give a new dimension to the wc system, because before him, it was very hard to get to extract the heat of the processors and to review to the fluid.

The lrww and the cascade are true monsters sucking heat, it remains we learn the best more use them.

I should place other in the amd soon, has more 3 here:D

I understand cathar perfectly, the p4 are hard to extract temperature measures, I test many processors in my wc, and I can assure that I never had the luck of having measured of similar temperatures, I don't get to do small it adjusts them in the rads and in my system in general in the p4, I have that make them in the amd where the temperature measurement is something more solid.

now as or maze 3 compared with the lrww, for god's love, I like den a lot, wonderful person, but the blocks of cpu of him are of the time of doing overclock in p3, 7 degrees of difference dissipating little heat in the processor, I with a rev3 compared to him, in an amd xp1600+ @ 2047mhz 2,3v palomino, to 1900mhz 2.0v had more than 10c with the maze3, breaking above that, while the rev3 took me to 2047 mhz , logical, that depends on the rest of the system...

I wonder if it would not be possible in the cascate or lrww, you optimize them a little more to the format incapsulated of the die p4, increasing the efective thermal change area of them

I believe that the differences of block change should be larger depending on the effectiveness of the remaining of the system, perhaps the cascade provides better temperatures in more cheaper systems, and that is something extremely desirable.

congratulations for the work, I loved the cascade

To optimise more for the P4's will result in less performance for bare-die CPU's.

I figured it like this. All the really hard-core P4 overclocker's pop the heat-spreader off the P4 anyway, which essentially turns a P4 into a bare-die CPU again.

I guess it's just a matter of targetting what matters most when it comes to balancing performance across a range of applications. There is a slight preferential focus to bare-die CPU's, but I was careful to ensure that for IHS-based CPU's the Cascade also gives very good performance.

The IHS really is a cooling obstacle. I can work with it slightly better (maybe 1C), but to the detriment of bare-die cooling performance, and doing so will also increase the cost of the waterblock as there will be extra machining time.

One thing I found of interest was the flow rate differance between the Cascade and the White Water in Copyman's testing.

The White Water showed almost 1/2 again more flow over the Cascade, or the Cascade was 1/3 more restrictive depending on how you look at the numbers.

This could make a noticable differance to systems with more than one block in series. For those systems the White Water would seem to be ahead of the Cascade.

One possible reason for the large differance, blockage of the tiny jets of the Cascade. Copyman, did you check your Cascade very carefully to make sure none of the jets are pluged? Even a few blocked jets would eleminated the Cascade's performance advantge and would also lower flow noticably.

Sorry to backtrack...

Cathar, I'm very interested in how quiet your blower is... I'm going to go WC but I'm having great difficulty choosing the Rad/fan combo. I don't really have space restrictions as the Rad/Pump/Res will be out of the case. I was originally looking for a heater core about the size of 2x120mm fans.

My real objective is to keep the noise low, so I assumed two slower 120mm fans would be pretty quiet. If your blower is quiet I would consider moving to a smaller core and raising the airflow (by looking at all of the specs shown here, raising airflow has a much bigger impact than everything else including water flow).

I'm curious if the noise from the blower is motor or air noise? If it's the motor, I'm guessing it could be damped further by enclosure?

Unfortunately, I also can't decide on the heater core since there is very little data on air flow capacity posted anywhere...

http://forums.procooling.com/vbb/sho...&threadid=7314

I'm hoping to get an automotive blower (running quietly at 5 volt), but a motorized impeller would be ideal.

You can silence the motor, but the air flow noise, you'd have to fix by adding some kind of turbulators, something to disrupt the airflow. If that doesn't work, you'll have to try another airflow rate: the whistling noise either comes from the air hitting the core, or from within the core.

For some airflow data, check out Bill's radiator roundup on overclockers.

Yes, I've read through Bill's posts extensively, which is why I want a rad with low static back pressure . As far as I can tell, he only has results for the 2 unnamed heater cores in the batch.

Unfortunately, that doesn't help me when I select my RAD of preference based on size.

my first blower

http://www.watercooler.com.br/f2/bl.jpg


but ugly camera

yes I disassembled it several times, if you render attention I make an alert one on that problem of blockage of the injectors, after the tests when I raised my rig and I dismounted the cascade again and there was not any obstructed pin


The data are totally with the two block's clean

[]'s

I'm not sure for the reason of the difference that copyman is seeing.

In my system I have 2 heater-cores in series, a sponge filter, two sweeped elbows, and about 4m of 1/2" ID tubing.

With the Iwaki MD-30RZ in my system the difference is 10.5-11.0lpm with the White Water and 10.0lpm with the Cascade.

With the Eheim 1250 I see 6.0lpm with the White Water, and 5.5lpm with the Cascade.

I'm using a regular Cascade as it would ship to people. No difference whatsoever.

The earlier Cascade prototypes had less tubes and larger jets, and were less restrictive than the White Water. With the shipping block that has changed slightly, but I kept a careful eye on the flow rate resistance so that this would not be an issue.

My blower is quite large. 33x15x15cm in size. It looks much like Copyman's above.

http://www.employees.org/~slf/images/fannage.jpg


It's installed in a large radiator box that helps to muffle the sound of it. The blower pretty much makes two noise types, that being the rushing air and the brushes in the DC motor. Incidentally the brushes are about 1/2" long rectangular blocks that are spring loaded. As they wear away the springs just keep pushing the brushes down. It'd take a long time for the brushes to wear away completely. I had to pull the blower apart and clean the brushes and contacts as the blower had been sitting on a shelf for 5 years and the oxidation of the copper over that time had left a thick black coating over the contacts such that the brushes were barely making any electrical contact. The noise of the brushes is a medium pitched low volume "whirring" sound, which gets completely muffled by the radiator box. The box however does amplify the inherent vibration of the blower and there is a very low pitched drone which changes in volume depending on the resonating pitch with the box's natural vibration frequency. Changing the voltage supplied to the fan by even a tiny amount is enough to almost totally nullify the effect of that noise.

The radiator box looks like this:

http://www.employees.org/~slf/rbox/radbox3.jpg

Except the back panel with the axial fans has been replaced by the blower. The blower takes up almost half of the back panel's area by itself.

I power the blower with a DC PSU which I can set to whatever voltage I choose, and it also allows me to keep an eye on its power draw. I typically have it set to draw about 25W, although I can put the blower to full speed (16V on my PSU) at which time it'll draw close to 250W of power and push around 750CFM of air-flow (in free-air rated mode). At the standard rated 13.6v it's rated for 650CFM of free-air flow. At the 25W setting I'd estimate it's pushing around a real 100CFM through the heater-cores, although I don't have anything to measure that with.

WOW, that's great. Interesting though, I assume with the blowers, the air going through the RADS is created by the suction in the box since I assume the blowers blow out the holes.

It looks like a car blower, but the voltages seem a bit off...

Another quick question... Why two heater cores instead of one double sized one? BTW, Which ones are those?

Interesting cathar, will take advantage of that will test other bomb of 1200gph 19 " shut off in the weekend and I will measure again, I still have here 3 cascades to play.

It is possible that it was obstructed in the test, but it was not more or final because I checked before

I am sad for not having to who to send my rads and your block cascade to evaluate at a laboratory that can supply necessary data.

I am with a ready complete group, rad, bomb, lrww, cascade and tubes for sending.

but for who?

bill doesn't execute more the tests, joe not this ready one executes it them, I am the wait so that my equipment can be tested in a complete kit with the cascade and lrww, and only the rad.

what sent for you it doesn't have that purpose, just for you to see if it likes... F2 Extreme it should take two weeks to arrive for you

As soon as possible I will send the models dual fan if you want

[]'s