Serial loop Vs. Parallel loop

Page 3 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
toolmaker_03

toolmaker_03

Honorable
Mar 26, 2012
2,650
0
12,960
a detailed parts list, of the two water loops used in this test setup.
2 MCR320-QP 360mm radiators
http://www.swiftech.com/MCRx20-QP.aspx
2 EVGA GeForce GTX 580 FTW Hydro Copper 2
https://www.evga.com/articles/00592/
1 XSPC raystorm CPU block
http://www.xs-pc.com/waterblocks-cpu/raystorm-cpu-water...
1 swiftech MPC655 pump
http://www.swiftech.com/mcp655.aspx
6 corsair SP fans
http://www.corsair.com/en-us/air-series-sp120-high-perf...
6 bits power 3 way 90's
http://www.frozencpu.com/products/12084/ex-tub-744/Bits...
1 Y block
http://bitspowerwork.com/html/product/pro_show.php?prod...
1 250mm reservoir
http://bitspowerwork.com/html/product/pro_show.php?prod...
8 3/8ID 1/2OD compression fittings
http://bitspowerwork.com/html/product/pro_show.php?prod...
2 T's
http://www.xs-pc.com/hose-fittings/g14-t-fitting-black-...
5 male to male fittings of varying length
http://www.xs-pc.com/hose-fittings/g14-5mm-male-to-male...
4 female to female fittings of varying length
http://www.xs-pc.com/hose-fittings/g14-18mm-female-to-f...
2 koolance flow meters with display
http://koolance.com/ins-fm19-coolant-flow-meter
http://koolance.com/dcb-fm01-flow-meter-adapter-with-di...
1 loop filter
http://koolance.com/ins-fltr03-inline-coolant-filter
2 temp sensors
http://www.frozencpu.com/products/10373/ex-tub-620/Bits...
1 alphacool bubble down
http://www.frozencpu.com/products/17333/ex-tub-1558/Alphacool_Bubble_Down_G14_Thread_Reservoir_Flow_Regulator_-_Deep_Black.html?tl=g30c97s169
1 bitspower aqua pipe
http://www.frozencpu.com/products/11863/ex-tub-713/Bitspower_G_14_Silver_Shining_Aqua-Pipe_I_BP-WTP-C17.html?tl=g30c97s169

This system is overclocked, on both the serial and parallel test setups.
CPU overclock @ 4Ghz
GPU1 overclock @ 850Mhz
GPU2 overclock @ 850Mhz
Memory overclock @ 1600Mhz
CPU stock @ 3.2Ghz
GPU1 stock @ 772Mhz
GPU2 stock @ 772Mhz
Memory stock @ 1333Mhz
this is the serial loop setup and what it looks like, and here are the hardware temps at idol, ambient temp of the room is at 28.6C
CPU 43C
GPU1 37C
GPU2 37C
the water temps at idol are
29C-30C
the hardware temps at load are
CPU 61C
GPU1 51C
GPU2 51C
the water temps at load are
37C-38C
the delta t of the system at load is 10C
http://imgur.com/Gi4wFZO,G1Mte0g,Pb8ot4j,k8LlLK5
Gi4wFZO.jpg

G1Mte0g.jpg

this is the parallel loop setup and what it looks like, and here are the hardware temps at idol, ambient temp of the room is at 28.6C
CPU 43C
GPU1 35C
GPU2 35C
the water temps at idol are
29C-30C
the hardware temps at load are
CPU 57C
GPU1 46C
GPU2 46C
the water temps at load are
34C-35C
the delta t of the system at load is 7C
http://imgur.com/pBZY5ne,yIT8rwA,CgNlfyH
yIT8rwA.jpg

pBZY5ne.jpg

this is the difference between the hardware used on the serial loop, and the hardware that needed to be added to the parallel loop.

http://imgur.com/U3yYYJl
U3yYYJl.jpg








 
Sort by date Sort by votes
ok so it is your belief that all the flow meters made by koolance are not calibrated by the company correctly, that is your opinion you are entitled to it.
so take that up with koolance.

I can go online and look at others that have the same flow meters I do and when hooked to similar hardware they have the same reading I do so I have no reason to doubt the accuracy of my flow meters.
 
Upvote 0 Downvote
That's great, but have you validated by another means to benchmark the flow meters? Just because they are all made by Koolance and read the same, they all could be wrong in the same way. The first thing I did when I got my flow meter was to test accuracy at 0.5 GPM, 1.0 GPM and 1.5 GPM. It is exactly correct.

Haveyouvalidated these are correct?
Click to expand...



You avoided my question. It was not about Koolance and them validating it. It was about you and your validation.
 
Upvote 0 Downvote
look, if I where getting a different reading on even one of my flow meters, I would test all of them.
the same can be said for you have you pulled the thermostat out of your A/C to see how accurate it is.
when you buy a new car do you pull the temp sensor out of it to see how accurate it is.
no one does, because its not necessary, unless you have a reason to doubt the accuracy of the device in the first place.
 
Upvote 0 Downvote
One of the biggest steps in recording and validating data is evaluating and benchmarking equipment to ensure you are getting expected readings from them.

I'm not using my A/C or car to test and evaluate data that I am presenting to a forum community, am I? So, there is no reason to do so. A car's cooling system and A/C thermostat are not testing equipment in the watercooling hobby; flow meters and temp probes are.

If I was a MLB pitching scout and had 3 radar guns to evaluate the pitching velocity of a prospect, I'd have to get each of them tested and calibrated. If they all 3 read incorrectly, but all 3 read the same, this does me no good to report that the kid throws 97mph (based on 3 inaccurate readings) but actually throws 93mph on a correctly calibrated and tested gun. What is the benefit of having 3 devices give you the same, incorrect reading if you haven't tested to know for certain? The fact they are all made by the same company means they all could be calibrated the same, but also means they all 3 could be reporting the same, incorrect readings.

There's no room for assumptions when the data you are presenting is intended to be factual and conclusive when your testing equipment having not being tested and calibrated itself is cause for failed validation.
 
Upvote 0 Downvote
I don't see how that is irrational by any means. It is standard procedure for any scientific methodology that has to do with reporting the data output of equipment.

You don't have to do anything, I was simply offering the standpoint that you have yet to do this set of steps.

Also please stop with the all caps typing; it is against forum Terms of Service rules.
 
Upvote 0 Downvote
@ toolmaker_03

If you were more open to suggestions and not thinking the worse from the questions you are getting, you may be able to refine this thread and make it useful presentable information.

But you need to realize this is an option of water cooling, not a necessity, especially when it comes down to the minimum amount of money a person wants to spend just to get into water cooling.

Just because we may disagree with something you've said, does not mean we are attacking you and saying everything you have done is invalid.

IMO there are simple things you need to do for presentation purposes:

#1 Reduce the clutter in your pictures.
#2 Post reasonable sized pictures that do not blow out the forum posting structure.
#3 Clean the dust from those radiators and repost the pictures.
#4 Outline your presentation with a plot of title and goal, don't start with a half page of parts listing.
#5 Be more descriptive regarding what your take is on Series Loop vs Parallel Loop. (Because some of what you show is actually Series/Parallel)
#6 Be more receptive and treat others with respect and you will get that back in return.
#7 Test all parameters and then end with a conclusion.
#8 Not everyone is going to agree with your conclusions even after all your proof is displayed, that is life, even if you are 100% correct.

Some of us will help you as long as you are reasonable, cussing us out in PMs does not put you at the top of our want to help you list.

Which leaves you all alone when there are years of water cooling experience here for you to take advantage of.

 
Upvote 0 Downvote
There ya go. It's good to validate across a wide range because like you found out, variances in tolerance do exist from one end of the flow curve to another with the same measurement tool.

Easiest way to test flow rate is a stop watch and a container of known volume. Close enough for everyone but astronauts and submariners.
 
Upvote 0 Downvote
ok so I went back and did a really in depth look at all the systems that I have ever water cooled, and this is a break down of them, in order from oldest to newest.
the first H2OQPower system was to cool a super socket 7 CPU only at 550Mhz.
the second H2OQPower system was a to cool K7 CPU at 1.7Ghz to start and later a Ti4600 video card with block, I also added a 240mm radiator to the system.
then I up graded the first system to a K8 CPU at 2.6Ghz, a 6800GT video card, with two HHD's blocks, and I added memory blocks to the loop, I also added a 240mm radiator, then I upgraded the CPU block to a MPW50 CPU block, for this build.
then I up graded the second system to a intel CPU 4.3Ghz extreme core, and a ATi all in wonder video card with block, and I upgraded the CPU block to a MPW50 CPU block for this system.
then I built a third system to cool a AM2 CPU at 3Ghz, a 8800GT ultra video card with block, and I added the memory blocks to that system and it has 2 X 240mm radiators, I added a reservoir, a MPW50 CPU block, and a pump for the loop.
then I up graded the first system to a AM3 CPU at 4Ghz, and a 480GTX video card with block, I changed the 2 X 240mm radiators to a monstra 360mm radiator, the MPW50 CPU block for a Raystorm CPU block and added a reservoir and a pump to the system.
then I up graded the second system to a intel 2011 CPU at 3.2Ghz, and two 580GTX video cards with blocks, I changed the 2 X 240mm radiators out for two swiftech 360mm radiators, the MPW50 CPU block for a Raystorm CPU block and I added a reservoir and a pump to the system. I am still working on this system, I want to have 3 radiators and 4 video cards with blocks in total for this system.

well I have been water cooling for some time now, and I did do a simple test like this about 2 years ago, that is how I figured out that the out put of the flow meter to my fan controller was in LPM, I did not post it as it was for my understanding of what I was looking at every day, and when I purchased the display made for the flow meter by Koolance it further showed me that it reads in LPM, but your right, I did not do a full test to see exactly how accurate it is all the way up the scale, I only ran the first test 2 years ago at 5LPM.
 
Upvote 0 Downvote
I am going to have to side with toolmaker_03 over rubix_1011 on this argument. Rubix_1011 stated below that when you split the tubing that yes you cut the flow rate
of one branch of the Y or T connector by half, but since you have two tubes now going at half the starting rate, both of them together equal the starting rate, in other
words no combined loss of flow. Also remember that you are filling two or more radiators at the same time instead of just one. Remember that moving water thru
the radiator that you will meet resistance.The higher the flow the more the resistance. By reducing the flow rate, by using a parallel set up, you lower the resistance
you encounter in the overall flow rate. So, in a way, you both are right. Now check this out. Pump #1 > CPU, GPU #1, GPU #2 or more inlet ports in parallel > CPU,
GPU# 1, GPU #2 outlet ports in parallel > reservoir #2 > Pump #2 > Parallel to radiators inlet > radiators outlet to parallel > reservoir #1 > Pump #1, repeat. This is
the set up the human body mostly uses. Think of the pumps as each side of the heart, radiators are the lungs( heat exchange instead of oxygen for CO2), and the
targets for cooling such as CPU, GPU's, and others are the rest of the body such as the brain, liver, kidneys needing oxygen. Different organs need different flow
rates to survive just like different components have different resistance or flow rates. When you do a serial set up, do you go by flow rate when you select the
sequence of your set up(higher flow rate first)? We all know it doesn't matter on serial set ups but we still want to answer the question so we satisfy this need by
converting it into choosing by aesthetics or how it looks.
Parallel does that for you by giving a higher flow rate over all and that gets the more coolant over unit of time and better heat exchange. That is by design. If
you want more flow to a certain component then buy the water block with the better flow rate. In serial, the flow rate is determined by the slowest component
not by need. Parallel goes by the best for each component based on flow rate or resistance. Now if you are always asking yourself which comes first in a set
up such as radiator, CPU, GPU then if you use a pump before and after the parallel radiators then everything comes first.
Plus parallel can look cleaner than series, so there!
Toolmaker_03 had a great idea, but I plan on improving on it. He doesn't have a reservoir on pump #2. You can't tell if they are in balance that way. When
you run one pump at a lower rate then you put extra ware and tear on the other pump motor and it wears out quicker and you won't know which one until it fails
completely and quickly ruins the other pump (untreated carcinogenic shock > DEATH). With my method the pumps last longer and you get one more thing to
tinker with and that is catnip to us enthusiasts.
To summarize:
1. Cool every thing with the same starting lower temperature water from parallel radiators. (Serial set up causes the temp to go up after each component you cool)
2. Will not loose pressure before or after the radiators better overall flow rate. (You don't have to choose what comes first after the pump... CPU, GPU, radiators - they all do!)
3. Every cooling target would benefit from all radiators together no matter how many you have. (with better cooling related to higher overall flow rate)
4. Less stress on the pumps
5. There could be some redundancy or early indication if one pump would start to fail or out of balance.
6. If one unit in the parallel set gets a blockage in the flow, the ones left over get more flow (serial > a blockage stops all flow) flow meters should be on branches not the trunk.
this causes a greater variation and makes it easier to see a full or partial blockage. Temperature changes from baseline can also be a sign of blockage.

Now if pump #1 and #2 go at different rates then you would have different levels in your 2 reservoirs, an indication of imbalance. In the body that is known as systemic and
pulmonary edema depending on which side of the heart is failing, or in this case, if one of the pumps has a faster or slower rate than the other. That can be caused by pump
failure or improper balancing. This can be fixed. You can re-balance the pumps if you are there or put a tube connecting the tops of the two reservoirs if you aren't watching
to temporarily compensate the imbalance until you are there (which will cause a shunting effect if left untreated thereby decreasing the efficiency of the system). Once you
fix the balance between the two pumps you have to make both reservoirs have equal coolant levels to resume monitoring.That can be achieved easily by connecting a tube
at the base of both reservoirs separated by a ball valve.(you need a valve to do this when needed or warm coolant will constantly mix cooler coolant without a valve or in
other words, create another shunt) Weather you believe in God or evolution, the body is designed well and we could learn something from it.
Tell me I'm wrong... I dare you.... 😛
 
Upvote 0 Downvote
@ y2kmedman

Welcome to Toms Hardware, I find it unusual your very first post here is to directly confront rubix_1011, by comparing water cooling to human anatomy you leave out the brains responsibility in regulating everything, so really there is no direct comparison between the miracle of the human body vs static run traditional computer water cooling.

All this series vs parallel argument is really a moot point, seeing as how if you had enough radiator cooling field to do the job, series parallel would not matter at all, toolmaker_03 is the one that felt this information needed to be force fed to us, because we were failing to grasp the concept.

Do you remember MC Hammer's song "You Can't Touch This", well that applies to serial vs parallel vs my chilled water cooling! (Now approaching 3 years in operation!) :)

But Welcome to Toms anyway and good luck with rubix_1011, but presently I think he is kinda busy reviewing the EK Predator 240 AIO cooler, he is kinda our traditional water cooling GURU here. :)
 
Upvote 0 Downvote
You may want to keep in mind because of manufacturing variations no 2 even identical radiators, or water blocks for that matter flow exactly the same, because of manufacturing variances, like too much solder partially blocking a flow tube, or a bent cooling fin in a water block, which can, and will affect flow, and in series parallel the higher flow volume takes the path of least resistance, thus creating a flow imbalance.

That sir is a fact of water cooling life! :)
 
Upvote 0 Downvote
Regardless of whether you consider a loop parallel, it is still serial at some point, otherwise you have 2 loops - UNLESS you run shared reservoirs; and even then, you simply have 2 serial loops to some degree.

Yes, lower flow also lowers restriction, we all know this.

The reason a Y-split doesn't lower the overall flow of the loop is because the serial line prior to the split then halves, running in parallel. But, at some point, those parallel lines Y-split back into a serial line - again (therefore, restoring the original, serial line) but, everything parallel is still serial in the end unless you meet the criteria where the only place the 2 loops interface is a shared reservoir. You also have to remember that any watercooling loop is just that - a loop. Pushing through a block is done just as much as pulling through a block. Filing a radiator is the same thing as emptying it. This is all because your loop is under constant flow equilibrium; which it operates at. If this fluctuated, it would be impossible to benchmark or determine performance of any watercooling component.

Also, remember that while lowering flow rate does lower restriction, if you look at every radiator performance curve, it also lowers cooling performance. Flow rate is a function of DT, so if you lower flow, you have to compensate either A) higher fan speed/airflow, or B) increase your radiator field. Think of it as a simple math equation: A+B+C = D. If you change any of the variables, in order for the equation to work out, you have to adjust the remaining variables to compensate.

I think the point that is being missed is that you're right; but so am I. This all depends on what your debate hinges upon. This discussion on flow rates, restriction, serial vs. parallel has been going on for decades.

 
Upvote 0 Downvote
I kinda question rubix_1011 being called a guru in that his rebuttal to toolmaker_03 stating that there was a loss in liters per min by half once
you split the line in two. He was addressing only one end of the Y connection and rubix_1011 represented it as the total of both ends L/min
loss due to splitting. I took a look at your set up. Now that is what I call cool. Pardon the pun. Elaborate but cool. When you talk about the
brain, the CPU comes to mind. It dies the quickest without oxygen much like the CPU with an inability to exchange heat. Although GPUs
can put off serious heat too. The body uses parallel circulation and when the mean arterial pressure drops below 50 mmHg the brain starts
to have problems. I am sure you can tell that I am new to this but I question everything I can so I learn more. When someone argues that
results need to be thrown out soley b/c the measuring tool might be off I am suspect. Take a common weight scale that is off by ten pounds
or even 10%. When a morbidly obese adult steps on it and a 10 year old does. You can tell which one it is based on the disparity. 2.4 vs 6.8L/min
assuming that the variation is constant and not random, it deserves further research and I intend to test that hypothesis. I am going to pay
close attention to detail and follow the advise you gave toolmaker_03 and the control will have little or no variation in the variables other than
series vs parallel. I plan to present my results and defend them. I just hope my need to be right doesn't cloud my objectivity which I think
happened to both of them.
 
Upvote 0 Downvote
Even though the flow rate is lowered by half, there are two radiators filling and draining at that rate.
The serial rate stays at the same rate based on the most and total restrictive flow. Parallel slows
down when splitting but more volume overall passes thru the component and speeds up to travel
to the next component when the tubes come together. That gives me an idea, keep the source
branches close together compared to the trunk to improve overall flow rate... Thanks!
 
Upvote 0 Downvote
Parallel slows
down when splitting but more volume overall passes thru the component and speeds up to travel
to the next component when the tubes come together.
Click to expand...

How can flow slow down (which is a function of volume and rate) and but also mean that more volume is present?

If you have 2 radiators in serial or 2 radiators in parallel, you still have 2 full radiators, just the flow rate of the total volume is different.
 
Upvote 0 Downvote


That's true, however the CPUs relation to water cooling is normally completely independent with most of the hardware independently controlled without the CPU controlling anything,
whereas comparing the brains relation to cooling the body is just one part of it's massive responsibilities to keeping the body alive.

Some water coolers use cooling that are to some extent motherboard controlled, but the mass majority use the power supply directly for pump and cooling fans.

A completely true parallel system is actually 2 complete independent loops as rubix said, or else it is series/parallel.

Any time the parallel loops merge it is no longer a true parallel system.

 
Upvote 0 Downvote


It is a term of honor I give him seeing all he has done for Toms Hardware over the years, it is not a tag he has claimed for himself!

He is very knowledgeable and you could learn quite a bit from him as he is a traditional ambient water cooler.

My expertise comes from experimentation in below ambient cooling, though my Graphics are traditionally radiator water cooled.

 
Upvote 0 Downvote
A completely true parallel system is actually 2 complete independent loops as rubix said, or else it is series/parallel.
Click to expand...

In this case, you still have 2 serial loops at the end of the day. The concept of 'parallel' in a watercooling loop is only as specific to the area in which split/merge exists. Outside of that, overall it is still serial. If you want to get really scientific about it; it's pretty much impossible to have a truly, complete parallel loop that is closed. This would only be possible if you had a scenario like a large bucket as a reservoir feeding a pump and then open-ended draining back into the bucket, and technically this is still serial. House plumbing feeding a loop and then draining down the sink would really be the only TRUE way.
 
Upvote 0 Downvote
Think of it this way. When you carry groceries in from the car, do you carry one out two bags for two trips or two out of two bags for one trip? If you choose to carry both at once you would go slower than if you carried just one, but you would get the job done faster, right? Since you lowered the resistance by lowering the speed by splitting the line, the over all time it takes to finish the job is lower. This is easy to test. Get one pump, 2 identical lengths of tubing, and 2 or 3 radiators and test one set up at a time. Measure the time it takes to for the water to reach the end of the line. Make sure everything but set up is the same for each set up such as the slope of the final run of tubing. Tell me what your finding are. I will be able to do this test in a month or so myself.
 
Upvote 0 Downvote
to rubix_1011:
I am sure we can both agree, we don't know each other. I am sure you have extensive knowledge and
experience that exceeds my own in this over all area and the reverse is true of me in other limited amount of subjects. With that being said, sometimes people with lesser or different knowledge than your own can come up with or inspire insights of your own that improves our overall general knowledge base. Sometimes you need to look past who is giving you the info and how they are doing it and objectively see the value of the insight that can be gained. Toolmaker_03 stumbled upon a possible truth that is counter-intuitive or goes against common sense and that is possibly a very valuable and rare thing and he felt overwhelmed with its possible implications and tried to, "force feed newbies" and you responded how I would probably had as first response. Being that I felt I was on the same path as Toolmaker_03 and came to a similar realization of the possible small benefits of using a totally parallel set up, I was researching and will continue researching and experimenting on this. I feel that my drive is in part driven by the fact that I am a disabled RN and miss the challenge of treating sick people and I want to bring my knowledge from that field to benefit me and possibly others. When I noticed some similarity in water cooling to the human body, I jumped on it like white on rice. It felt good to be able to predict the problems with a total parallel set up that mimicked known problems that I have been taught to respond to. I hope that you and I can have a positive interaction in the future and I hope to benefit from your knowledge. Thank you for your time
-Sam
 
Upvote 0 Downvote
I think your view of this thread is what you think actually exist for all members of the Tom's community in the sense of watercooling. However, there is quite a deep history between toolmaker_03, 4ryan6 and I all together. The culmination of this thread is the result of a few years of interactions and hundreds of threads in which we have each contributed. So, while you are reading this instance, please note that there is a substantial backstory to the entire relationship. That being said, there have been some heated back-and-forths between us all. For example, I know for a fact that if there is anyone on these forums that will challenge what I write, it's 4ryan6. I respect his work, his research and data presentation that he has done for his forum community. Likewise, he has often said the same about me, for which I am grateful.

So, with that being said, I do my very best each day to openly attempt at teaching and helping those new to the hobby. Am I a wordly expert? No. There are guys over at OCN and XS that make me seem like an everyday guy. Have I learned a lot and done what I can to learn, in order to help others learn? Yes, very much so.

There are no bullies around here, even if there is a belief otherwise. I started watercooling 13 years ago and had absolutely no idea what I was doing...even for years this was true. However, I will stand firm on what I know exists as well-known watercooling knowledge as proven by people way smarter than I. It is why I wrote the sticky and put together the radiator estimation sheet.

Everyone is welcome here, we just ask that we keep it friendly, professional, fact-oriented and open minded. I love a good debate - but I will not tolerate name calling or dancing around facts in order to counter with retaliatory data.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom