Introducing my new watercooling design concept!

The alternative to direct air cooling

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Sun May 09, 2004 6:29 am

thanks for your reply Michael_qrt, always a pleasure to listen to a fellow sydney-sider :) (assuming you're from the australian variant of sydney)

portability will actually be very easy with the setup i'm planning. i'll be using quick-disconnect couplings which means that you can just disconnect the tubing from the unit (or computer) by hand, and off you go. another thing to carry, and maybe a few drops of leaked coolant upon disconnection, but that's it.

i too would like to see a good dedicated wc+silent case. if this project turns out to be a success, i will definitely pursue this path. it is a great deal more work though, and is of debatable use given the uprising of SFF and in the future, BTX cases. this external box solution i'm working on allows compatibility with both of those.

waterblock design is a whole different arena, but i have looked into it and i have some jet impingement designs drawn up in CAD software already. i'll see how things go. the radiator will just be bought probably, i am extremely short on the expertise required to build one of those :)

thanks for your good wishes Michael_qrt, i like reading long posts and yours was indeed helpful!

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Sun May 09, 2004 6:34 am

hi eniacs,

i definitely aim to sell the system for under £250. that is already up around the price range where phase change cooling becomes appealing :)

believe me, my mum is pretty illiterate when it comes to computers :) perhaps the "mum test" is too harsh - perhaps the "girlfriend test" will suffice.

wood is an attractive option as it has very nice sound dampening qualities, but i haven't looked into it very much and i'm worried about warping issues etc.

thanks for the boost in confidence regarding the electronics, my friend whom i was counting on to provide that expertise has been crushed by a truckload of uni assignments and so any advice is most welcome. would you be able to provide a diagram or a link to some useful resources regarding that circuit?

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Sun May 09, 2004 6:43 am

just thought i'd post an update with regard to this project.

it's coming along very well, i've got most of the airflow management sorted out and the layout of parts etc. the structure has almost been finalised, as has the base plate. cover and aesthetics will follow shortly.

due to space constraints, i'll be using 2 80mm fans and 1 120mm fan. the pump will be an eheim 1048. all fans and the pump will be decoupled.

for fans i'm thinking 2 panaflo 80L's + 1 papst 4412fgl, which should provide adequate airflow.

i've decided to skip electronic flow rate detection as it adds too much restriction to the loop, and quite simply isn't necessary unless the user deliberately clamps a hose shut or kinks it beyond recognition. a visual system will be used, just as the zalman reserator flow rate monitoring system is a visual one, although mine differs significantly.

the dimensions are looking to be around 360mm long x 220mm wide x 180mm tall. so it's not small, but it'll fit nicely on top of a tower case.

that's about all for now, i've been doing a lot of work with respect to the internal structure and am now moving on to the outsides, now that i know the insides will stay together and not fall over itself :)

thanks for all your support to everyone that has replied so far, it has helped tremendously.

Michael_qrt
Posts: 154
Joined: Thu Oct 17, 2002 5:53 pm
Location: Sydney

Post by Michael_qrt » Sun May 09, 2004 5:18 pm

Yep, the aussie version of Sydney.

Anyway it sounds like you've really got this coming together. I guess it makes sense to just buy the heatercores, at least at first. though maybe for V2 you could come up with a custom design. As for waterblock design jet impingement designs are used in the highest performance water blocks, the real trick will be to make them cheap and easy to manufacture.

Anyway, I hate to go back to the whole case topic again but i wonder if it would be feasable to run a WC setup which uses huge heatsinks on the case sides as the radiator. Not a performance design but a potentially very quiet one and competition to that zalman heatpipe case except it'd be much more flexible, kind of like a reserator built into the case.

Just another question, are you planning on having manual fan control or temp based fan control or both?

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Sun May 09, 2004 6:08 pm

woohoo go sydney :)

innovatek has something similar to what you describe, they're essentially about the size of a case panel and bolt right on. passive cooling isn't my forté at the moment but i'll look into it after i've got this project done.

fan control will be manual. i'll set it so that the minimum setting will be 5v, which should provide enough airflow to prevent things from overheating, so there won't be any need for temp control to handle dangerous situations. it'll essentially be a 'volume' control :)

Michael_qrt
Posts: 154
Joined: Thu Oct 17, 2002 5:53 pm
Location: Sydney

Post by Michael_qrt » Sun May 09, 2004 7:31 pm

chylld wrote:fan control will be manual. i'll set it so that the minimum setting will be 5v, which should provide enough airflow to prevent things from overheating, so there won't be any need for temp control to handle dangerous situations. it'll essentially be a 'volume' control :)
That's a fair comment, and I'd prefer manual anyway.

What do you have lined up in terms of acoustically damping the enclosure, do you think it'd be worth it to make a "hushbox" type enclosure for the pump or will the 1048 be quieter that the fans even on their lowest setting with your decouple mounting? I don't have any experience with this.

And yeah, go Sydney. though it'll be a bitch sending samples to reviewers in the US. But such is life.

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Sun May 09, 2004 7:39 pm

Or Canada. *cough mikec* hehe :)

Acoustic dampening is a possibility, I know where to get some really good stuff and it won't add too much expense to the box. However it will increase it's weight as well as the dimensions of the box, so I am aiming to get it noise and vibration free without the use of such materials.

The 1048, properly decoupled, should be silent, as should the fans at 5v. At 12v they would most likely be noticeable, but not irritating.

sthayashi
*Lifetime Patron*
Posts: 3214
Joined: Wed Nov 12, 2003 10:06 am
Location: Pittsburgh, PA

Post by sthayashi » Sun May 09, 2004 8:21 pm

Another input thought that I have that I'd like to see in a watercooling system is a threaded connector system (if that's possible). I don't know how to accurately describe what I'm talking about (or if such a thing even exists), but it would look something like a Banana jack on a speaker.

When it comes to water, I don't like the idea/thought that if I pull on a tube that it will come off semi-easily. I think that a threaded solution may be more robust against "Tube Fatigue." Which is one of my subtle fears on Watercooling.

Am I being way off base here?

BTW, chylld. Hurry up with the W/C'ing system. My proc temps are at 66C right now with case temps at 54C, and summer hasn't peaked yet!!

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Sun May 09, 2004 8:26 pm

i'll try to find compression fittings, as they are very easy to install and almost foolproof.

sorry guys but i've hit a snag - i received a new waterblock this morning, so i put it in, but unfortunately in the process of removing the old waterblock i damaged the cpu core badly. it's still running, but 10-15 degrees hotter than it should be so no intensive tasks, which includes designing the wc system :(

i'm ordering a new processor in from america (athlon xp-m 2600) and as soon as that one arrives, i'll get back to work.

sorry again

Greg_R
Posts: 27
Joined: Thu May 06, 2004 1:23 pm

Post by Greg_R » Mon May 10, 2004 4:23 pm

chylld,
I recently built a system using the Zalman cooler (1st WC system). Here's my reasoning behind the purchase:

- supports a mild overclock
- looks are incredible
- good support on their website
- absolutely silent (0dB) With pump it is still inaudible (but may read on an SPL meter)
- Everything comes in the kit (except water) incl. hose clamps
- system is compatible with other waterblocks (have not tried this)
- promised support of future video cards and CPUs

I looked very closely at the various Danger Den kits (which would have run ~$100 more). However, I decided to go with the more silent (albiet lower performance) option.

Good luck with your project...

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Mon May 10, 2004 4:36 pm

the reserator is a very very nice product, i've seen it myself (a friend just got one). the objectives of my project compared to the reasons behind your purchase of the zalman are as follows (not to put anyone down here):

- supports a heavy overclock
- looks are pretty good
- personal via-email support :D
- very quiet, only sound will be a bit of unavoidable air turbulence, but from experience this sound is actually soothing
- everything comes with the kit, i'll be including more stuff than the zalman one
- compatible with p4 / k7 / a64 (same as zalman)
- support for some gpu's (i wish they'd establish a gpu mounting hole standard!)

compared to the zalman, essentially mine will be more expensive and a little noisier, but it's smaller, it'll allow rather extreme overclocks and it won't fail during the summer :) kinda like comparing a lotus elise to a honda accord. :)

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Mon May 10, 2004 4:39 pm

another update: turns out my cpu wasn't cracked too badly (although it is visibly damaged) - i just mounted the cpu block bad. (it was only contacting half of the core, even with arctic silver). i seem to have a good mount now and it's showing to be around 3C better than my old leaky waterblock :) overall, a success, and i can get back to designing :)

Michael_qrt
Posts: 154
Joined: Thu Oct 17, 2002 5:53 pm
Location: Sydney

Post by Michael_qrt » Mon May 10, 2004 5:37 pm

Definatly keep us informed with how it's going. Too bad you damaged your CPU but at least it still works and when you get your new one you can put together another folding box :)

Anyway, what are your plans to get this thing manufactured in quantity? Obviously a few prototypes can be put together easily by someone with skills at making things but if this unit turns out as good as we all hope then lots of people will want one. I'm just curious about this side of the project as its another aspect of it which I have no experience with.

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Mon May 10, 2004 5:41 pm

I already have a company in mind that will be able to mass-produce the product. however i'll be putting them together by hand (i don't trust slave labour hehe) so i think the limiting factor will be how many i can produce in a given week/month.

the production side is actually quite iffy because i don't think i can afford to patent the design, so i'm going to be modifying the internal structure a bit so that the parts look meaningless by themselves and hopefully no one at the manufacturing company will copy it.

Michael_qrt
Posts: 154
Joined: Thu Oct 17, 2002 5:53 pm
Location: Sydney

Post by Michael_qrt » Mon May 10, 2004 6:10 pm

Well if you get each individual component manufactured seperatly and just put it together yourself then the manufacturers wouldn't really have the overall system design in front of them. If you know the people there and know they are ersonable then it should be ok.

Another thing I don't understand is what about the design could you patent? I mean your product is basically an improved version of similar products already on the market.

Also if you are successful enough to be limited by how many of the things you can physically put together you could easily hire a few people to work with you on it and train them to do it properly, hell I'd do it.

It'd just be a shame to become a victim of your own success and not be able to meet the potential demand for your product.

Gooserider
Posts: 587
Joined: Fri Aug 01, 2003 10:45 pm
Location: North Billerica, MA, USA
Contact:

Post by Gooserider » Mon May 10, 2004 6:41 pm

Chyld, I don't understand your comment about air pocket problems with a single pass rad as opposed to a dual pass. Seems to me that a dual pass is MORE likely to have air pockets since no matter how it's mounted the coolant has to make an up and down trip.

OTOH, with a single pass rad, it can be plumbed so that the coolant comes in the bottom and out the top, I can't think of a better way to get rid of pockets...

As to your design being copied, I wouldn't worry to much about the folks at the fab shop, those sorts of people routinely sign restrictive contracts and NDA's to prevent them from copying a client's work. Aside from legal repercussions, if they stole your design and you squawked it would cost them more in reputation than they'd ever make from the theft. Your copying problem would be more from someone purchasing one of your finished products and reverse engineering it, and that would be far harder to prevent.

Gooserider

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Mon May 10, 2004 6:45 pm

reverse engineering is nearly impossible to prevent, indeed.

the radiator is going to be laid on its side in the box, hence why i chose a dual-pass with inlet at the bottom and outlet at the top. i haven't thoroughly research single pass configurations yet but the dual-pass one suits my current design perfectly.

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Tue May 11, 2004 3:39 am

UPDATE

i put in a lot of hours work today - a LOT. the design is nearly almost finished. dimensions look to be 37cm long x 22cm wide x 18cm tall.

the 'essence' of the case cover is done. the base is done. the internal structure also looks to be done. i've planned routing for all the electronics and put spots for the IEC plugs and an on/off switch on the back.

one thing that i've been paying attention to is leaks. during the design process i assume that if it can leak at some point, then it will. the internal structure is designed to handle this and route water away from electronics. hopefully. :)

it appears to be structurally sound now, although i won't be able to verify that until i actually build a prototype, which should happen in the coming weeks.

thank you all for your support, given that it's been all text and ramble from me. to show that i'm not fooling you around, here's a teaser wireframe render for your viewing pleasure:

Image

not much to see there, well actually, too much to see there. don't strain too hard :)

i also have a solid waterblock design down that focuses on minimal machining time whilst still using some characteristics of jet impingement heat transfer. beauty of it is, it's easily adaptable to a gpu waterblock.

pictures of that one to come :)

Seal
Posts: 522
Joined: Sat Oct 25, 2003 4:39 am
Location: Uk

Post by Seal » Tue May 11, 2004 7:50 am

Good work chylld!

With all my watercooling experience, ive always seen the zalman reserator and been quite skeptical about it. I'd say a rather a good solid watercooling system put togeather well would easily outperform that thing in most aspects including quietness.

In my experience anything with a large reservoir instead of a radiator works well for computers that arent on 24/7 because the water gradually heats up because when it reaches an exceeds a certain temperature, it cannot get rid of more heat than it is being given. The zalman does look well designed in the sense that it has fins but on the other hand experienced watercoolers would never put up with anything less than 1/2"tubing and it *looks* as if zalman use 3/8". Waterblocks with that small barbs are known to be very bad performance wise (well compaired to the professionally manufactured likes of dangerden etc..).

The water pump that zalman uses looks very low powered too which dosent look like it could match the performance of an eheim 1048. which kinda makes sense as theyre only useing smaller tubing.

... just my opinion really... [open to discussion]

Michael_qrt, that was very interesting to read that not only your performance/value expectations are very similar to mine not to mention your wish to watercool with a portable, high performance, overclocked, QUIET computer. I was in your shoes wanting all that when i looked into watercooling and realised all that IS actually possible and not only that, i actually went ahead and did it myself and amazed not only myself but others on this forum too with my temperatures!

Image

Thats an old image but you get the idea of what ive done in my little 3700 bqe. Ive removed the enitre front bay section for the hard drives for space for my radiator. The pump is an eheim 1048 which is very quiet and MORE than adequate to cool a processor cooling the gpu IMO is ott and most of them can be cooled passively with a zalman. The best part of my system is that my waterblock is quite old and my tubes need changing yet i get amazing temperatures from that setup you see there with the 120mm fan on only 5v. My setup is a Atlhon xp-M 2500 overclocked to 3200+ equivalent speed (the chip can handle a LOT more) and my cpu temperatures at a 23 C ambient are currently 43 C on 100% load (my comps on 24/7). The temperatures speak for themselves.

Portability wise, i couldnt be a finer example. As im at uni, i transport my computer to and from uni and home about 5-6 times a year. The only problem that i have is its a little heavy, otherwise transporting it with watercooling is NO hasstle whatsoever.

Also in the maintainence side at first is heavy as you get used to how it works but now i leave mine and dont touch it. I dont even look at it because i cant as its on the floor with the case closed.

What kinda rad setup you going for chylld? I had an interesting concept/idea of having like a square wind tunnel with a heatercore at each end and a single 120mm fan in the center. So with one single fan you get double the performance in theory although one rad would get slightly warmer air.

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Tue May 11, 2004 1:38 pm

iirc the zalman system uses 3/8" OD tubing. quite shocking really! however it is adequate for the performance targets they are aiming for. for us, we require bigger... in fact we could fit the zalman tubing inside our tubing, with room to spare :)

i'm going to be using a 220mm x 140mm core face area rad, cooled by 2 80mm fans and 1 120mm fan pulling air through it.

the square wind tunnel idea imo wouldn't work too well... there'd be so much restriction the fan would be moving hardly any air :) also, 2 120mm rads will cause a much greater pressure drop than 1 220/240mm rad, hurting flow rate and thus performance.

Sardaan
Posts: 42
Joined: Mon Apr 12, 2004 6:12 pm

Post by Sardaan » Tue May 11, 2004 2:06 pm

This is what I imagine (my own personal design so slightly off topic :-):

Outside deminsions of the external water cooling solution:
13"x13"x19" (the 13"x13" plays a huge role, the 19" is deep enough to hold a BQE or similar sized case on top).
The outer shell of the box is made of 1/2" plywood. This makes the interior deminsions 12"x12"x18".

The front of the box is open. This allows you to place a 12"x12"x1" air filter, one that is normally used for a small air handling unit (sells at home depot for $3).

The inside of the box contains the pump, res. (if you want a res), heat core (2x120mm) and 2 - 120mm fans.

Also in the corner is a mount for a PSU, this way you can easily power your fan's and pump (if it will use an ATX PSU). You could choose to power everything from the comp CPU with some longer cables.

Finally, I have put alot of thought into how the fans should be mounted. I have decided on a wood duct (about 10 inches in length/depth) that the fans are mounted mid way through and the radiator is mounted on the end. This way the fans are pullling air through the radiator and expelling the warmed air out the back side of the case. The air is of course filtered via the front filter and everything should be somewhat dust free.

This box would be heavy and not very mobile, but I don't do LAN parties so its not a problem for me!
Last edited by Sardaan on Tue May 11, 2004 2:39 pm, edited 1 time in total.

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Tue May 11, 2004 2:33 pm

not really OT Sardaan, feel free to share any ideas in this thread.

i might have misread your post, but i don't understand how you can fit a 2x120mm heatercore and 2x120mm fans in an enclosure 12x12x18?

OHH.. right, you're talking in inches. bloody metric education got me confused. thought you were talking in cm :)

my original idea for my radbox was to have the fan(s) at one end, the rad at the other, and everything else in between. problem with this is, unless you have a very aerodynamic pump, there's too much of an obstacle in the airflow that it takes a hit on performance. that's what i'm trying to avoid with my current design, which has isolated air channels for all fans that avoid most obstacles.

another thing: a computer PSU to power the fans and pump is absolute overkill. instead, head over to your local electronics store and pick up a 12v AC adapter. if you're like me, you'll find that they've reproduced themselves around the house anyway so you can easily find one for testing.

for my box, i intend to use a 12v/500mA unit, which is more than enough to power the fans which require 240mA@12v.

Greg_R
Posts: 27
Joined: Thu May 06, 2004 1:23 pm

Post by Greg_R » Tue May 11, 2004 2:37 pm

iirc the zalman system uses 3/8" OD tubing.
It's 8mm inner, 12mm outer (3/8" ~= 9.5mm).
the objectives of my project compared to the reasons behind your purchase of the zalman are as follows (not to put anyone down here)
No problem, just wanted to let you know what this newbie looked for in a WC system! Hopefully your kit will match up favorably to the Danger Den offerings (price / noise / performance / convienience).

Sardaan
Posts: 42
Joined: Mon Apr 12, 2004 6:12 pm

Post by Sardaan » Tue May 11, 2004 2:45 pm

chylld wrote:not really OT Sardaan, feel free to share any ideas in this thread.

i might have misread your post, but i don't understand how you can fit a 2x120mm heatercore and 2x120mm fans in an enclosure 12x12x18?

OHH.. right, you're talking in inches. bloody metric education got me confused. thought you were talking in cm :)

my original idea for my radbox was to have the fan(s) at one end, the rad at the other, and everything else in between. problem with this is, unless you have a very aerodynamic pump, there's too much of an obstacle in the airflow that it takes a hit on performance. that's what i'm trying to avoid with my current design, which has isolated air channels for all fans that avoid most obstacles.

another thing: a computer PSU to power the fans and pump is absolute overkill. instead, head over to your local electronics store and pick up a 12v AC adapter. if you're like me, you'll find that they've reproduced themselves around the house anyway so you can easily find one for testing.

for my box, i intend to use a 12v/500mA unit, which is more than enough to power the fans which require 240mA@12v.
I fixed the measurments to make it easier to read :-)

I see your point on the pump. Maybe the radiator is mounted directly behind the air filter and then there is a "chimney" to the back of the case (fans in the middle of chimney) where the heated air is expelled.

Mentally I have to think about this some...

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Tue May 11, 2004 4:18 pm

yep that's kinda what i've done with my current design. the trick is tailoring it for ease of production :)

Michael_qrt
Posts: 154
Joined: Thu Oct 17, 2002 5:53 pm
Location: Sydney

Post by Michael_qrt » Wed May 12, 2004 2:38 am

Thanks for the confidence in water cooling Seal, I'm getting more and more tempted to go that way. It'll at least have to wait till I get hold of a bit more disposable cash though and I'll have to read up on it a bit more too.

Just a question about the picture of your rig, is that extra vertical pipe there for filling/bleeding purposes? I guess it must be since you don't seem to have a res. Anyway the T joint in the piping looks to be a fairly large flow restriction compared to the rest of the tubing is that avoidable or just "too bad" you gotta have it.

Obviously the tubing used has to have a greater ID than the fittings but what can be done to minimise the restriction caused by the joints in the system?

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Wed May 12, 2004 2:59 am

Michael_qrt wrote:what can be done to minimise the restriction caused by the joints in the system?
well the simple fact is, joints are going to be a restriction no matter what. so the best thing to do is get rid of them altogether, or at least get rid of the ones that make sharp bends.

and what seal uses is called a t-line.

i've hashed a bit on both these topics in a thread i just posted; my guide to watercooling. a little more info there on both t-line and loop restriction.

Michael_qrt
Posts: 154
Joined: Thu Oct 17, 2002 5:53 pm
Location: Sydney

Post by Michael_qrt » Wed May 12, 2004 10:43 pm

I just erad your WC guide and it's a good summary of what you need to know to build a good WC setup.

The only thing I can think of to alleviate constriction due to joints is to have the tubing fluted at the ends and use fittings with an ID equal to the ID of the tubing. This would be quite impractical though as it seems you need to cut the tubing to the right length etc. Also I don;t hink it'd have much effect on overall cooling ability of the system.

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Wed May 12, 2004 11:26 pm

that's actually pretty easy to do, Michael_qrt. all you have to do to fit 1/2" ID hose onto 1/2" ID barbs (as opposed to the regular 1/2" OD barbs) is put the hose in hot water for a few seconds to soften it up, and then push it on. the benefit would be marginal, but it can be done very easily.

UPDATE

scouted around half way across town today to find some vibration control products, and couldn't find them. looks like i'll be ordering some stuff in from america, probably sorbothane (with which many of us forum users are familiar).

i've chopped a whole centimetre off the width so it's only 21cm wide now. i'm pretty happy with the way it's looking at the moment.

waterblock design is almost finalised as well.. so things are looking good for some prototyping to be done by the end of the month :)

Seal
Posts: 522
Joined: Sat Oct 25, 2003 4:39 am
Location: Uk

Post by Seal » Thu May 13, 2004 3:05 pm

Nice work chylld!

BTW michael yeah i was justa bout to say the same too about using bigger barbs,at the moment i use a T line and a L as well, it is restrictive, but arguably i dont need that extra degree or 2 as my temps are already pretty amazing. Using bigger barbs is a good idea as theoretically there is no or very little restriction.

Post Reply