Good upgrade from eheim 1250?

[DarkenReaper57]

Well, I've been using an eheim 1250 for about a year and a half...and I bought it used. Anyway, I am redoing my w/c system this summer, and the system will be more restrictive with two heatercores.

What is a good pump upgrade? Keep in mind that I don't want to spend too much...as the MCP600 or MCP650s are like...$80.

I was thinking along the lines of a danner mag 3 or mag 5, leaning towards the latter. They are reasonably inexpensive, yet effective. What do you guys think? The mag 5 provides more head and a higher gph at that head than the Eheim 1250. I also have reason to believe that my current pump is...wearing down...and my flow isn't optimal. I could be wrong.

Oh yeah, I read somewhere that some of the mag pumps put out a lot of heat. Will this be a significant issue? I am trying to keep my coolant, and ultimately my gpu and cpu, cool with the new setup I am going with.

Thanks for any input.

 

[orkan]

I went from a 1250 to a Laing D4. I am pleased.

 

[plywood99]

C-Systems is putting out a new mag pump that will retail for around $50. I'm thinking of getting one myself but I'm waiting for some reviews to come out. I think Dtek has it on their site.

 

[DFI Daishi]

oh, the danners put out quite a lot of heat, but most of it seems to be put into the air, and not the coolant.

 

[DarkenReaper57]

C-Systems is putting out a new mag pump that will retail for around $50. I'm thinking of getting one myself but I'm waiting for some reviews to come out. I think Dtek has it on their site.

Thanks for the heads-up. Looks like a nice little pump, but I'm not sure if it provides the head I want...the cheaper mag 5's stats are better if memory serves me right.

I'll need to wait until benches, of course.

 

[DarkenReaper57]

Another thing I thought of. I am currently using a d-tek WW. Will I see a benefit in upgrading my pump with this block? I would think so, since this block is rather restrictive. I am also using a d-tek GPU cooler on my 6800 GT, if that matters at all.

So, does anyone argue to an upgrade to a Mag 5? Any other suggestions that aren't too expensive?

 

[DFI Daishi]

maybe go to an aquarium store and see what they have with similar flow ratings at 3-4 feet of head, at comperable to lower prices.

that is to say that you should compare the danner's flow at 4 feet to other pumps at 4 feet. they should all have performance curves on the boxes and in the packages. the staff could probably help you.

 

[DarkenReaper57]

maybe go to an aquarium store and see what they have with similar flow ratings at 3-4 feet of head, at comperable to lower prices.

that is to say that you should compare the danner's flow at 4 feet to other pumps at 4 feet. they should all have performance curves on the boxes and in the packages. the staff could probably help you.

Is 4 feet of head a good target number to look at? Or were you just using 4 as an example.

Hmm, I've been looking at some charts online, comparing head at different flows. The idea in my mind is to look at a really good pump, like a mcp650, and see what the flow is at that level. I can then deduce the head from a chart at that flow. I can then take a look at the charts of other pumps and compare the head and flow of the good pump to that of the pump I wish to buy.

Basically, here is the setup I plan on using (I already have most of the equipment)

-Inset pump here (currently own eheim 1250)
-D-tek WW CPU
-D-tek GPU
-77' Bonneville and 86' Chevette in parallel (I will use y-connectors, and punch the end in which there is more flow, so I can have equal flow through the radiators)
-Maybe a res...I currently have a bay-res which I like the aethetic reasons, but since I am going to an external setup, I doubt I will use it. I might, however, integrate it somehow in the box so it looks cool and doesn't go to waste. I might end up going with a t-line for convenience, though.
-1/2'' ID tubing
-Note that I will use quick disconnects.

Taking all of this into account...what do you guys think the max flow will be? It will help a lot in figuring out what kind of pump I will need.

 

[plywood99]

Is 4 feet of head a good target number to look at? Or were you just using 4 as an example.

Hmm, I've been looking at some charts online, comparing head at different flows. The idea in my mind is to look at a really good pump, like a mcp650, and see what the flow is at that level. I can then deduce the head from a chart at that flow. I can then take a look at the charts of other pumps and compare the head and flow of the good pump to that of the pump I wish to buy.

Basically, here is the setup I plan on using (I already have most of the equipment)

-Inset pump here (currently own eheim 1250)
-D-tek WW CPU
-D-tek GPU
-77' Bonneville and 86' Chevette in parallel (I will use y-connectors, and punch the end in which there is more flow, so I can have equal flow through the radiators)
-Maybe a res...I currently have a bay-res which I like the aethetic reasons, but since I am going to an external setup, I doubt I will use it. I might, however, integrate it somehow in the box so it looks cool and doesn't go to waste. I might end up going with a t-line for convenience, though.
-1/2'' ID tubing
-Note that I will use quick disconnects.

Taking all of this into account...what do you guys think the max flow will be? It will help a lot in figuring out what kind of pump I will need.

You really don't need 2 rads with your setup, just complicates things. Stick with just the Bonne core.

 

[DFI Daishi]

Is 4 feet of head a good target number to look at? Or were you just using 4 as an example..

it seems to me that 4 feet of head is about where most DC pumps start pulling ahead of most AC pumps in GPH flow, and so good flow at or past that point is what i would consider to be desirable in a high performance AC pump.

 

[DarkenReaper57]

You really don't need 2 rads with your setup, just complicates things. Stick with just the Bonne core.

I have done that with an external setup before. My temperatures were alright, but not as good as I wanted them to be, even after some really good lapping and making sure everything was mounted correctly. Heck, my water temperatures in the bay res were a bit high at about 8C above ambient if I remember correctly. This means either the pump is not pushing water quickly enough, or the radiator is not up to the task of dissipation. It could have been a combination of both.

Oh yeah, I live in FL and when I go home for the summer, my parents keep the house at about 80 F because they are cheap . I'd like to keep things close to ambient so I don't need to bother with tweaking my clocks, etc.

I also want 2 cores because...well...it will be fun to make and is [H]. I have even considered going with dual bonnies, just for the fun of it. It will be an external box anyway, so its not like I need to fit things inside the system. I'm an engineering major...so I like complicated stuff, hehe. It's fun to work on.

I also want to future-proof my system since I will be upgrading to either the venice core or dual cores.

Basically my aim is to get my water temperatures closer to ambient than with the single bonnie core, in effect getting my cpu and gpu temperatures lower.

One more thing. I've done a bit of research on the mag 3 vs. the mag 5. Some say that the mag 3 is better overall because in the head vs. flow graph segment that watercooling falls under, their performance is very close. The mag 5, however, puts out more heat (it is a 45 watt pump, the mag 3 is 35 watts). I have also read that a lot of the heat is dissipated out to the body of the pump, thus not affecting the water temperature too much...so I really dont know what the difference in performance overall would be between a mag 3 or 5. What do you guys think?

 

[plywood99]

A typical magnetic pump will dump between 80 - 90+% of it's heat into the water. A good number to shoot for is a pump that puts out less than 50 watts, less is better. 50 watts in an average setup will raise temps around .5c. Yes that is point 5 degrees celcius....

In water cooling head is far more important than gph. For example the DDC only puts out around 90gph but it's head is 13 feet.

I know you said you want to go cheap, but it's hard to beat a mcp600/aq50z for cost/performance. They really are some of the best pumps out there, even at the cost of $70 duckies.

As for the rads , since you are going ahead with two of them, I really recommend the above said pumps. Why cheapen your system?

 

[DarkenReaper57]

A typical magnetic pump will dump between 80 - 90+% of it's heat into the water. A good number to shoot for is a pump that puts out less than 50 watts, less is better. 50 watts in an average setup will raise temps around .5c. Yes that is point 5 degrees celcius....

In water cooling head is far more important than gph. For example the DDC only puts out around 90gph but it's head is 13 feet.

I know you said you want to go cheap, but it's hard to beat a mcp600/aq50z for cost/performance. They really are some of the best pumps out there, even at the cost of $70 duckies.

As for the rads , since you are going ahead with two of them, I really recommend the above said pumps. Why cheapen your system?

Thanks for the input. As for the rads...it is true that I am going with 2, but they will be in parallel...so the restriction will probably be even less than a single rad...

Well, I really don't feel like dumping $80 into a pump (after shipping). Do you guys think I would see a difference if I bought a danner mag 3 or 5 to replace my eheim? The danners have a max head of about 10' and a higher flow at all head values. This eheim, while reliable, only has a max head of 6.5' or so...

 

[DFI Daishi]

Thanks for the input. As for the rads...it is true that I am going with 2, but they will be in parallel...so the restriction will probably be even less than a single rad...

Well, I really don't feel like dumping $80 into a pump (after shipping). Do you guys think I would see a difference if I bought a danner mag 3 or 5 to replace my eheim? The danners have a max head of about 10' and a higher flow at all head values. This eheim, while reliable, only has a max head of 6.5' or so...

well, i know that you said that you are an engineering student (used to be one of those myself) however the science student way of testing to see if a larger pump will help is to measure your coolant temps going into and out of your blocks and rad. if the delta T across the block is significant, then a stronger pump will help somewhat provided that the delta T across your rad is similarly high. if you don't have both of those conditions, then the pump is not the bottleneck that you should be worrying about.

 

[DarkenReaper57]

well, i know that you said that you are an engineering student (used to be one of those myself) however the science student way of testing to see if a larger pump will help is to measure your coolant temps going into and out of your blocks and rad. if the delta T across the block is significant, then a stronger pump will help somewhat provided that the delta T across your rad is similarly high. if you don't have both of those conditions, then the pump is not the bottleneck that you should be worrying about.

Thanks for the input, I didn't think of that. I guess I need to figure out an easy way to measure temperatures in 2 places, as before I just stuck a thermometer into the reservoir (I'm not so sure how accurate this was).

It is difficult to do this in the first place because I am only using the 86' chevette core. I would need to redo everything, go though the hassles of testing, then find out whether or not the delta t is significant.

I would prefer to jump past that stage if possible. I guess I could do it if I really needed to, but it would mean that I would need to outfit the case with quick disconnects, drain, reroute tubing, buy and place thermistors, and test. Might sound easy, but things tend to take twice the amount of time I initally estimte to do the job right.

Going from a single 86' chevette to a the same chevette with a bonnie core in parallel (at optimized flow to each core for temperatures), do you guys think I would benefit from upgrading my pump from an eheim to a mag 3 or 5? Any other suggestions? It is probably difficult to gauge, but I just don't want to waste money.

I suppose I could always try it with this pump, putting in probes during the build process (routing these to the mobo for permanent use) and judge from that.

On another note, does the type of tubing have a significant effect on temperatures, assuming there is negligible difference in kinks? I have always used the cheap home-depot stuff (vinyl?) since it is cheap. I use coolsleeves so kinking isn't an issue.

EDIT: I've pretty much decided to stick with the pump I have after doing quite a bit of research. Going to my new setup with probably even lower overall restriction. If there is more, it is probably not too significant. Thus, a new pump will probably make a 1-2C max difference. Not worth $50-75. I'll do some testing this summer, though. If my flow rates are like...1 liter/min (about 1.75 to 2 is good for my Whitewater), then I will consider buying a new one.

The price of the danner mags 3 and 5 seem to have gone up about $10-$15 over a year too...

At least I learned a lot more about the details of head loss and fluid mechanics in a computer system

 

[DarkenReaper57]

I plan on using the quick disconnects by colder (1/2'' barbed). I knew they were somewhat restrictive, but apparently they only allow 3/8'' nominal flow. Is there any way to tell how will this affect cooling performance with the above setup?

Argh. I really want to use quick disconnects...or some sort of lever disconnects...with my next setup for mobility purposes. I don't, however, want them to worsen my cooling.

As for levers or valves like this...are these viable? I have seen them at home depot and lowes, but they arent barbed.

 

[plywood99]

If the quick connects are 3/8 i.d., it should have minimal effect on your system.

 

[DarkenReaper57]

If the quick connects are 3/8 i.d., it should have minimal effect on your system.

Really? I figured it would hamper my system quite a bit, and in effect bringing me from a 1/2'' ID system down to a 3/8'' ID system. This is not true?

Also, how would the restriction of a bonnie and a chevette core compare to this if in series? What about in parallel with tweaked flow (so one does not get more than the other)?

I'm trying to weigh out all of the pro's and con's before I build.

 

[Rix2357]

If your really worried about the flow, you can get the colder quick disconnects in 5/8" size. That is equivalent to 1/2" flow rates. It will take a little effort to get the tubing on, but it does stretch and fit. Getting it off will be impossible however.

 

[TehRoflcopter]

1/2" modded DDC 12V.

Best thing I ever did. I did a test to see how much more water it pushed out with the mod compared to no mod. Wow, it was like a friggin jet stream. The unmodded DDC appeared to push out some water like a half way open tap would, but the 1/2" modded one just launched it RIGHT out there.