• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

48C high for TDX

H

hanlin

Limp Gawd
Joined
Oct 20, 2004
Messages
243
Here's my hardware:

FX-55 @ 2.96 1.76v (using multimeter)
DD 12V pump
TDX nozzle 4
Maze4 for gpu
single heatercore w/ shroud and Delta fan

Here's my setup

pump -> res -> cpu -> gpu -> heatercore (mounted outside) -> pump

Here's the thing, my cpu runs goes up to 48C after hours of folding. I have taken off the block and reapplied the AS5 many times. It's been more than a week since the last application (a dab in the middle and put the block right on). My goal is to run my FX at 3ghz stable for 24/7 use. However, I think the 48C is what's holding it back. Is this too high for water cooling?
 
OK, I think the setup should go,

Pump>CPU>GFX>HeaterCore>Res>Pump.

However, I think the maze4 may be hindering your performance because (correct me if i am wrong) it is a higher pressure, lower flow piece, this means that the tdx isn't getting the flowrate it needs. This supposedly can be helped with the right spray nozzles in the TDX. Also, the type of heatercore and fan attached will play a huge role as well.

M
 
Which spray nozzle should I use? I've tried only 1 and 4 and 4 seems to be doing better than 1. The heatercore is the one from the dangerden site. I have an 120mm fan and it's on full which is around 130cfm.
 
you you have some way ot checking your coolant temps?

some exposed brass fitting that you could tag a thermal probe on to?

if you coolant temps are around 42, and your CPU is 48, then it's not the waterblock that is the weak point in that system; you might need a two fan rad.

and yeah, i think that the res should be before the pump, as opposed to after it.
 
You should be arrested and sent to Cuba for putting the res after your pump, half the point of using a reservoir is to provide a nice supply of water to the pump.

And feel the block's copper, if it's not very warm then your readings are messed up or your block isn't mounted correctly. My RBX rarely even feels warm, I love water cooling :)
 
The block definitely doesn't feel too hot. Water temps are 27C measure with a thermometer in the reservoir. It idles at about 32C no cool n' quiet.
 
thewhiteguy said:
You should be arrested and sent to Cuba for putting the res after your pump, half the point of using a reservoir is to provide a nice supply of water to the pump.

And feel the block's copper, if it's not very warm then your readings are messed up or your block isn't mounted correctly. My RBX rarely even feels warm, I love water cooling :)
Click to expand...
harsh man. don't even joke about that.
 
hanlin said:
The block definitely doesn't feel too hot. Water temps are 27C measure with a thermometer in the reservoir. It idles at about 32C no cool n' quiet.
Click to expand...
do you have a video card and software to check the graphics temperature? if that is also running hot, then your coolant may not be circulationg properly. check pump is spinning properly, hoses are freee of kinks, not blockage anywhere.

if the vid card temp is where it should be, then you might be looking at a wonky sensor.

what board do you have?
 
Video card is fine. I have an X800 Pro. Ati tool reports 33C max. My mobo is MSI K8n Neo2 Plat, bios 1.5 official.

Also, I noticed the temps jump extremely quickly, going from the 32C idle to 48C load in a about 2 seconds.
 
you should probably move the res regardless, but this is sounding like either a poorly mounted block, or a wonky sensor.

have you tried alternative temperature monitoring software?

have you checked to make sure that the heatspreader on your proc is nice and flat?

you said that you have already tried taking off and putting on the block sever times, and from the tone of your initial post, i assume that you always recorded similar temps.
 
Yes, I get similar temps after adjustments. I just put a dab of AS5 in the middle and screwed the block down. After removing it, it was spread quite evenly over the core and almost reached the edges. It seems to be making good contact.

So far, I've used CoreCenter and Sandra. I will try some others and see if they report the same.
 
whoa... I hit 52C on Prime95 with a 3500+ at 1.7V 2.64Ghz, same setup as yours minus the DD accelerator nozzle. Im not worried because stock cooling at stock clocks hits 54C. You have very low temps :D
 
I would say that you've reached the cooling limit of a single 120mm rad. I bet a double rad would get you closer to the temps your looking for. Since you have it externally mounted it shouldn't be too much of a problem to switch it out.

Just to reiterate, the res really should feed the pump.
 
Sideroxylon said:
I would say that you've reached the cooling limit of a single 120mm rad. I bet a double rad would get you closer to the temps your looking for. Since you have it externally mounted it shouldn't be too much of a problem to switch it out.

Just to reiterate, the res really should feed the pump.
Click to expand...
are you interpreting his having 27C coolant temps in his res diffrently than i am?

if his coolant temps are that low, i think that the problem does not revolve around the rad.

it might hold him back if he starts to go harder/farther, but at the moment, i think his stopping block lies elsewhere.
 
My room temps are about 22C so I don't think I can get the water to get any cooler. Right now, I'm thinking of removing the IHS. I read over at xtremesystems that most people find it very easy to do and are getting much better temps. I'm still debating whether I should do this or not? What are you guys' opinions on this?
 
hanlin said:
My room temps are about 22C so I don't think I can get the water to get any cooler. Right now, I'm thinking of removing the IHS. I read over at xtremesystems that most people find it very easy to do and are getting much better temps. I'm still debating whether I should do this or not? What are you guys' opinions on this?
Click to expand...


Check out my thread here:

Going naked drops temps!

My temps dropped 7c removing the IHS. But, YOU HAVE TO KNOW WHAT YOUR DOING!!!
It's much easier to get a improper mount without the IHS, meaning more of a chance you'll crack and or chip your core. The core is very small on winnie's, around half the size of a Barton AthlonXP core, plus there are no neoprene buttons to help with the mount.
 
I don't have a Winchester. I have a Clawhammer. I know I have to be careful if I do decide to do it.
 
DFI Daishi said:
are you interpreting his having 27C coolant temps in his res diffrently than i am?

if his coolant temps are that low, i think that the problem does not revolve around the rad.
Click to expand...

You may be right. I forgot he gave his coolant temp. Could be that his rad can handle it until he puts some load on it? Maybe a water temp at full load after running for an hour or so would be more telling? That should tell us if he's getting good heat transfer or not.

Just as an example my coolant temp is 26c with an ambient of 20c. I'm sitting at 39c while typing this and my load temp is 46c. The rig in my sig is being cooled by a TDX cpu block and a Polarflo vga block on a 1/2" system with a dual 120mm hc powered by a AQX-50Z pump. Now that I've given you that info, I'm not sure what your going to do with it but for some reason I felt compelled to share it with you. ;)

hanlin said:
Right now, I'm thinking of removing the IHS.
Click to expand...

I don't think you need to go that far. Put a straight edge on it and see if it's flat. I've heard of some that can be concave. If that's the case I would consider lapping it before prying off that ihs. Also check to make sure that you wb is nice and flat too.
 
Sideroxylon said:
Just as an example my coolant temp is 26c with an ambient of 20c. I'm sitting at 39c while typing this and my load temp is 46c. The rig in my sig is being cooled by a TDX cpu block and a Polarflo vga block on a 1/2" system with a dual 120mm hc powered by a AQX-50Z pump. Now that I've given you that info, I'm not sure what your going to do with it but for some reason I felt compelled to share it with you. ;)
Click to expand...
excellent, you have fallen for my ruse. i will now use this information to further my plot to take over the world!!!!!!!!!!

or not.

i like double rads/heater cores as well. i reccomend them to anyone making a go of watercoling. it might be a good idea for him to do as you propse to verify that his rad is not a a weak point, but i think that evidence thus far is against it.

no hard feelings either way, right?

hanlin: i think that there are safer things to try, before you go and pry off the IHS.
 
I have been folding now for probably 2 hours now. Water temps are 28C. I don't think the heatercore is my problem here.
 
hanlin said:
I have been folding now for probably 2 hours now. Water temps are 28C. I don't think the heatercore is my problem here.
Click to expand...

but thats the point.

you DO want water temps to be pretty warm because that means that the system is doing its job by taking the heat away from the CPU.

most people have this misunderstanding that if their heatsinks are cool to the touch then they automatically assume that the CPU must be just as cool. NOT! there again, you want a heatsink to be warm or hot to the touch because that means (like I said earlier,) that the heatsink is taking the heat away from the CPU.
 
ozziegn said:
but thats the point.

you DO want water temps to be pretty warm because that means that the system is doing its job by taking the heat away from the CPU.

most people have this misunderstanding that if their heatsinks are cool to the touch then they automatically assume that the CPU must be just as cool. NOT! there again, you want a heatsink to be warm or hot to the touch because that means (like I said earlier,) that the heatsink is taking the heat away from the CPU.
Click to expand...


Not exactly true. Modern blocks will dissipate the heat into the water in the small area directly over the core. If the block feels hot then the heat is NOT going into the water. A hot block is one that is of a poor design. A good block will only be a few degrees hotter than the water, say around 8c - 12c under load, higher than water temp.
 
First thing, move the pump after the resovoir. If you have the pump before the res, you're just letting the suction of the pump pull the water through the loop, and that will be weaker (especially after all the restrictions of the rad and other coolers).

Then, if that doesn't help, to see if it's the block or not, clock, down your CPU, and take the delta temp between idle and load. Then raise your CPu a bit, and take the delta again. If you're block is limiting you, then you'll notice a larger increase in Delta then if it was the radiator.
 
hanlin said:
pump -> res -> cpu -> gpu -> heatercore (mounted outside) -> pump
Click to expand...

I think you should tinker with your water circuit before spending any more $$.

this is how I'm set up atm. (reccomended by dfidaishi)
res -> pump -> xchngr -> GPU > CPU > res
 
I'll change my loop in the next couple of days. I have 2 options:

pump -> radiator -> cpu -> gpu -> res -> pump

or

pump -> cpu -> gpu -> radiator -> res -> pump

The second option will require a little extra tubing though. I have my heatercore mounted outside and the tubing goes through the bottom of the back of the case. My reservoir is mounted in the 5 1/4 bays towards the top. If I go the second route, I will probably need an extra 6 inches of tubing. Will this be fine? I have probably about 4-5 ft right now.
 
hanlin said:
I'll change my loop in the next couple of days. I have 2 options:

pump -> radiator -> cpu -> gpu -> res -> pump

or

pump -> cpu -> gpu -> radiator -> res -> pump

The second option will require a little extra tubing though. I have my heatercore mounted outside and the tubing goes through the bottom of the back of the case. My reservoir is mounted in the 5 1/4 bays towards the top. If I go the second route, I will probably need an extra 6 inches of tubing. Will this be fine? I have probably about 4-5 ft right now.
Click to expand...


Out of those two setups, go with what is most convenient for you. Less tubing equalls less resistance, but six extra inches wont affect you much anyway..
 
i think that having the pump go into the blocks first is preferable for a flow-impingement block like your's.

probably only a small difference, though.
 
DFI Daishi said:
i think that having the pump go into the blocks first is preferable for a flow-impingement block like your's.

probably only a small difference, though.
Click to expand...

So you're saying to go with the second setup.
 
hanlin said:
So you're saying to go with the second setup.
Click to expand...
yes i am.

i am also noting that the difference between the two will most likely be very small, though.

both layouts "fix" the res issue, so it's all good.

might want to give your IHS and block a quick pass with some 1500 or 2000 grit wet sandpaper while you are at it.

check for flatness on the IHS as well.

i remember that you said that the paste spread out the way it should, but if the IHS is concave, then it'll be thicker in the center than is optimal.
 
Well, I went with pump -> rad -> cpu -> gpu -> res -> rad. This was much easier to do than the other one. I'm not getting any temp differences though. I still can't overclock any higher than I was able to before the change.
 
That varies too much based on ambient temps equipment etc. Also, for the overclocking, if you hit a wall, sometimes the voltage is lacking. It's hard to isolate what is holding you back (CPU, Chipset, ram). The best way I find out my max oc is to relax the ram timings a bunch, give the cpu and ram some juice and up the fsb in 3 mhz increments. After the max fsb is attained, try tightening the ram timings and stop when it becomes unstable. Then, I see if that configuration runs with lower voltage. If not, than voltage may have been holding you back, if it does run, than you probably hit the limits of your configuration.

M
 
Voltage doesn't get it any higher. My guess is that the heat is what's holding back the overclock. 46C (dropped a couple degrees changing loop) seems to be what a lot of people are getting on air. I see a lot of people with water running under 40C. This is my goal. If I can do that, I'm sure I can do 3ghz stable.
 
well..........this is a pretty lousy situation.................

a bigger pump might help, or it might not.

removing your IHS might help a lot, or might not help so much.

you might have hit a brick wall.

if you're feeling McGuyver-ish you could: wrap you tubes and blocks in rags for temporary condensation proofing, disconnect the fan on your rad, get a basin full of ice/dry ice, fill you loop with normal antifreeze/water mix(car ratio), fill the bucket with the same, take the tubes that you normally have going in and out of you res, put some nylon stocking material over the end of the intake hose, stick the hoses in the basin, and run sub-ambient for a while to see what kind of speeds you can hit then.

at least that would let you know if your proc has more in it or not.
 
I just remembered something. I'm using regular car antifreeze at a ratio of 10:1. I just read that antifreeze doesn't do as well as products like hydrx and zerex. Is this something that could cause higher temps. It seems that the antifreeze causes a layer of something to build up on the tubing. I don't know what it is but it comes off very easy when I try to wipe it off. Also, will using nozzle 5 get me lower temps? That's the one with the circles on it.
 
I recently moved my comp to my basement. Ambient is around 16C. I've updated my bios to 1.8 beta which supposedly fixed the temp bug. Now I'm getting 23C idle and 34C running large fft in prime95. I hope these are the actual temps which would be very nice.
 
did you let the AS5 break in???

You mention JUST putting the AS5 on and reapplying it a lot. It needs about a week to reach low temps. Temps can actually go UP if you don't wait!

I would not OC that much until I am sure the AS5 is broken in.
 
You must log in or register to reply here.
Back
Top