• 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.

Monitoring software for Pumps?

DeathPrincess

Fully [H]
Joined
May 15, 2010
Messages
18,205
I use the open hardware monitor for monitoring my tepretures and other things like RPMs and various whatever the hell is that doing things. But theres always been one problem. The pump is connected to the CPU fan header (it's a DD-CPX-Pro 12V) and so I get the readout as a fan on the monitor. But theres a problem...

33xa7ut.jpg3


337500RPM!!!

Unless I have the world's sexiest and quiestest for RPM (unless someone else has a 19dB 337500 fan) then theres something not being monitored right here. It uses the yellow wire to vary the speed.

Is there any software for monitoring pumps properly?
 
As far as I know, the tach lead on most pumps is purely meant as a failsafe in case the pump stops running, so you get some kind of audible alert from motherboard and/or the bios shuts down the system automatically. If you don't have a variable speed pump, that's all that matters really.

I'm pretty sure the RPM values don't correspond directly with fan RPM speeds. At the very least, if the pump is slowing down or you adjust the speed, the rpm values should change accordingly.
 
It just gives random results, it constantly goes between 0, these kind of numbers and infinity (you can have infinity RPM!?)
I'm not actually sure if it is varible to be honest. I used a fan controller while it was being built to vary the votage, and that did seem to work. But I also burnt out the controller because I forgot it was only 6w per header (it really was burnt out, with smoke and everything!).
I'd just like some way of monitoring what it's doing other than holding tubes which gets impracticle after a while!
 
motherboard monitor used to have dividers you could check to help with high/low speed fans, where the fans may spin and trigger the tach faster than the polling rate of the header, or for fans that spin so slowly they miss the poll for more than a few turns and trigger the fan fail alarm...

dose your software have anything like this, might be able to find a sweet spot where it actually read the tach correctly.
 
The only option it has is C/F for the tempreture!
Is the motherboard monitor software still good?
 
yep check the bios.

I would set cpu fan speed control to disabled (100% speed) until I figured out what was going on.

Speedfan has rpm readouts. Think MBmonitor will be too old, it has not been worked on in ages. Cant hurt to try I guess.
 
The only option it has is C/F for the tempreture!
Is the motherboard monitor software still good?

no, way outdated.. you might get lucky and find a motherboard in the list that has the same sensor as yours, but there is no guarantee that it will be wired any where near the same. the guy (alex van cam) that used to script it had special license from the mother board makers so that he could get the correct memory interface values so that he could read the chips. there started being to many to keep up with and trying to track down the difference in every slight rev. on the motherboard, he just up and said stuff it.. but because he had to have special permission, he couldnt open source the project and it died off in... mid 2004 i think it was...
 
Most mainstream motherboards come with a software suite for monitoring/overclocking/etc. I would try those first.

Third party software like Everest and HWMonitor are also pretty good, though they're usually behind on support when new boards come out.
 
I think it is more a hardware than a software problem here. Are you sure that you connected a correct rpm sense signal to the cpu fan header?
 
Thanks everyone!

Check the readings in the bios. Are they random as well?

I didn't think of that before. I checked in the bios and it comes up constantly as 1730rpm +/- 15 (which sounds reasonable). Also, the tempreture of the CPU is very different. On the desktop I get 40c tempretures, but in the bios its 33c. It's a ASUS P6X58D-E

yep check the bios.

I would set cpu fan speed control to disabled (100% speed) until I figured out what was going on.

Speedfan has rpm readouts. Think MBmonitor will be too old, it has not been worked on in ages. Cant hurt to try I guess.

no, way outdated.. you might get lucky and find a motherboard in the list that has the same sensor as yours, but there is no guarantee that it will be wired any where near the same. the guy (alex van cam) that used to script it had special license from the mother board makers so that he could get the correct memory interface values so that he could read the chips. there started being to many to keep up with and trying to track down the difference in every slight rev. on the motherboard, he just up and said stuff it.. but because he had to have special permission, he couldnt open source the project and it died off in... mid 2004 i think it was...

The tubing coming out from the CPU does seem cold still,but ill set it to 100% for now until I can get something definate.
I tried motherboard monitor (which I had already downloaded at some point, just never used) it just gave me a load of neverending "driver didn't load" errors, and never got around to opening after holding down enter for 3 minutes, so i'm guessing it's unfortunately not compatible, which is a shame because it looks like a good one!
Speedfan, apart from being ugly has a constant reading at 1740. and provides a third tempreture reading.
Oh and @BillParrish I got the motherboard with the ''blown'' motherboard back today, so i'll be looking at that later!

3162blu.png


Thought the GPU (theres something running and only one connected) AUX and system all come up identically, the CPU tempretures are different, i'm not quite sure what the symbols next to the tempretures mean. The bios says 33c, but neither are around that, and I don't think it would get colder at desktop. The configuration doesn't have my motherboard on the list, also i'm unsure of what clock chip it uses. ICH10R?

Most mainstream motherboards come with a software suite for monitoring/overclocking/etc. I would try those first.

Third party software like Everest and HWMonitor are also pretty good, though they're usually behind on support when new boards come out.

I tried HWmonitor and got a new set of different tempretures, though the RPM was the same as speedfan. They are all 2c below that of OHM. I dont think that the CPU is 40c(104f), because the watercooling has two external pipes coming directly from the CPU and neither feels above room tempreture (I know that doesn't make much sense, but it doesn to me. Its the way a measure tempretures one is hot, one is cold).
ASUS motherboards come with Xpert fan controller. I never actually installed it when this thing was built because at the time it was a 64GB SSD only, and because it looked like shovelware. Ok after downloading at a rather retro 15kb/s it's running its up and running. It has the same 1740 rpm for the speed. But, whats intresting is that after calibrating for the different speeds, everything is 1730 to 1740, but the top speed is 337500RPM. This has the idle tempreture at 35c.(95f) the same as the bios.

I think it is more a hardware than a software problem here. Are you sure that you connected a correct rpm sense signal to the cpu fan header?

Its a 3pin into a 4pin header, but the notch on the plug only allows it to be conected one way, but i've been known to do more foolish thing in the past. But from what I can tell the wires correspond to the right connectors.
 
hiya,

Keep in mind all this software takes readings over some time period and averages the values so minor differences will be seen just based on the averaging used in the code and how often it updates the screen figures.

In the bios the CPU temp is read by a sensor located under the cpu socket while the "core" temps are read by sensors actually embedded in the cpu itself. So the bios temp typically reads lower and does not respond as fast.

On speedfan it keeps track of the trends of the temps os the blue down pointing arrows say that compared to the last few readings/samples the temps are going down in general. Kick in a load and you should get red arrows pointing up.

The "core" temps are the ones you want to pay attention too under load.

It appears that the open hardware monitor rpm sensing code has trouble with the pump signal where the code for speedfan understands it. Tach signals vary slightly in design from device to device for reasons that I do not understand (one would think the fan people and pump people could talk it out but /shrug ) so since the Open Hardware Monitor guy has a thread here a post to him might be in order. Likely he can code a fix fairly shortly.

The fact that the bios is reading it right confirms you got the 3 to 4 pin properly placed and the 1700 rpm is about right for pump speed so I think everthing is ok.

Its difficult to write software that works with every board and every accessory. So no worries.
 
For you, does speedfan gives 8 readings for cores, even though 4 of those are imaginary. They show different tempreture readings, and not in pairs. So i'm intrested what they are reading for that? Also, I've wondered often, where is the CPU sensor, is it the BIO sensor, or is it a central on chip, or just an average of all the others(which wouldn't make sense)? Strangely CPU is missing from OHM. Speedfan is a good program, and I like having the main tempreture in the taskbar.
Groups of people doing things together is miraculous when it does happen, as everyone has to be right in their opinions at the same time. Everybody has to be winning for everyone to win. Or I suppose, if everyone is losing it works the same. Damnit why still no ATX style laptops...that would be perfect. Oh because 700$ un-upgradable laptops (apart from yay moar ram!) are what the consumer wants.

I hunted around and found the OHM post so i'll tell them as i'm sure it would happen with other people too. Maybe ill get the original thing to work anyway...
 
Also, I've wondered often, where is the CPU sensor, is it the BIO sensor, or is it a central on chip, or just an average of all the others(which wouldn't make sense)?

Its a per core sensor, 4 sensors for 4 cores, depending on the software, i dont see it unthinkable to have the 4 sensors added together to get a common reading, its not like they are 4 separate cpus in their own socket. what happens on one core, (where heat is concerned) will be transmitted via the heat spreader to the other cores, so even if only 1 core is getting worked over, the other cores should still rise in temp just the same. but they should stay just a hair lower.

As for maybe getting 8 readings only thing i could think of would be if there is a temp sensor for the HT logic. but as far as i know, they dont do that. its more likely either a hardware bug, or some ghost in the software.

However, i know there is sensors else where on the motherboard. there may be one in the north/south bridge, one by the mosfets... one reading ambient temp within the pc case. one reading the graphics cards ect. the software dosent really know what the sensors belong to (outside of the cpu) as there is just a chip that monitors the diodes, that signal from that leg of the chip could be anywhere on the board. and the way a pro board is wired could be totally different from that of a le board. with graphics card, it should all read the same. but because of different expansion slot configurations it may not be at the same memory address from board to board... most of the software is pretty good about getting this right however.

Damnit why still no ATX style laptops...that would be perfect. Oh because 700$ un-upgradable laptops (apart from yay moar ram!) are what the consumer wants.

in my opinion laptops are a dyeing breed, although no one will willingly admit it. lets face it, trying to cram more and more cpu/gpu power into a laptop and expecting the performance of a desktop is getting to be harder and harder, laptops have, and always will lag behind desktops in every since, at least where performance is concerned. battery life has just about stayed the same (well it has gone up, some...) partly do to battery tech advancing (li-ion up from ni-cad) and party do to individual chips trying to do better at using less power.

can you imagine trying to fuel a laptop with a 480gtx gpu???
 
Rather than monitor the pump directly, another option is to get a flow meter with adapter to allow you to hook it up to one of your motherboard's fan headers:
Koolance Flow Meter with Power Frequency Adapter

It's not all that expensive and would also let you know if one of your blocks is getting gummed up and flow begins to drop - well before it would lead to a pump burning up due to a blockage or heavily restricted flow.
 
If the BIOS and other tools can monitor the pump correctly, then it is most likely a bug in the Open Hardware Monitor. You can help me correct the bug by doing the following:

- Download the latest version 0.1.37 of the Open Hardware Monitor and check if the problem is still there

- Save an OpenHardwareMonitor.Report.txt (File Save Report...) and post the report here

- Close the Open Hardware Monitor

- If the CPUID HWMonitor can read the fan correctly, please save a report (File / Save monitoring data) and post it too

- If you have an official monitoring tool form your mainboard manufacturer you could create a screenshot showing all the values (voltages, temps, fans), and save an OpenHardwareMonitor.Report.txt more or less at the same time.
 
Its a per core sensor, 4 sensors for 4 cores, but because of different expansion slot configurations it may not be at the same memory address from board to board... most of the software is pretty good about getting this right however. /snip

in my opinion laptops are a dyeing breed, although no one will willingly admit it. lets face it, trying to cram more and more cpu/gpu power into a laptop and expecting the performance of a desktop is getting to be harder and harder, laptops have, and always will lag behind desktops in every since, at least where performance is concerned. battery life has just about stayed the same (well it has gone up, some...) partly do to battery tech advancing (li-ion up from ni-cad) and party do to individual chips trying to do better at using less power.

can you imagine trying to fuel a laptop with a 480gtx gpu???

I was going to say something about the first part but my mind has just gone blank...
Oh yes! When I went to the OHM thread, there was someone saying the same thing about speedfan with another setup. It must be these new fangled 8 pretend cores maybe.
Haven't you heard the news! the mobile 480s are coming babay!
But yes, I always go through the same, ok so I like to edit video, and do VJ stuff, and I always end up lying lots of weird places with weird faces (I travel and know people, im not easy). So the caculations go on in my cranium and the laptop route begins. But at the end of the day even the most expensive hardcore laptop fails at most performance/cost things when looking at the most uber hardcore things I want to do. So I always just grow a new box in my 10 computers in a 2m box room and go nuts with everything ill think ill need, and it still comes under the cost of most "gaming laptops". Gaming laptops are a joke when you look at SFF computers as well anyway. Most people with serious laptops just end up putting it on a desk in an office anyway. And when 2 desktops can be bought for the same price as a decent laptop anyway the home/office arguments end up being eroded. Actually. I kind of need one again. I need one for touring with a show. I don't pack lightly so I need something that fits in the overhead locker, but can power 2 screens of VJ displays/visualisations/multilayered 3d imagery. And lots of audio processing/synths and things. I might wait for the refresh.

Rather than monitor the pump directly, another option is to get a flow meter with adapter to allow you to hook it up to one of your motherboard's fan headers:
Koolance Flow Meter with Power Frequency Adapter

It's not all that expensive and would also let you know if one of your blocks is getting gummed up and flow begins to drop - well before it would lead to a pump burning up due to a blockage or heavily restricted flow.

I was thinking about this kind of thing eventually too. Is there anything with a tempreture sensor for the water? I know it's rather overkill, but I'm a controll freak sometimes when it comes to this thing. I'll probably get something like this when I replace the tubing (needs more purple, and the clear UV tubing reflects too much, going plain tygon)
 
I tried HWmonitor and got a new set of different tempretures, though the RPM was the same as speedfan. They are all 2c below that of OHM. I dont think that the CPU is 40c(104f), because the watercooling has two external pipes coming directly from the CPU and neither feels above room tempreture (I know that doesn't make much sense, but it doesn to me. Its the way a measure tempretures one is hot, one is cold).
ASUS motherboards come with Xpert fan controller. I never actually installed it when this thing was built because at the time it was a 64GB SSD only, and because it looked like shovelware. Ok after downloading at a rather retro 15kb/s it's running its up and running. It has the same 1740 rpm for the speed. But, whats intresting is that after calibrating for the different speeds, everything is 1730 to 1740, but the top speed is 337500RPM. This has the idle tempreture at 35c.(95f) the same as the bios.

For Asus, their monitoring software is PC Probe, which you can download directly from their site. It looks like crap, but it works well, and it's currently the only monitoring software that works for all the sensors in my Asus board.
 
Thanks for the reply! I put this answer seperate because otherwise it would get buried in a wall of text as I rant.

- Download the latest version 0.1.37 of the Open Hardware Monitor and check if the problem is still there

I did this and the problem is still there. But it show slightly different readings (this one shows 667500, something like that). I tried opening both at the same time to see if the reading was the same but its different. Sometimes its stable, and other times it swings around wildly.

- Save an OpenHardwareMonitor.Report.txt (File Save Report...) and post the report here

I used the stock settings for this (for tempreture stability, also its hot today, and i'm only checking websites).

Report http://www.sendspace.com/file/dfekss

- Close the Open Hardware Monitor

- If the CPUID HWMonitor can read the fan correctly, please save a report (File / Save monitoring data) and post it too

That seems to have the same problem with the sudden jumps to crazy RPMs. Ill include it anyway http://www.sendspace.com/file/hiqjhj

- If you have an official monitoring tool form your mainboard manufacturer you could create a screenshot showing all the values (voltages, temps, fans), and save an OpenHardwareMonitor.Report.txt more or less at the same time.
[/QUOTE]

The ASUS motherboard only came with AI suite that shows fan speeds and voltages. I got two, one when it was stable:

ngc689.png

Report: Here

And one when it was bouncing around (I included the fan Xpert)
2vhwrpi.png

Report Here
When I tired to add a system fan to the probe software it added another one to OHM, but nothings connected to it, so thats why theres the extra fan header above CPU.

I hope this is helpful!

Added new pictures and reports with ASuS probe. (yes it iis ugly...)
 
Last edited:
For Asus, their monitoring software is PC Probe, which you can download directly from their site. It looks like crap, but it works well, and it's currently the only monitoring software that works for all the sensors in my Asus board.

That wasn't on the motherboard disk! damn asus. Ill get it from the site and redo the above post.
 
Koolance has a water temperature sensor, also. I recently installed a Koolance TMS-200, which is a board that you can plug your pump, temp and flow sensors, and up to three fans into, and control everything via a software interface. It is pretty customizable and has let me create a pump and fan profile to minimize noise when idle, and kick it up when the water temps start rising.
 
Thank you very much.

I think the problem is caused by running more than one hardware monitoring application at the same time. This can cause conflicts when communicating with the monitoring chip, leading to wrong readings.

I have built a new version which should have corrected labels for the W83667HG chip and your mainboard.

http://openhardwaremonitor.org/openhardwaremonitor-v0.1.37.1-alpha.zip

Please try if the problem with monitoring the pump is still there if you run only the Open Hardware Monitor and keep Asus PC Probe, Speedfan, HWMonitor, CPU-Z or anything similar closed.
 
Koolance has a water temperature sensor, also. I recently installed a Koolance TMS-200, which is a board that you can plug your pump, temp and flow sensors, and up to three fans into, and control everything via a software interface. It is pretty customizable and has let me create a pump and fan profile to minimize noise when idle, and kick it up when the water temps start rising.

I would like to try something like that with a tempreture probe and flow rate, but I am always wary of fan controllers because of the mod thats being done to the case, it makes the front panel kind of inaccessable, so the adjustment would be difficult! Also I only have a single bay left that cant be more than 2 inches deep from the from because of the pump(s) being wedged in the 2 low 5.25 bays with ballistic jelly stuff, and the top bay having a ghetto CCFL switch device mounted in it. How small is it depth wise?

Thank you very much.

I think the problem is caused by running more than one hardware monitoring application at the same time. This can cause conflicts when communicating with the monitoring chip, leading to wrong readings.

I have built a new version which should have corrected labels for the W83667HG chip and your mainboard.

http://openhardwaremonitor.org/openhardwaremonitor-v0.1.37.1-alpha.zip

Please try if the problem with monitoring the pump is still there if you run only the Open Hardware Monitor and keep Asus PC Probe, Speedfan, HWMonitor, CPU-Z or anything similar closed.

Thankyou for the quick work on the issue! I deleated the older 1.35 that turns on on startup and then made sure everything was closed.
Opened up the new one and it has the values for voltages and GPU usage;
35d6xqw.png

The only item left off the list is the SSD, but i'm not even sure they have tempreture probes in them. And the auxillary probe that was there before, but I am assuming it wasn't a real one anyway.
But yes the problem is still there, I timed it and it seems to happen either every 1 minute or 30 seconds for about 3 seconds then return to normal. It happens slightly less than before though!
I tried out your theory and opened PC probe at the same time as this build, and it did seem to unsettle the OHM values for a few seconds. I also used it to check the voltages againsed your new build and they seem to be perfect!
 
I would like to try something like that with a tempreture probe and flow rate, but I am always wary of fan controllers because of the mod thats being done to the case, it makes the front panel kind of inaccessable, so the adjustment would be difficult! Also I only have a single bay left that cant be more than 2 inches deep from the from because of the pump(s) being wedged in the 2 low 5.25 bays with ballistic jelly stuff, and the top bay having a ghetto CCFL switch device mounted in it. How small is it depth wise?

The Koolance TMS-200 doesn't have to go in a 5.25 bay. It doesn't have any knobs or switches that need to be adjusted; instead it plugs into either an external USB port or internal USB motherboard header. You control everything via software.

You wouldn't be able to fit it into a bay with only two inches of depth. You can put the card in a pci or pcie expansion slot, in which case you'd use an external USB connection from the back of the device to a rear panel USB port. It doens't actually interface with the PCI / PCIE slot; the PCB just inserts into the slot as a means of mounting.

Or you can just mount it somewhere else in your case, like a HD bay, but this might take some minor modding.

The Aquacomputer Aquero is another option, and that may fit in a bay with two inches of depth. But you should double check it is powerful enough for the pump you have.
 
The Koolance TMS-200 doesn't have to go in a 5.25 bay. It doesn't have any knobs or switches that need to be adjusted; instead it plugs into either an external USB port or internal USB motherboard header. You control everything via software.

You wouldn't be able to fit it into a bay with only two inches of depth. You can put the card in a pci or pcie expansion slot, in which case you'd use an external USB connection from the back of the device to a rear panel USB port. It doens't actually interface with the PCI / PCIE slot; the PCB just inserts into the slot as a means of mounting.

Or you can just mount it somewhere else in your case, like a HD bay, but this might take some minor modding.

The Aquacomputer Aquero is another option, and that may fit in a bay with two inches of depth. But you should double check it is powerful enough for the pump you have.

No PCI slots either. It goes;

Tubing bulkhead passthrough thing
Wireless card
470
470 (Currently missing)
260 (Currently missing)

Its a 18w pump which means the number of fan controllers able to handle it is few. Would it fit in a 3.25 bay as theres 3 more harddrives to go in the HDD bays so thats kind of out. Anything would need to be pretty, or eaisily hidden as the window on the side shows everything from the powersupply in the bottom left to the back of the DVD drive in the top right and its a becoming a work of art soon. Just it needs to be functional too!
 
I'm pretty sure the TMS-200 won't fit in a 3.5 bay. Take a look at it and its specs on the Koolance website and see if you can fit it in somewhere, keeping in mind you need to run all the sensors, fan control cables, USB lead, and a 4-pin molex power connection to it. The TMS-200 supports 2A per pump or fan channel, with a board total of 6.5A. That should be enough for your pump, but the koolance page isn't clear about official support for non-koolance pumps; check the xtreme systems forums for other users of the TMS-200 and non-koolance pumps, if you want. Or write Koolance and ask them.

The mCubed TBalancer can fit in a 3.5 bay, but I haven't installed that. You'll need the analog sensor hub add-on to control a pump. It is also software-controlled and does not have any knobs or whatnot for external control. It's software is a bit older than the TMS-200, but some people still use it.
 
In the top picture it looks like its in a 5.25 bracket and has quite a bit of room...so possibly...widthways. The length might be a problem. But luckily it is black, which would fit in well with my black black case. 2A should fit nicely I think its maximum is 1.98.

The word analog has always worried me! But for things like this something slightly older proabably wouldn't be a problem.

I'm going to get something like this when its tube and fluid change time. The current pump is sucking the tubing (tygon) out of shape. I've put coils over it,but it doesn't look so nice. So this might happen sooner rather than later. Or when the GPUs get water, which would free up the slots anyway. Or Vapor. If eraser wasnt so ugly...
 
Back
Top