Is fan port safe for pumps?

[Dave]

Get alot of email asking same question, so here is a simple example how to determine if this is so.

1. Try to find the spec's of you MOBO, in manual or there WEB site, you need around 1A average or

2. Find the spec's of your processor stock fan- this is easy to find on AMD or Intel WEB site.

I will use Pent 4 -775 for example.

ftp://download.intel.com/design/Pent...s/30255303.pdf

Sections 5.1.5 "Fan performance for Active Heatsink"
Peak fan current draw - 1.5A
Average - 1.1A
Start- 2.2A

As Mag uses under 1A, you should be fine. In fact you can likely use a DDC off this fan port (just a suggestion, I am not a Laing rep).

If your not sure, use a 3 to 4 pin adaptor.
However as you can see, the current requirement of processor fans is increasing.

[pHaestus]

And I remember when my socket A Asus boards were impressive for providing 1A (12W). Time marches onward

[BillA]

Swiftech sells the MCP350 with a Molex type connector, that is the intended power source
fan headers are not recommended because there is no necessity to add risk

[Marci]

Any chance of seeing the DDC / MCP350 with rpm monitoring cable separate tho to run to mobo? Just makes life handy for utilising bios emergency shutdown on "fan" failure... would love to see every 12v pump with some form of RPM header that's usefulness is purely for determining alive or dead status rather than any form of particular flow rate... or have I missed that conversation?

[BillA]

Originally Posted by Marci

Any chance of seeing the DDC / MCP350 with rpm monitoring cable separate tho to run to mobo? Just makes life handy for utilising bios emergency shutdown on "fan" failure... would love to see every 12v pump with some form of RPM header that's usefulness is purely for determining alive or dead status rather than any form of particular flow rate... or have I missed that conversation?

http://www.swiftnets.com/products/mcp350.asp
"RPM sensor plugs to 3 pin motherboard fan connectors and reports impeller rotational speed (effective 1-24-05)"

thats how it is configured (on a separate lead), need to add a photo showing the connector

[Dave]

Given the current output required for a 775 board, is there really any risk Bill?

[BillA]

not how I consider it
what is the need/benefit in using the fan header ?
and if one fan header is ok, then all fan headers will be considered ok irrespective of any specific board evaluation
not worth it, may pick up consequential liability

[Dave]

"what is the need/benefit in using the fan header ?"

Direct RPM / CPU control, we have examples there of.

[BillA]

ok
is the cooling better ?
is the noise actually lower ? (something perceptible outside of a closed case)

??

[Dave]

Bill, I think you need to understand, it is up to the customer to decide if a feature is wanted, not you.

On board fan control will allow a reduction in pump noise, based on demand.

Do I think it is needed?

My opinion doesn't mean anything, if the customer wants this feature.

You asked "why", and I answered, if you what to argue, find someone else.

I find the power savings, and possible component life extension is more appealing then noise, but this is just my opinion and interest.

[BillA]

ok

"You asked "why", and I answered, if you what to argue, find someone else."
to question (me) and to answer (you) is "to argue" ?

I shall find someone else I suppose

customer perception is an aspect of marketing
I was not asking about marketing, simply if using a fan header yielded a demonstrable technical benefit
which you answered "My opinion doesn't mean anything, if the customer wants this feature."

no argument here
??

[jaydee]

Man, I hate it when I see from a manufacture "it is up to the customer to decide if a feature is wanted".... This how all the idiots of the world control the world... No offence intended Dave.

[csimon]

The fan header rpm monitoring feature is nice to have and use just as if it were a cpu mounted fan itself. It can be monitored by the bios and shut down the system or prevent it from booting in case of failure. It can also provide an assurance of it's working order to the customer via the rpm's since the pump itself is inaudible. Inaudible to me anyway. The accuracy of the rpm's is as unimportant to me as is the accuracy of the cpu/sys temp's being monitored ...but it sure helps in managing the overall system at a glance.

Thanks again Dave.

[jaydee]

Originally Posted by csimon

The fan header rpm monitoring feature is nice to have and use just as if it were a cpu mounted fan itself. It can be monitored by the bios and shut down the system or prevent it from booting in case of failure. It can also provide an assurance of it's working order to the customer via the rpm's since the pump itself is inaudible. Inaudible to me anyway. The accuracy of the rpm's is as unimportant to me as is the accuracy of the cpu/sys temp's being monitored ...but it sure helps in managing the overall system at a glance.

Thanks again Dave.

Yeah, that's been covered. The question was using a mobo header to power the pump not just monitor rpm.....

[Ls7corvete]

They pay the $, give em what they want. They blow their mobo running too much current so be it. This is WCing, I think too much fan header load is the least of the problems someone will have if they are incapable of following recommendations.

[nicozeg]

Originally Posted by unregistered

what is the need/benefit in using the fan header ?

Lot cleaner cable setup. Crucial feature for the modding market.

[Razor6]

Using software to control the fan header and consequently the pump, now that sounds like it could be really useful. Constant performance/noise management using the CPU temperature seems like a marketable feature. Are there any problems with electrical interference from the pump going directly to the board?

[Etacovda]

Originally Posted by Ls7corvete

They pay the $, give em what they want. They blow their mobo running too much current so be it. This is WCing, I think too much fan header load is the least of the problems someone will have if they are incapable of following recommendations.

all it is, is more chances for idiots to mouth off after doing something stupid and giving a product a bad name.

is it the products fault? not strictly.
Is it going to be blamed on the product or company? more than likely...

power savings on something that uses 7w? lol...

Unless its CONSIDERABLY quieter, its adding complication and another point of failure for nearly no benefit.

[Marci]

thats how it is configured (on a separate lead), need to add a photo showing the connector

Forgot to add into that the MCP655 / D5... as you're the man in the know when it comes to Laing's products have they been poked to add to that model as well? Would be the icing on the cake for the D5, and was amazed they didn't put it on there already...

Anyways, simple solution to the above 3pin header argument is, supply the pump with Molex and RPM connector, and 2x spare pins for the RPM header so u can snip and solder them on and convert it to 3pin should the end user so desire... just means dropping the remaining pins into the box...

As imo if a customer can't solder two bits of wire together then they shouldn't own a waterkit (just on a basis of iq level / ability - badly phrased but can't think of right phrase)...

Or a simple adapter to accept molex and 3pin with rpm only and convert down to single 3pin...? Add into paperwork a "use at own risk" statement covering why it isn't like that to start with...

[Dave]

I do not believe everyone understands what feature it is that Bill and I are discussing.

As you know, there are external fan controls, well there is now onboard fan controllers as well.

We designed the MAG to work with these systems, Bill thinks that is a waste of time, and I believe if the customer requests it, as they have in the past (even on this board) we should include said feature.

I never liked "stand by" mode in computers, as the OS always seem to screw something up, but with XP, we use this mode all the time and never turn systems off. This feature allows the pump to slow down or even shut off, just as the system controls the fan.

And yes, we submitted a patent request on this feature.

[Marci]

Aye but requirement for this feature is minimal from the end-user in this particularly scene... most just want a simple pulse that the bios can recognise as fan working or not working (pump working or not working). If not working, switch PC off. IE: if pump fails and doesn't send more than 2200rpm signal (most mobos tend to whine at 2200 or below depending on manufacturer) then doesn't let PC boot, or if occurs during use and rpm drops below threshold, PC switches off.

That's what most people want an RPM feature on a pump for. VERY few ppl want to slow a pump down, especially one that's silent to start with... economics wise, the power saving is also minimal when you're talking the kind of consumption these particular pumps take...

Very few folks in this scene use "standby"... it's PC on with screensaver, or PC off... PowerSaving features are put into play by enduser purely for noise based issues - slowing fans, but if pump is silent to start with (which the mag is) then it doesn't need slowing. Most go to watercooling so they can leave it running 24/7 downloading pirate crap tbh without it keeping em awake at night, or leave it folding 24/7, so standby isn't an option for them... and that covers a big chunk of the market...

RPM builtin to pump is far more convenient than flow indicator/meter. Most end-users only want a flowmeter to get an RPM to use the bios emergency shutdown feature, not to measure actual waterflow or control pump speed...

[Dave]

^ Again I agree, however I fully and openly state, when our test systems go into "stand by".... you can clearly hear the pump.

The test rigs we have, turn power supply fan off as well, just like a laptop.

I think this feature is likely more welcomed with the combo unit, which is a direct OEM replacement for CPU fan / sink.

[Marci]

Now that's wierd. Two CSP-Mags in parallel at full 12v off a passive PSU, mounted in a case on antivibe mountings and I can't tell they're on - s'literally perfectly silent... no fans in the system whatsoever...

[Long Haired Git]

Power in my area/house is too shitty for standby, but since I work on laptop I use hibernate a fair bit on all my PCs.
I, personally, don't care about an RPM sensor on the pump. All I need is the CPU temp sensor hooked to shutdown. That way if ANYTHING goes wrong, then it shuts down.
KISS and all that.
There is "silent" and then there is "silent".
I do like the idea of slowing down a pump.
I am not after overclocks, I am after silence, as I do some late night computing where you can hear pins dropping. Lots of other guys have their PC in their bedroom.
I can see a benefit of having the pump able to be controlled to slow down and go from "9pm TV on in another room, odd car driving past" quiet to 4am quiet.

[Dave]

Figured as much Marci, I think it is the MOBO itself transmitting the vibrations, and producing the volume

I know it will not pass our OEM clients spec, unless we include control or I figure out how to solve transmit issue.

It would be easy if it was just bearing noise, but it appears to be simple pumping production.


Ok- Dave is busy for awhile
LTR