Can fan controllers/monitors reliabily report fan speeds down to 100RPM?

[Cathar]

I have a liquid flow meter that sends out pulses much like fans do.

Now it sends out 1 pulse per 10ml of liquid that flows through it. I need to be able to get a reliable reading on liquid flow rates down to 1 litre per minute, which if we do the math, equates to an effective fan speed of 100RPM. This is kinda nice, because if I saw, say, 432RPM, that would equate to 4.32 litres per minute.

I tried plugging the flow meter into my motherboard's fan headers, and get fluctuating readings which is undesirable. The sample period of the fan pulses seems to be too low. If I watch it for long enough I found that the highest value that it reports is the actual flow rate and it will fluctuate with values below that point.

What I wondering is if there are any of those fan controller/monitor devices out there which can reliably report fan speeds down to 100RPM? Anyone with any experience with seeing their unit being able to report such? Any help is greatly appreciated.

[fareastgq]

Uh, I don't think so man, every fan controller I have ever had varies so much in voltage that the rpm's jump up and down way too much for what you want, you'll prob have to design a circut. Most likely the rheostat/bus type controllers will prob be better than any digital ones though. a high quality rehostat would prob do the trick for u. maybe none of the fan controllers on the market are even good enuff for what you want to do.

[greenman100]

Uh, I don't think so man, every fan controller I have ever had varies so much in voltage that the rpm's jump up and down way too much for what you want, you'll prob have to design a circut. Most likely the rheostat/bus type controllers will prob be better than any digital ones though. a high quality rehostat would prob do the trick for u. maybe none of the fan controllers on the market are even good enuff for what you want to do.



read his post again.

He doesn't need the power capibility od the fan controller, only the RPM reading.

That said, I haven't seen one display that low. I'm surprised it doesn't work, though.

[Marci]

Cathar - Lemme check this T-Balancer over tomorrow and I'll let you know

[MaxxxRacer]

T-balancer???

[Marci]

Heh - http://www.mcubed-tech.com/eng/produkte.htm m8... they're superb. I'm in love with them other than the lousy flow meter that they supply - more designed for Euro-NarroBore systems than our style ;)

However, can use a Swissflow with em... :)

Not accurate enough for MAJOR testing as only accurate to 0.5 deg C afaik on the temp measuring front, but for programming in automation - ie: if liquid temp = more than 30 run fans at 100%. If liquid temp = 28, fans at 90% etc.

You appoint any one of it's 16 possible temp probes to a fan channel, and program in a response curve specifying "AT X temp FANSpeed = Y%"

Can also hook lights up to the fan channels and put in programmed lighting patterns.

Partner this with a MatrixOrbital and the relevant plugin and you can pass all your temps over to an LCD Display for full temp monitoring.

T-Bal also sits between your ATX ON switch and your mobo, so u can set switch-off points based on temp of water, flow rate, temp of whatever u have sensors on.... they really are a superb piece o' kit...

Have a look here - s'wot I used for the PA vs HE Series thang...

http://www.over-clock.com/ivb/index.php?showtopic=13545

I originally stumbled over them when looking for phasechange controllers with the capabilities of the commercial offerings used by Asetek and ECT in the VapoLS and MachII's... not suitable for that job sadly, but for watercooling / TEC cooling protection and monitoring I've never had anything more useful.

[fareastgq]

read his post again.

He doesn't need the power capibility od the fan controller, only the RPM reading.

That said, I haven't seen one display that low. I'm surprised it doesn't work, though.

I guess I was a bit unclear, I know he doesn't need the power capability, I'm saying most retail units will not be precise enough for what he wants to do.

[greenman100]

I guess I was a bit unclear, I know he doesn't need the power capability, I'm saying most retail units will not be precise enough for what he wants to do.


then why diod you say the voltage of the fan controller would jump up and down?

he's not using the voltage out on the controller, only the RPM monitor...?

[Marci]

K... struggling to find the answer to that question as I don't have anything that'll actually produce 100rpm for me to attempt to measure. Just wondering, can pulse be simulated by hand in any way???

[Marci]

OK... I'm an arse. Should've just read the product specs page...


0 rpm speed possible: if temperature rises, the controller speeds up the fan according to the response curve

Yes, it reads all the way down to 0 so will happily read 100rpm. Wish I'd have known that earlier - could've popped one in the box I just sent out to ya d00d! If getting one over there's proving diccifult, drop me a line and I'll send one over with "the missing next-step" ;)

[fareastgq]

then why diod you say the voltage of the fan controller would jump up and down?

he's not using the voltage out on the controller, only the RPM monitor...?

because it does jump up and down. If you were going to use just the digital output, why bother getting any kind of fan controller? He said all he did was plug it into a fan header to get a reading, to me that implies he wants to use the whole thing, hence, the fluctuations which is exactly the same problem he alrdy had before.

[greenman100]

because it does jump up and down. If you were going to use just the digital output, why bother getting any kind of fan controller? He said all he did was plug it into a fan header to get a reading, to me that implies he wants to use the whole thing, hence, the fluctuations which is exactly the same problem he alrdy had before.
ok, let me try again.

a fan header has 3 pins. +12v, GND, sensing.

he is using a flowmeter. all this does is output a series of pulses. It is not dependent on the 12V source.

he said fan controllers/monitors because many fan controllers have displays.

[Marci]

Flowmeter is unpowered. All it has to do is send a pulse to anything capable of receiving an RPM, be it a fancontroller with builtin rpm monitor, mobo rpm monitor... anything capable of monitoring that pulse, but capable of doing so at 100rpm. A lot of cheap-end rpm monitors built into fanbuses don't show less than 700rpm and other such figures...

The fluctuations Cathar talks of are in the rpm's that the mobo reports, not the power it supplies as he isn't drawing any... the mobo just isn't capable of reading sensitively / slowly enough to accurately report 100rpm... I've seen some that just sit and flip from 0rpm to 600rpm as fan drops below and above 600... whereas Cathar needs something able to report right down to 1rpm if needs be without falling foul of the monitoring chips' capabilities.

All the flowmeter has at the end of it's wire is yellow (pulse) and black (grnd) - no red.

[fareastgq]

thx for explaining that marci. btw, ygpm

[Cathar]

Thanks for the discussion guys, but it seems like I may just end up doing something with Linux and a serial port. Hook up the flow meter output to the RX pin of a serial DB9 connector, and write a little app to poll it in real-time, or get the OS to notify the app in an interrupt driven fashion.

[Marci]

How is the pulse generated tho d00d? As in in general on fans, pumps, flow meters... is it always an electronic pulse that's specific..? or could a pulse be recreated in analogue fashion? Just curious for various purposes... not got round to googling-up on it yet...