Water Flow Meter

kalvnhobbz · 06-03-2002, 04:44 PM

http://www.highspeedpc.com/index.htm?item135.htm

It's a flow rate meter that'll plug into a motherboard to give RPM monitoring. Seems like it'd go a long way in saving your cpu if the pump fails. It's expensive though... and I sort of wonder how it the paddle attaches to the outside electronics without leaking.

Plus a blue paddle would be much better than red.

TonyM · 06-03-2002, 05:31 PM

seems kind of expensive just to detmine flow, but maybe its just cause im poor.....

mhayenga · 06-03-2002, 05:41 PM

I agree with tony....

Flowmeter, excuse to upgrade, flowmeter, excuse to upgrade, flowmeter, excuse to upgrade!

Plus I've heard that the flowmeters really cut down the flowrate of the system.

bigben2k · 06-03-2002, 07:32 PM

I've seen that flow-"meter" on an european website. Glad to see it's here in the USA! But I surely won't pay $40 for one.

I highly doubt that it is accurate for measurement, but it is definitely good for sensing wether or not there is flow. I don't think that the RPM can actually be equated with flow, but I could be wrong. There's no info on this.

It is driven by magnetic field, magnets being imbedded in the impeller. It also has 0.2" ID connections, which means that it would impede flow, especially with 1/2" ID tubing (0.5").

$40? I don't think so... I'm still searching.

kalvnhobbz · 06-03-2002, 09:40 PM

http://www.overclockers.com/tips642/

something similar there, but not connected to rpm monitoring. also pretty complicated.

anyone know how to even make this (the one with rpm, not the one i linked to)?

Divided · 06-03-2002, 10:26 PM

both the homemade version and the over-priced online version use the same concept..

ah.. damn i forget what its called -- the ______ effect.

basically like bigben2k? said.. its all about magnetic fields.

think RPM monitoring -- oh but it would be horribly inaccurate

no way to equate the RPM of that little paddle to GPM or GPH.

Lmandrake · 06-03-2002, 11:03 PM

It is a hall-effect sensor. They come in different flavors, but the most common is one that switches on when a magnet passes by and switches off when the magnet passes by going in the opposite direction.

I am going to be doing a review of the unit sold by HighSpeedPC in a couple of weeks for www.pcabusers.com

The one I made for myself paid for itself one night when low coolant killed my flow from my pump...

crane · 06-04-2002, 10:00 AM

LMandrake... I am attempting one of your flow sensors, and I have a question. The Hall sensor has 4 leads coming out of it, correct? When I put a continuity tester on 2 outside terminals and pass a magnet by the sensor I get a brief flicker on my tester. When I attach the continuity tester to the other 2 leads, I show no sensor activity. Possible I have a crapped out sensor, huh. What I'm getting at is what possible testing/troubleshooting can I apply to these hall effect sensors?
I probably burnt the sensor using a powered tester, correct?

BrianH · 06-04-2002, 10:18 AM

http://www.gemssensors.com/TOCResult...nContentsID=74

These look nice... very expensive

gmat · 06-04-2002, 10:28 AM

1 - generally hall effect sensors need to be powered. Usually at high voltage / low current. That explains the 4 leads.
2 - if you calibrate the sensor you can relate RPM to flow rate - because basically it's a volumetric sensor (most flow sensors work this way BTW). The accuracy loss will then be due to turbulence passing between the impeller and the casing without pushing the impeller in the right direction. This will only happen at very low / very high flow. Thats why most volumetric sensors have an operating range...

[edit] typos + voltages fumble

Lmandrake · 06-04-2002, 10:50 AM

Boy, I have killed a lot of fans in my quest for knowledge on these sensors. I doubt that you killed yours by using a powered voltmeter.

Most I have seen have three leads, positive, negative, and output. Some have four leads, positive, negative, and two outputs. I don't know how you can tell which one are which without the circuit board it was used on or a data sheet.

The first thing I would suggest is that you look for numbers on the hall effect chip itself and do a google search with the numbers. That may lead you to a data sheet on the chip. If not, the sensor you have may be one where almost all of the fan control circuits are integrated on the chip. I am using one of those now since I ruined my last one by handling it too roughly and breaking it off the circuit board. In the four pin sensor I have, there are positive and negative leads in for power. There are two outputs.

The outputs are not what you may expect. If this is a chip where the whole fan control circuit is on the chip, the outputs are probably not "hot" or positive outputs. They are, instead, tied to ground. When used in a fan, which usually has two coils, the 12v "hot" or positive lead is tied to one side of each coil and the other side of each coil is tied to these outputs. The fan control circuit, which is essentially two transistors controlled by the hall sensor, activates the coils by providing a path to ground (the negative side). When the coil is not activated, the output for that coil stays "high" - i.e. near or at the "hot" or positive input voltage. With no voltage differential across it, the coil does nothing. To activate that coil, the chip output goes "low" - i.e. close to ground and the voltage differential across the coil becomes great enough for it to become active and do work.

I would suggest that you either find data on the chip or experiment on the hypothesis that the two "output" leads are paths to ground. If you pulled the chip off a fan circuit board that just had circuit traces on it and not any or many components, I would bet that the whole fan circuit is in a single chip and it operates as described above. If you have some LED's with integrated resistors or LED's wired in line with resistors so they won't blow, you should try them with their "negative" sides tied to the chip outputs.

I have played with far too many hall sensors (I have a pile of discarded AT power supplies that I pull fans out of) and it is possible to burn up the fan circuit by operating it without a load, but I have only burned up one that way.

LMK if this helps. BTW, if you, or anybody else is interested, I would be willing to share how to use this thing to automatically shut down your box if flow stops.

06-03-2002, 04:44 PM	#1
kalvnhobbz Cooling Neophyte Join Date: Feb 2002 Posts: 27	Water Flow Meter http://www.highspeedpc.com/index.htm?item135.htm It's a flow rate meter that'll plug into a motherboard to give RPM monitoring. Seems like it'd go a long way in saving your cpu if the pump fails. It's expensive though... and I sort of wonder how it the paddle attaches to the outside electronics without leaking. Plus a blue paddle would be much better than red.

06-03-2002, 07:32 PM	#4
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	I've seen that flow-"meter" on an european website. Glad to see it's here in the USA! But I surely won't pay $40 for one. I highly doubt that it is accurate for measurement, but it is definitely good for sensing wether or not there is flow. I don't think that the RPM can actually be equated with flow, but I could be wrong. There's no info on this. It is driven by magnetic field, magnets being imbedded in the impeller. It also has 0.2" ID connections, which means that it would impede flow, especially with 1/2" ID tubing (0.5"). $40? I don't think so... I'm still searching. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

06-03-2002, 10:26 PM	#6
Divided Cooling Neophyte Join Date: May 2002 Location: NY, USA Posts: 81	both the homemade version and the over-priced online version use the same concept.. ah.. damn i forget what its called -- the ______ effect. basically like bigben2k? said.. its all about magnetic fields. think RPM monitoring -- oh but it would be horribly inaccurate no way to equate the RPM of that little paddle to GPM or GPH. __________________ Pentium IV 1.8A NW @ 2.65Ghz Alpha Heatsink w/ Panaflo H1A ASUS P4S533 Mobo 512Mb Mushkin Level2 DDR-SDRAM VisonTek GeForce3 Ti500 (254/540Mhz) 2 x 60Gig Seagate HD's W.D. Caviar 8Mb Buffer - 80Gig HD HighPoint Rocket404 Raid Controller Plextor 40/12/40A CD-RW

06-03-2002, 11:03 PM	#7
Lmandrake Cooling Neophyte Join Date: Nov 2001 Location: D.C. Posts: 26	It is a hall-effect sensor. They come in different flavors, but the most common is one that switches on when a magnet passes by and switches off when the magnet passes by going in the opposite direction. I am going to be doing a review of the unit sold by HighSpeedPC in a couple of weeks for www.pcabusers.com The one I made for myself paid for itself one night when low coolant killed my flow from my pump... __________________ The technology is within the reach of even the smallest nuclear power. Last edited by Lmandrake; 06-03-2002 at 11:11 PM.

06-04-2002, 10:00 AM	#8
crane Cooling Neophyte Join Date: May 2002 Location: so cal Posts: 82	LMandrake... I am attempting one of your flow sensors, and I have a question. The Hall sensor has 4 leads coming out of it, correct? When I put a continuity tester on 2 outside terminals and pass a magnet by the sensor I get a brief flicker on my tester. When I attach the continuity tester to the other 2 leads, I show no sensor activity. Possible I have a crapped out sensor, huh. What I'm getting at is what possible testing/troubleshooting can I apply to these hall effect sensors? I probably burnt the sensor using a powered tester, correct? __________________ dirt rules

06-03-2002, 05:31 PM	#2
TonyM Cooling Neophyte Join Date: Mar 2002 Location: Mount Prospect Posts: 78	seems kind of expensive just to detmine flow, but maybe its just cause im poor.....

06-03-2002, 05:41 PM	#3
mhayenga Cooling Neophyte Join Date: May 2002 Location: Earth Posts: 46	I agree with tony.... Flowmeter, excuse to upgrade, flowmeter, excuse to upgrade, flowmeter, excuse to upgrade! Plus I've heard that the flowmeters really cut down the flowrate of the system.

06-03-2002, 09:40 PM	#5
kalvnhobbz Cooling Neophyte Join Date: Feb 2002 Posts: 27	http://www.overclockers.com/tips642/ something similar there, but not connected to rpm monitoring. also pretty complicated. anyone know how to even make this (the one with rpm, not the one i linked to)?

06-04-2002, 10:18 AM	#9
BrianH Cooling Savant Join Date: May 2002 Location: Ottawa, Ont, Canada Posts: 115	http://www.gemssensors.com/TOCResult...nContentsID=74 These look nice... very expensive

06-04-2002, 10:28 AM	#10
gmat Thermophile Join Date: Mar 2001 Location: France Posts: 1,221	1 - generally hall effect sensors need to be powered. Usually at high voltage / low current. That explains the 4 leads. 2 - if you calibrate the sensor you can relate RPM to flow rate - because basically it's a volumetric sensor (most flow sensors work this way BTW). The accuracy loss will then be due to turbulence passing between the impeller and the casing without pushing the impeller in the right direction. This will only happen at very low / very high flow. Thats why most volumetric sensors have an operating range... [edit] typos + voltages fumble Last edited by gmat; 06-04-2002 at 10:30 AM.

06-04-2002, 10:50 AM	#11
Lmandrake Cooling Neophyte Join Date: Nov 2001 Location: D.C. Posts: 26	Boy, I have killed a lot of fans in my quest for knowledge on these sensors. I doubt that you killed yours by using a powered voltmeter. Most I have seen have three leads, positive, negative, and output. Some have four leads, positive, negative, and two outputs. I don't know how you can tell which one are which without the circuit board it was used on or a data sheet. The first thing I would suggest is that you look for numbers on the hall effect chip itself and do a google search with the numbers. That may lead you to a data sheet on the chip. If not, the sensor you have may be one where almost all of the fan control circuits are integrated on the chip. I am using one of those now since I ruined my last one by handling it too roughly and breaking it off the circuit board. In the four pin sensor I have, there are positive and negative leads in for power. There are two outputs. The outputs are not what you may expect. If this is a chip where the whole fan control circuit is on the chip, the outputs are probably not "hot" or positive outputs. They are, instead, tied to ground. When used in a fan, which usually has two coils, the 12v "hot" or positive lead is tied to one side of each coil and the other side of each coil is tied to these outputs. The fan control circuit, which is essentially two transistors controlled by the hall sensor, activates the coils by providing a path to ground (the negative side). When the coil is not activated, the output for that coil stays "high" - i.e. near or at the "hot" or positive input voltage. With no voltage differential across it, the coil does nothing. To activate that coil, the chip output goes "low" - i.e. close to ground and the voltage differential across the coil becomes great enough for it to become active and do work. I would suggest that you either find data on the chip or experiment on the hypothesis that the two "output" leads are paths to ground. If you pulled the chip off a fan circuit board that just had circuit traces on it and not any or many components, I would bet that the whole fan circuit is in a single chip and it operates as described above. If you have some LED's with integrated resistors or LED's wired in line with resistors so they won't blow, you should try them with their "negative" sides tied to the chip outputs. I have played with far too many hall sensors (I have a pile of discarded AT power supplies that I pull fans out of) and it is possible to burn up the fan circuit by operating it without a load, but I have only burned up one that way. LMK if this helps. BTW, if you, or anybody else is interested, I would be willing to share how to use this thing to automatically shut down your box if flow stops. __________________ The technology is within the reach of even the smallest nuclear power. Last edited by Lmandrake; 06-04-2002 at 11:39 AM.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)