DigiFlow 8000T Flow Rate Meter Review

[Martinm210]

I found one of these on ebay for $35, and since it came with threaded 3/8" NPT, I was happy to know that I could put 5/8" barbs on it.

There were not much for specs on it other than is said .4-2.0 GPM with 3.5GPM max and the 3/8" NPT ports.

When after fitting it with some nylon 5/8" x3/8" NPT barbs, this is what it looked like:
Front

Back:

What you get (Minus the barbs)

The display in operation. The little line on top sort of counts segments as the turbin spins. While GPM reads to the .01GPM it actually jumps about .1 GPM between steps, so it's really only resolved to about .1GPM in reality.

The turbine (Sorry for the poor picture, it's about a 3/8" Diameter turbine):

Internals: (Could easily be taken apart and the LCD screen, battery pack, and speaker could be easily relocated)


Test Results (It had around 8-15% error). It's good enough for general info, but not for pressure drop testing or something that needs more than that amount of accuracy.


Pressure Drop Guestimate:
I can usually get a pretty good guess at how restrictive something is, by blowing through it. I'm still waiting for my digital manometer to get here, so in the mean time...my guestimate is a "Fairly free flowing" rating. Seems to be less restrictive than my MCW-60 sitting here on the bench, and probably more restrictive than an MCW-30. So...restriction isn't too bad and probably better than most flow rate meters.

Anyhow, I don't know that I would run one in my system all the time, but it's usefull to get an approximate reading and might make a fun project to tear apart and put into a custom fabricated case.:thumb:

Update 12-15-07 Pressure Drop Test Complete

I did a whole bunch of pressure drop testing today, this being one of many. Turns out this flow meter is actually LESS restrictive than an MCW-30. I also went ahead and captured flow rates from the King Instruments Flow meter as a means to completely retest flow rates again from scratch. What I found was nearly identical, instead of my previous average of 11% high, I got 13% high...that's some good consistency..

On with the pressure drop results: I included my MCW-30 curve for a relative, which happens to be the most free flowing block to date. So, this little flow meter, while having a tendency to read fairly high, doesn't have much of a hit on your flow rate...so that's good.

[joecop120]

Would that be able to be rewired to accept power from a 3v PSU rail?

[Exahertz]

Would that be able to be rewired to accept power from a 3v PSU rail?

it says it takes two AAA batterys, so im sure you could solder wires to the + and - on the circuit and connect to the PS.

BTW, exelent review! ive been looking for a flow meter and i kind of like this one!
I wonder if you could take it apart and move the whole circuit and lcd to some other part of the case and leave the in-line sensor intact? it looks like only two wires and could be easy.

[MrToad]

Thanks for the review ^^

Price-wise you can't possibly beat it. More so when you consider that it has it's own display.

However, as you point out, accuracy-wise it can't compete with the digimesa ones. They have an average error of +-2% and slightly lower pressure drop @ 2 gpm than a MCW60.

And of course, if you can put up with the horrendous pressure drop, you can't beat the accuracy for your money of the SFF800.

Still, a funny gadget.

[Martinm210]

Yeah, that's the part that was suprisingly nice about this one. You could very easily take the unit apart, wire the power to your 3V, and mount the LCD in the computers front panel.

I'll follow up with the real pressure drop when I get my manometer, but it's not too bad. The accuracy really isn't there for precision work, but for a general system monitor and because the pressure drop is fairly low, it's a pretty cool little project for anyone with a soldering iron and some creativity.

It also has quantity options, so you could set an alarm after X number of gallons or days to remind you of something. I think that function is really designed for filter changes in aquariums, but I suppose if you wanted a reminder to change your coolant after 5000 gallons, you could do that. Too bad you couldnt set it up for a minimum flow rate or something like that though...

[headala]

I'm assuming the display isn't backlit?

[Martinm210]

I'm assuming the display isn't backlit?

No..unfortunately it is not, just black letters.

[Martinm210]

Martin, have you plugged all the numbers for you loop into that spreadsheet to see how accurate it comes out?

Yeah, I've tested both loops before and each one was accurate with less than 10% error of the estimator.

I'm also awaiting delivery of a King Instruments flow meter that's supposed to be accurate within a couple of percent, so this one is just a toy...

[welshtom]

What do you use to measure the flowrate for comparison?

I guess you are basing this ones accuracy on the accuracy of your own instrument which seems flawed to me?

[Martinm210]

What do you use to measure the flowrate for comparison?

I guess you are basing this ones accuracy on the accuracy of your own instrument which seems flawed to me?

I use a graduated 5 gallon reservoir that I made myself and a stopwatch. I stubbed out a vertical section of tubing and graduated gallon marks on it. Then I use a stopwatch to measure time and typically measure from the 1 gallon mark to 4 gallons.

Simple, but effective and accurate too. Only problem is it takes soo long to get lower end flow rates.

[welshtom]

Thanks for the info

[Martinm210]

Sure,

I think a good portion of the variability in the error is due to the inability of the meter to actually vary by .01 GPM, instead it has defined steps that it reads out that jump a little more than .1 GPM.

A couple of my tests it would jump back and forth between the two, so I averaged those. In others it would hold constant, but because the steps are only .1GPM apart, anything that is within .05 or so GPM would be acceptable for a specific step.

There may be something to the barbs used too. I used 5/8" barbs which are good flow, but there is still a fairly abrupt step at the end of the threading, perhaps this added turbulence is increase the velocities along the wall perimeter.

Actually the more I think about it, perhaps the cold temperature of the water I used has a bigger role in the error of my results. The water used was cold tap water at probably 10C which has a dynamic viscosity around 1.31 (cp), where as a typical aquarium probably runs at a much warmer 26C with a viscosity of .862 (cp).

I should run the test again at a more typical computer water temperature 24C and a 5% ethelene glycol mixture to get a more accurate error representation to what we would see in a system. That's just alot of work...

[The_Beast]

It seams pretty bulky and ugly

[Exahertz]

It seams pretty bulky and ugly

Very true. But i think it could be dissasymbled, display moved someware in the case and the inline sensor unmounted from the cover and rewired.

[Martinm210]

Yeah, with nothing more than some very basic soldering skills to extend the wires, you could easily take apart the turbine sensor and relocate the LCD somewhere more convenient.

Also since in runs on two AAA batteries, you could also easily tie that into the 3V power feed on the PSU, not that it takes much power...but at least you'd never need to change batteries that way.

I'll post some updated results when I get my digital manometer so you can all see the pressure drop, and I may redo my error graph with warmer water to more closely match a water cooling system. The more I've thought about it, the more I think the increased viscosity of cold water could be throwing my results off and it's more accurate than that. It's at least worth a couple of verification tests. Most digital flow meters are no better than 5% accuracy anyway.

That's what I would do if I had planned to run one in a system permanently. These may be good indicators of the health of a system, like if your CPU block started plugging up full of debris, or the pump was starting to die. A flow meter even if not accurate, would provide a good relative indicator for potential problems.

[Exahertz]

Also since in runs on two AAA batteries, you could also easily tie that into the 3V power feed on the PSU, not that it takes much power...but at least you'd never need to change batteries that way.

If you remove everything from the case you would have to tie it into the 3v rail, considering that the battery bank is part of the case.

[Boogerlad]

what wires do use the 3.3 volt line these days? In other words, where can u hook up the meter to the 3.3 volt line on the psu?

[Martinm210]

what wires do use the 3.3 volt line these days? In other words, where can u hook up the meter to the 3.3 volt line on the psu?

I havn't done it and would use a multimeter to confirm, but this seems to indicate the orange colored wires are the typical +3.3V, so you'd have to patch into the main connector or a SATA connector:

http://www.highpowersupply.com/hpc5x0A12C.html

[Exahertz]

thats kind of a touchy subject tapping into the ATX power, but with such a low power draw im sure it wouldnt have any effect on the motherboard

[ladderman]

thats kind of a touchy subject tapping into the ATX power, but with such a low power draw im sure it wouldnt have any effect on the motherboard

Just found the same unit over here in the UK, but as usual it’s twice the price £35.00 = $72.40.

http://www.bellflowsystems.co.uk/LCD...m-p-19867.html

I’ve ordered a Swissflow SF800 flowmeter direct from Swissflow for 25 Euros = $37.10.

http://www.swissflow.com/en/SF800

When it comes I intend do some flow tests to see how restrictive it is before I decide to use it. Mine will be pretty basic testing. Like filling a container and see how long it takes to empty a certain volume of water, (With and without the Sensor).

All I want the sensor to do is shutdown the system if the D5 pumps fail

Just while I on the subject of using flowmeters. The only way you can use a flowmeter to shutdown your system on a EVGA 680i mobo is to use a mCubed Extension Set – analog sensor hub. Because there is no option in the BIOS to set a shutdown using the CPU fan RPM header.

In fact there are no options in the BIOS System Monitor to do any form of system shutdown if your temps get too high.

I’m sending an e-mail to EVGA too find what happens if the CPU temp get too high.

[Marci]

Pressure drop specs etc for the Swissflow SF800 can be found in procooling.com's forums somewhere...

[MrToad]

Pressure drop specs etc for the Swissflow SF800 can be found in procooling.com's forums somewhere...

Pressure drop curve for the SFF800:

http://www.xtremesystems.org/forums/...3&postcount=19

Rest of the relevant specifications can be found on the manufacturer's web page:

http://www.swissflow.com/en/SF800/Fl...Specifications

[headala]

Make sure that the mCubed can handle the high pulses/liter that the swissflow puts out. My AC Aquaero can't. :-(

[Xilikon]

Make sure that the mCubed can handle the high pulses/liter that the swissflow puts out. My AC Aquaero can't. :-(

True, but I heard someone made a converting board so it can work with the Aquaero and I'm really curious myself about how to make one so I can add a SF800 in the future with the Aquaero setup.

[headala]

If you find that, I would love to build/buy the converting board. Please PM me if you find anything out about that.