DigiFlow 8000T Flow Rate Meter Review

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
http://www.overclock.net/gallery/dat.../DigiFlow1.jpg
Back:
http://www.overclock.net/gallery/dat.../DigiFlow3.jpg
What you get (Minus the barbs)
http://www.overclock.net/gallery/dat.../DigiFlow4.jpg
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.
http://www.overclock.net/gallery/dat.../DigiFlow2.jpg
The turbine (Sorry for the poor picture, it's about a 3/8" Diameter turbine):
http://www.overclock.net/gallery/dat.../DigiFlow5.jpg
Internals: (Could easily be taken apart and the LCD screen, battery pack, and speaker could be easily relocated)
http://www.overclock.net/gallery/dat.../DigiFlow6.jpg

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.
http://www.overclock.net/gallery/dat.../DigiFlow7.jpg

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..:D

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.
http://img265.imageshack.us/img265/9...ow8000tjc4.jpg

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

Quote:

---

Originally Posted by joecop120

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.

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.

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.:up:

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.:shrug: Too bad you couldnt set it up for a minimum flow rate or something like that though...

I'm assuming the display isn't backlit?

Quote:

---

Originally Posted by headala

I'm assuming the display isn't backlit?

---

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

Quote:

---

Originally Posted by DerekFSE

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...:D

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?

Quote:

---

Originally Posted by 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?

---

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.:up: Only problem is it takes soo long to get lower end flow rates.

:) Thanks for the info

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...:D

It seams pretty bulky and ugly

Quote:

---

Originally Posted by The_Beast

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.

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.

Quote:

---

Originally Posted by Martinm210

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.

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?

Quote:

---

Originally Posted by 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?

---

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

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

Quote:

---

Originally Posted by 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

---

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.

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

Quote:

---

Originally Posted by Marci

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

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

Quote:

---

Originally Posted by headala

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.

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

http://forum.effizienzgurus.de/showp...&postcount=134 seems to be the right circuit board for the SF800 divider, too bad it's in german so I need someone with german or electronics knowledge to confirm this.

the forum page with this translated : http://translate.google.com/translat...hl=fr&ie=UTF-8 then just jump on page 6 for the diagram and the notes.

The diagram on page 6 or the revision on page 7?
Its 40 years since my high school german. Wasn't too good at it even then. lol

Quote:

---

Originally Posted by Jedda

The diagram on page 6 or the revision on page 7?
Its 40 years since my high school german. Wasn't too good at it even then. lol

---

Missed the revised version, thanks for pointing it out.

http://www.turbokeu.com/myprojects.htm is a page with some details about the SF800 and his own schematics to directly read the flowmeter feed in a LCD. I'm sure he know how to build a divider circuity so contact him at :

turbokeu at turbokeu.com

info at turbokeu.com

turbokeu at skynet.be

I also made a post on the Aquacomputer forum in hope someone got something to hook the SF800 to the Aquaero.

Quote:

---

Originally Posted by Exahertz

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.

---

It would look bad if you did that, where do you hide it when you have an acrylic case :rolleyes:

Quote:

---

Originally Posted by Xilikon

http://www.turbokeu.com/myprojects.htm is a page with some details about the SF800 and his own schematics to directly read the flowmeter feed in a LCD. I'm sure he know how to build a divider circuity so contact him at :

turbokeu at turbokeu.com

info at turbokeu.com

turbokeu at skynet.be

I also made a post on the Aquacomputer forum in hope someone got something to hook the SF800 to the Aquaero.

---

I've just read through this and I have to tell you that I’m totally out of my depth trying to understand and interoperate what I will need to do when my Swissflow SF800 comes.:confused:

I think I need to get myself on an electronics course, because all of a sudden I feel totally inadequate ha ha .:eek:

I hope all you gifted guys that are knowledgeable in this area will be helpfull.:up:


This schematic is ment to digitally display the output of my electronic Swissflow SF800 flow sensor. This sensor puts out an open collector square wave signal (like a fan RPM monitoring signal) between 50Hz (=0.5 liter/minute) and 2000Hz (=20 liter/minute). By dividing the measured frequency by 10 and putting the decimal point at the right place you get directly a display in liter/minute. I don't use the first digit output (output D) from the 74C925 to obtain a display of 10 to 9990Hz or xx.x liter/minute. Resistor R1 (2k2) is there for accomodating the open collector output of the SF800 to a 5V TTL signal, available at connector J1.
The schematic I designed is based on a old well known Fairchild Semiconductor counter IC, the MM74C925. It's basically a 4-digit counter with seven segment-decoder & multiplexer and counter & latch registers.
The other sections of the schematic include a precise 0.5Hz timebase for the measuring time (during the positive cycle of the clock = 1 second = 1measure/2 seconds), based on a Philips HEF4521 crystal oscillator & 24-stage divider. This oscillator oscillates at 4.194308MHz and the output signal is taken from output Q23 which represents a dividing factor of 2^23 (=8.388.608) and serves as gate signal for IC3B (CD4093 NAND Schmitt trigger). Integrator IC3A creates the periodical latch enable signal and integrator IC3C produces the periodical reset signal, also from the 0.5Hz timebase signal.
The 7805 regulator delivers a stable 5V power supply from the 12V PSU line for the 74C925, the HEF4521 and the CD4093.
The segment driver output of the 74C925 is buffered by an ULN2003. With the shown values of R6-R12 (150 ohm) the peak output current per segment through the displays is about 40mA to compensate for the duty cycle of 25% and the multiplexing frequency of 1000Hz.
The 7-segment displays are (ofcourse) blue displays, the 0.56" common anode LT-5501 from Liteon. Blue leds have a forward voltage of about 3.8-4.0V, and taken into account the 0.9V saturation voltage of the ULN2003 output and the 0.7V saturation voltage of a BC327, +5V power supply was insufficient, and the displays are powered from the the +12V line. Since the digit driver outputs of the 74C925 only outputs 3 to 4 volts, it's not sufficient to saturate driver transistors T2, T4 and T6 (BC327 PNP's) and I added T1, T3, T5 (BC547 NPN's) to bypass this issue.
My Eheim 1250 pump delivers a flow of 1200 liter/minute at 0" head. Taken into account the pressure drop from radiator, water blocks, tubing and flow sensor it's quite possible that the final water flow will be under 10 liter/minute (600 liter/minute). In that case the first digit will always indicate "0". Therefore I will show both the schematics and PCB layouts for XX.X liter/minute display and for X.XX liter/minute display.

Quote:

---

Originally Posted by The_Beast

It would look bad if you did that, where do you hide it when you have an acrylic case :rolleyes:

---

Well from a cosmetic standpoint i see were your coming from (you have to be real carful about things when it comes to the acrylic case). but i only have a window so i could hide it out of sight.

just a quick thought... do they make flow sensors that you can plug into the mother board as a fan and send flow rates as RPM? cuz you could easily convert the RPM into GPM, LH or any other unit in software. the hardest part i think would be making the correction charts for RPM/GPM

Quote:

---

Originally Posted by ladderman

I've just read through this and I have to tell you that I’m totally out of my depth trying to understand and interoperate what I will need to do when my Swissflow SF800 comes.:confused:

I think I need to get myself on an electronics course, because all of a sudden I feel totally inadequate ha ha .:eek:

I hope all you gifted guys that are knowledgeable in this area will be helpfull.:up:

---

Yeah well.
I think you'd be far better off trying to get this guy to make what you need.
Going off on a tangent setting up the gear to create your own PCBs is counter productive.
What he's described is a PCB display for the flow meter, rather than a signal dividing circuit so the flowmeter could hook to your M-cube or Aqueiro.

Hehe. You guys are over my head on the electronics piece of the SSF 800, I can wire just about anything automotive, or household, but when it comes to real electronics....I was pretty happy with myself finishing the first half of my electronics learning lab exercises...:D

Quote:

---

Originally Posted by Jedda

Yeah well.
I think you'd be far better off trying to get this guy to make what you need.
Going off on a tangent setting up the gear to create your own PCBs is counter productive.
What he's described is a PCB display for the flow meter, rather than a signal dividing circuit so the flowmeter could hook to your M-cube or Aqueiro.

---

I’m aware of what this guy is building. I was just making comments on the degree of expertise needed to do it.

My intension is to use an SF800 in conjunction with the mCubed Sensor hub. What I am saying is I may struggle to understand the schematic that comes with the SF800 if I need to wire it correctly to the mCubed Sensor hub.

This is a nice flowmeter, very compact and includes a bult-in readout. The thread about my Protues flow meter didn't appear to raise as much interest, though...

Quote:

---

Originally Posted by MegadetHCl

This is a nice flowmeter, very compact and includes a bult-in readout. The thread about my Protues flow meter didn't appear to raise as much interest, though...

---

That's a much more robust setup there, great work!

Quote:

---

Originally Posted by MegadetHCl

This is a nice flowmeter, very compact and includes a bult-in readout. The thread about my Protues flow meter didn't appear to raise as much interest, though...

---

Just read your thread, love the set-up. But on cost alone would be totally unrealistic. If a Flowmeter like that cost $295 it would cost double that here in the UK, and you are using two. The Flowmeters must be worth more than the rest of your rig :eek:

They look great though:up:

I also posted this in the OP, but I just wanted to follow up with some actual pressure drop testing of this flow meter that I ran tonight.

I also did a complete retest on the flow reading and got nearly the same results. My original test which used a graduated bucket and stopwatch identified an average 11% High reading delta to actual flow rate.

Doing that same comparison using my new King Instruments flow meter resulted in a 13% High reading delta...consistent results:yepp:

The pressure drop test, actually less restriction than an MCW-30 which is the least restrictive block tested to date. Nice and low restriction for a flow meter:up:
http://img265.imageshack.us/img265/9...ow8000tjc4.jpg

That's really a good result then, Martin.
You're getting a very nice test bench together. Should help lift your spread-sheet tool to even higher hights.

Man, you are outta control...in a good way! I'm glad to see you're enjoying your new gear.

Have you tested a swissflow for accuracy or pressure drop?

Quote:

---

Originally Posted by headala

Man, you are outta control...in a good way! I'm glad to see you're enjoying your new gear.

Have you tested a swissflow for accuracy or pressure drop?

---

+1 for a Swissflow SF800 test :)

Thanks guys!

The new test gear is alot of fun, it's amazing how much faster it is to test than the old bucket/stopwatch method. Then there was the actual manometers...all 16' of them...:eek:

Now I've got instant results, and so far even more consistent/accurate too. I've had a few bad graphs in the past outdoor manometer setup, and so far with this..I'm consitently getting a very nice set of data that fits a nice parabolic curve very well (very little deviation).

Anyhow, no I have not tested a swissflow, although I suspect it's a much higher precision tool. This little DigiFlow is pretty cheap and low on quality, so I'm not too suprised by the reading errors.

I'd be happy to test any extras anyone might have..;)

Quote:

---

Originally Posted by Martinm210

Thanks guys!

The new test gear is alot of fun, it's amazing how much faster it is to test than the old bucket/stopwatch method. Then there was the actual manometers...all 16' of them...:eek:

Now I've got instant results, and so far even more consistent/accurate too. I've had a few bad graphs in the past outdoor manometer setup, and so far with this..I'm consitently getting a very nice set of data that fits a nice parabolic curve very well (very little deviation).

Anyhow, no I have not tested a swissflow, although I suspect it's a much higher precision tool. This little DigiFlow is pretty cheap and low on quality, so I'm not too suprised by the reading errors.

I'd be happy to test any extras anyone might have..;)

---

I will see if I can order a Swissflow myself then borrow it to you for some testing. I don't need it asap ;)

We need some guys (me included) with a SF800 to build a good setup for flow reading.

Quote:

---

Originally Posted by Xilikon

I will see if I can order a Swissflow myself then borrow it to you for some testing. I don't need it asap ;)

We need some guys (me included) with a SF800 to build a good setup for flow reading.

---

Sounds great! Every time I test something new, I'll plan on adding it into the estimator to so you can tinker with it..:up:

Quote:

---

Originally Posted by Martinm210

Sounds great! Every time I test something new, I'll plan on adding it into the estimator to so you can tinker with it..:up:

---

I have sent a request for infos to Swissflow and I'll let everyone know how we can order a few. I'm feeling it would be impossible to do a normal order for a single unit but only in bulk so if it is the case, I'll see if I could organize a group buy (with a deposit) and lend one to Martinm210 for testing and adding in the flowrate estimator.

This is unless someone know where I can buy a single SF800 new or used. PM me if you have the infos or own one yourself for sale.

I bought mine as a single direct from Swissflow in Europe. I think the guy's name was Dirk.

Thanks!

I know someone offered me to borrow a swissflow before, but it was before I had my internal test rig built and I didn't want to borrow it for a long time. Now I'm just interested in running a quick curve for it and checking flow rate accuracy.

Swissflo will sell you one direct for 25 euros and asked 8 euros to send it to australia.
They've set up a paypal account to handle it too.
I had an exchange of emails with them last year and they set it up for direct sales to LC enthusiasts then. Before they were whole sale only.
Good guys, great price!

Quote:

---

Originally Posted by Jedda

Swissflo will sell you one direct for 25 euros and asked 8 euros to send it to australia.
They've set up a paypal account to handle it too.
I had an exchange of emails with them last year and they set it up for direct sales to LC enthusiasts then. Before they were whole sale only.
Good guys, great price!

---

Thanks for the info. I saw a old post saying they will only take orders of 50+ at a time but it was in 2005. Looks like they recognized the enthusiast market and setup to help them...

I also received a email from them with a pps presentation and asked me for what purpose I'm interested to the SF800. I will explain what I wanted and see what I can do :)

info@swissflow.com
I dealt with Dirk B.van Wezel

Dirk B.van Wezel dealt with my order too. It cost 25Euros + 7Euros for shipping. It only took four days to arrive.

I’ve just spent all day doing flow comparisons against my blocks and rads to see how restrictive it is.

Will be making a thread on the results shortly. So your input will be very much appreciated.

They will be no were near as sophisticated as your tests so please don’t mock me too much please.

Quote:

---

Originally Posted by Jedda

info@swissflow.com
I dealt with Dirk B.van Wezel

---

Yes, that's the same one who replied to my email. I'm just waiting for a quote and if it's ok, I'll order one.