Insane Hydraulics
Bold and audacious blog about fluid power
I finally managed to test the Webtec FI750-120ABOT variable-orifice flowmeter that I tore down the other day, and here are the results and, as usual, my thoughts. I am not sure if I already mentioned it or not - I fished this baby out from scrap some years ago when a client of ours got tired of the constantly sticking piston and replaced it with a thermal mass flow sensor (it was a "very peculiar" application), and then it lay in my "too good to throw away" box for many years till I finally decided to see if I can find a use for it (or not).
So, here's the setup - the FI750 in series with a Parker turbine, hooked up to our test bench:
There's also a volumetric counter at the end of the line - these oval gear meters are extremely precise and are perfect for creating a credible flow rate reference (at least when you are willing to spend time tweaking the flow rate till you hit a desired volume in a desired amount of time). For example, this is what 5 l/min looks like:
Now, before we begin looking at the test results, I need to comment on something real quick - the turbine flowmeter used in this test has the amplifier board that runs my DIY firmware, and you'll see in a minute that its reading is slightly (albeit consistently) offset - that is so because when I programmed the flow-to-voltage constants - I was using another turbine flowmeter for my flow reference, which means that I basically copied whichever offsets it may have had, and also - I am starting the test with the oil at 30Cº, and turbine flow meters can present readings that are lower than the actual flow rate by 10-15% due to the increased viscosity of cold oil. I could offset this error with a custom reading profile if I connected the flowmeter to my Bluetooth monitor, but for this test, I decided to use Parker's very own SCM-150 display, which essentially converts the analog 0...3V volts into a 0...600 l/min with a 12-bit resolution.
Also, let us decipher the FI750-120ABOT, shall we?
I tore this flowmeter down to the last bit, which means that whatever calibration it may have had is screwed up now. There's an interesting phrase in the user manual: "...Recommended period between calibrations is 12 months. Maximum period between calibrations is 36 months. Unit accuracy may be affected by operating cycle, fluid condition or extended periods between re-calibrations..." - Who doesn't like a subscription-based business model, right?
As far as I can tell - the only thing that you can really calibrate is the position of the needle (unless the spring itself looses its "springiness" due to an extremely abusive action). When I reassembled the flowmeter, I positioned the needle so that it begins to lift from the stop when the piston has moved by about 0.5 mm - it was purely a hunch, but it didn't turn out too bad. And, in any case, most of the time what I want from a flowmeter is not absolute precision, but rather relative - meaning I want to be sure that my instrument can reliably read by how much a flow changes from a given baseline. I can tolerate consistent offsets all day long.
Here's some info from the manufacturer - the brochure and the user manual. And here are the test results - the second and the third columns show the readings of the turbine flow meter and the FI750, and the last two columns show the relative flow increments for each of the flowmeters:
| Flow (l/min) | Parker (l/min) | Webtec (l/min) | Parker incr. | Webtec incr. |
| 5 | *** | 7 | *** | *** |
| 10 | 8.4 | 13 | 8.4 | 13 |
| 20 | 19.0 | 28 | 10.6 | 15 |
| 30 | 28.2 | 37 | 9.2 | 9 |
| 40 | 37.8 | 46 | 9.6 | 9 |
| 50 | 47.4 | 53 | 9.6 | 7 |
| 60 | 58.4 | 64 | 11 | 11 |
| 70 | 67 | 72 | 8.6 | 8 |
| 80 | 77 | 79 | 10 | 7 |
| 90 | 86.4 | 87 | 9.4 | 8 |
I actually took pics at key points as well, and if you want to go over the readings with a fine-tooth comb, be my guest!:
Now, there's a flaw in the test - I forgot to check the pressure drop of the instrument, and if you look at the 90 l/min picture, you'll see the pressure gauge marking almost 9 bar, but that's not the pressure drop of the Webtec, that's the pressure-inducing valve downstream the turbine flowmeter. The FI750 is supposed to introduce a 2-bar drop with a 25cSt mineral oil at 120 l/min, and I actually did another quick test to verify this:
As you can see, the pressure drop is what they say it should be. Note that I am testing the drop with VG46 mineral oil at about 55Cº (which you can conveniently see on the built-in thermometer).
I started the test at 30C (the VG 46 mineral oil at about 75 cSt) - and you can see the flowmeter "can tell the difference" because the readings, especially at the beginning of the range, are definitely "exaggerated". However, when the temperatures rose (you can see the thermometer readings very clearly in the pictures) - the precision improved. I would say that when the oil hit about 50C, the readings fell well within the declared accuracy of 4%. I especially liked the lower range then - when the needle was pointing at 10 l/min, the actual flow rate was 10.6 l/min, the next bar below it - 8 l/min, the next - 6.6 l/min, and the 4 l/min mark - 5 l/min. I am impressed!
There's another very interesting phrase in the user manual: "...When panel mounting ensure that rear and bottom faces of the unit are at least 12 mm (1/2”) from any ferrous material such as an iron panel or base. The piston contains a magnet that can be affected by close proximity of ferrous material..." I already suspected that this would be the case when I disassembled the FI750 and saw how the magnetic needle coupling worked, but if you wonder just by how much a "presence of something ferrous" could affect such a meter - check out this video:
In other words - this is a decent tool but you must know how to use it, and it will bite you in the read end if you don't.
I would not call it "beginner-friendly." Forget that it's "one-way only" - and you can literally blow something up with your diagnostics. Use it with cold oil - and get yourself incorrect readings, which may lead to incorrect conclusions. Put in on a metallic surface - and, once again - get a wrong reading with all the following. You see what I mean, right?
But if you remember about the one-way, aim for 25 cSt, and keep it away from metal - you'll get yourself some decent readings all right! Can this be applied to all troubleshooting scenarios? No, it can't. This is why I would not recommend such a flowmeter as your primary flow-metering tool. Turbine is still the king! But it is definitely a decent secondary or a backup. If you ever get one of these - I would strongly advise running a similar test on your test bench so that you can see for yourself how it behaves before using it in the field.