Insane Hydraulics
Bold and audacious blog about fluid power
Dear friends, I would like to welcome you to the project "Super Pressure Gauge," through the course of which I will be attempting to make a digital pressure gauge that truly satisfies my exceptionally demanding pressure measuring needs (the project was actually supposed to be titled just "Super Gauge" but, as it turns out, this name is already taken...)
It sounds strange, doesn't it? Why could I possibly want another digital pressure gauge when I already have access to high-quality instruments from Parker, Minipress, Stauff, etc... and even self-hacked but time- and battle-proven wireless pressure gauges based on industrial sensors? Let me explain.
I've been using "good" gauges for a very long time, and I can firsthand attest to their reliability and trustworthiness. I am talking about our ubiquitous daily drivers - i.e "normal" digital pressure gauges with 0.5% FS precision one would typically use in the field - the likes of Parker's Service Juniors or Stauff's SPG-DIGIes - designed around a Wika's TTF-1 pressure transducer hooked up to the MCU's built-in 12-bit ADC.
On a side note - now that I mentioned the 12-bit ADC, let me show you something cool real quick - a pressure gauge of a friend of mine glitched out the other day and got stuck in this weird ADC mode after he replaced the batteries:
Notice that units read "adc" and the full scale is 4095, which totally proves that this is a 12-bit system. We took the batteries out and put them back in - and this special mode was gone. I don't know - maybe there's a combination of buttons you can press to activate it? In any case - I am glad I caught it!
I call these gauges "hydraulic Fluke 87Vs" because, just like the Fluke meters, they've been around long enough to be universally trusted, and they are accurate enough for 99.9% of real industrial scenarios. So, the question stands: "Why re-invent the wheel?"
First - since I have a good idea of how these gauges are built - I can see both the advantages and limitations of their design. While I totally appreciate the simplicity and bulletproofness - I can't stop wondering how much better could the same transducer perform with a dedicated ADC, or how much nicer and more informative could a graphic display be instead of the segmented LCD. I just wrote down two points, but whenever I take one of these gauges out of my toolbox - there's like a thousand other ideas rushing through my head all falling under the category of "it could be a tiny bit better if..." Am I even right in my assumptions/pretensions? Is it even physically possible? I guess there's only one way to find out - build it and see for myself. So this is exactly what I am going to do.
And second - I truly enjoy inventing and making stuff. It's educational, it is fun, and it helps me keep my ADHD (mostly) under control.
Now, let me "rant" a bit more about the gauges I use.
Take, for example, the counts. A typical 600-bar Service Junior has 6000 counts (0.1 bar resolution due to the over-sampling). This is not bad, but could it have more? There are digital scales with 10,000 - 100,000 counts (e.g. a 1000g x 0.01g scale) - why can't my 600-bar gauge have that? I realize that noise issues would make this impossible with the scanning rate of 100 Hz, but why can't I have, for example, a gauge with two modes - a slow high-resolution one and a fast normal one? When I am setting a hydrostatic drive at 450 bar, the high-speed low-count mode is perfect, but when I am monitoring the case pressure of a pump - I want to be able to reliably read low pressures with high resolution - something along the lines of 0.05 bar, or maybe even better. Do I really need to? It's arguable... But I want to!
And now that I mention reading low-pressures with a 600-bar digital gauge - here's a question for you - Did you know that your 600-bar digital pressure gauge (I am assuming that you, like everybody else, use the same model/clone that I do) has a low-pressure threshold below which it reads zero purposefully engineered in the firmware?" I didn't realize that until we got a new batch of Stauff 600 bar digital pressure gauges (both individual gauges and kits):
Me being me - I immediately hooked one of these up to the Pressure Maker II just to see what I would get - and I was surprised to discover that while the firmware of my old 600-bar Parker SCJN gauge cut low-pressure readings below 0.6 bar, the newer Stauff gauges "raised the bar" to 1.2 bar! Anything below just reads out "0"! I checked this on several gauges, and all of them (firmware 5.07A) seem to have the 1.2 bar (18 PSI) lowest reading threshold (even though the spike-catching display clearly shows that the sensor does pick up pressures below 1.2 bar):
I suppose this counts as an incentive to make people invest in gauges with lower range, doesn't it? And why on earth are these gauges so "afraid" of displaying a negative offset? I wouldn't mind it in the slightest. In fact - I would prefer knowing that my transducer has drifted below zero over time before I activate the zeroing function.
Let me show you what I want my 600 bar pressure gauge to be able to resolve. This is my low-pressure gauge with a capillary test hose connected to it:
There's some oil in the capillary (the hose is about 1 m long). Watch what happens to the reading when I stretch the hose vertically upwards, and when I stretch it downwards:
How cool is that? The gauge can actually "see" (albeit very slowly - it takes several seconds for the needle to rise) the hydrostatic pressure of the column of oil in the capillary! And this is my best attempt at creating static pressure with my lungs:
None of my gauges with a 600-bar range can do that. But could the same TTF-1 600-bar transducer cell see such small pressure changes if it were hooked to a proper ADC? I will not sleep until I find out!
So, here's my plan for determining the viability of this project:
Here's what I am hoping to achieve at this initial stage:
Let me show you the supplies I've got so far:
I already have several busted Parker gauges that can donor a 600-bar transducer cell, and this old Silicon Labs development kit that has a BGM13P22 Bluetooth module with a 32-bit MCU and a 128x128 graphic Sharp memory LCD. The module's micro-controller should be enough to drive an ADC and host a simple interface, and this memory LCD, in my opinion, is the best reflective monochrome LCD an instrument can have because it is super sharp and can be easily driven with a refresh rate of 20+ HZ - which is just perfect for that analog bar graph. Its only drawback is its tiny size - but it should be more than enough for this stage.
So, wish me luck, and stay tuned for the updates!
P.S.
I'm going to go ahead and dig out my "thinking hat," for there's a lot for me to invent now...