Insane Hydraulics
Bold and audacious blog about fluid power
The assembled PCBs with analog-to-digital converters for the super pressure gauge project have arrived, and what a treat they are:
I don't have any affiliations with the fabrication house that manufactured the PCBs and did the assembly - the JLCPCB - but I can state that my experience with these guys has been problem-free so far. And, believe it or not, the total amount I paid for the five assembled PCBs, including shipping and customs, is less than the price of 5 bare AD7195s if I were to acquire them locally, for example, from Mouser Europe - who currently retails these chips at €17.98 a pop plus shipping. Insane, isn't it?
Anyhow - now's time to figure out if I can get the ADCs working. As I explained before - my plan is to use the 600 bar transducer that I salvaged from a dead Parker digital pressure gauge, and according to the datasheet, it is supposed to have bridge resistance of 6.5 ±1.3 kΩ and an output of 1.7...2.3 mV/V, so the first thing I need to do is create a similarly-spec'd Wheatstone bridge emulator. Let me show you how I did it by employing another DIY tool that I use quite often in my day job:
Two pots, two trimmers, and two screw connectors mounted inside a dollar project box - 5 bucks worth of parts with virtually unlimited potential! I have been using this doodad for driving ratiometric inputs of proportional valves (which, in my case, defaults to Danfoss PVG32s) for years. The pots are 1K and 2.2K - for no specific reason, just something that I had "lying around" when I built it. Sometimes I connect them in series for "redneck 20 mA loop emulation purposes". I am not even sure what current these no-name pots are rated for, but I haven't burned them so far, and you wouldn't believe how many times this little box came in handy in my practice. Feel free to steal this exceptional design.
Here's how I used it in combination with some SMD resistors that I slapped on a prototype board to get the desired 6kΩ Wheatstone bridge with a regulated millivolt output:
The execution is... "prototypy," but I think we all can agree that the design is genius! If I drive this bridge with 5V, I can use the 2.2K pot to get an output signal that varies from 0 mV to about 10 mV, which is exactly what I need. With real-life parts, I actually managed to go from 0 to 11 mV (with 5V excitation):
Finally - the last step - connect one of the boards to the makeshift Wheatstone bridge emulator and my Silicon Labs development kit and see if I can communicate with the ADC. Once again - one of the reasons I chose this converter was the existence of a free driver library in C one could pretty much copy-paste into a project without the need to "reinvent the wheel."
I am very pleased to report that the board functions as intended (at least the one you see in the picture), and I managed to read and write to and from all of the ADC registers. Here you can see the AC excitation of the bridge (the yellow line, ACX running at 150 Hz), with the blue and the magenta lines being the serial clock and data ready/data out pins of the ADC:
As you can see, at the end of every AC cycle, some communication happens, and if we zoom into what at current resolution looks like negative spikes, we'll see that about 90 microseconds after an excitation cycle is fished, the data-ready pin is pulled low (signaling that the conversion data is ready to be read) and then about 40 microseconds later my development kit starts pulsing the serial clock line and the ADC shifts out a reading:
Whew! It works! What a relief!
Now I have to come up with an interface that can display the ADC value on the graphic LCD of the kit (in required units), connect a real pressure transducer to the board, and finally see how many of those 24 bits I can actually use.