Slowly getting there. I need to play with resistor values on the amp circuit as at the moment I can only get a 300mV change in voltage and I’d like something closer to 5V for the A3/A4 pin on the RAMPS AUX-1 connector. The dips in the 'scope are my applying weight to the load cell 5 times with my hand. It’s not a linear output, the sensitivity is much greater at the low load end, just what we need for a spool weight monitor.
What would you do with the spool weight measurement?
To tell you when you are running low
I see… I use a Photon Reflectivity Capture Unit 
Do you have a circuit diagram?
Who knows where this will end up. It’s just one of a number of ideas I’m experimenting with around data.
I’m also making a separate respooling device that will measure the length of filament as I wind it from a 2.3Kg spool to my working 1Kg spool. Given length and weight we can start doing things like warn that for a given print you’re close to running out of filament or worse definitely don’t have enough.
By giving our printers more information we can get the code making sensible adjustments or doing things differently. I really like the dynamic filament diameter monitor Thomas Sanladerer posted a few days back, if it sees the diameter drop the printer can step up the extrusion rate.
Low res thermal imaging could let it know if the cooling fan needed adjusting to optimise speed and adhesion. Or pre-print it could step around the bed to check the temperature was even.
Taking things a step further with weight what if the slicer could talk to the printer too? Imagine a prompt “You don’t have enough filament loaded for 80% infill” with buttons for 60%, 40% and Cancel.
@Chuck_McManis when I’ve got all the component values finalised out I’m sure I could sketch one out.
@Mike_Thornbury you eyeball it? Good if you’re there, what if you’re printing remotely via OctoPrint or have an automated device to clear the print bed ready for the next print and are running a batch?
I was thinking that if you had a schematic it would be straight forward to figure out the gain required and thus the component values needed. To that though one needs to know the opamp part number and the strain gauge part number.
Thanks Chuck. LM358, Proto-pic 3133_0 - Micro load cell CZL635
@Mark_MARKSE_Emery I have one printer - it is in the workshop with me.
What if’s are great, but what if I don’t have Octoprint, let alone a remote camera? What if I don’t have a an automated bed clearance device? What if I print less than a spool a week? What if I make mostly one-off items, one at a time?
I, like hundreds of thousands of other printers, looks at the spool and estimates the remaining amount. It seems to be pretty effective.
Sorry, I wasn’t dissing your device, I think it’s an innovative design, but for me, it has little real-world applicability. I was interested in what your application was, that’s why I asked.
@Mark_MARKSE_Emery
ok that makes sense, according to this (http://www.robotshop.com/media/files/pdf/datasheet-3135.pdf) it is 1 mV/V for a 50kg full scale reading means you will see about 40uV difference for a 2kg spool on vs off. To get a 5V reading from 40uV you need a gain of approximately 100,000. You could do that with half the 358 but I would suggest that making one 100x feeding the other with is 1000x. These guys (http://www.phidgets.com/) seem to hav a thing called a ‘phidget’ which can signal condition this for you.