Ok, been tinkering with the scales again.
Sawed the existing circuit board in half (as you do) and re-fitted the buttons. Didn’t take long:
Then I finished the display driver board…
The 6 pins to the left of the IC are where the display sits:
Following that, I reassembled the main part of the scales (the load cell bits) and dug out the INA125P for a quick play.
All a bit of a mess, as usual, but I was playing around with resistors to set the “gain” on the instrumentation amplifier at this point. I settled on 33 Ohm which gives a gain of, er, 4+ 60000/33 = 1822. With the voltages I’m seeing then that means the scales will go up to about 12.1 kg. Not bad. Resolution is 3g so, if I do it right, accuracy should be around +/- 1.5g.
Here’s a 2985g weight on there:
I’m not entirely happy with the content of these posts. I will provide circuit diagrams for this, and links to whatever software that I end up writing. I would sooner get it right first though. Having looked at lots of INA125 circuits on the interwebs it seems that most of them are just flat out wrong. Most of them don’t follow the datasheet. I don’t get it, unless people rush to show the circuit and then do testing and don’t update it?
The main problem seems to be that they use the 5V “precision voltage reference” when driving the thing with, er, 5V. Datasheet states that you have to have +1.25V headroom to use any reference voltage. I’ve tested 7.5V, 9V and 12V and it’s rock solid. Then I tried 6V and it’s all over the place. So, stick with the datasheet and it’s fine. No surprise there then. I can’t believe anybody would take the output of a simple LDO over a “precision voltage reference” that’s been designed to be, er, a “precision voltage reference”. But they do. Weird.
I can’t really moan about this unless I publish my working circuits, so I will 😉 If I’ve screwed anything up then I would hope to get shot down in flames and learn something. That’s really what this is about for me.
Right, in “a stroke of good luck” I seem to have bought a couple of MCP3202 12-bit ADC converters a couple of months ago. Dunno why 🙂 That should give me the accuracy I need. You can see the thing on the far right of the breadboard, it’s an SPI device so no big deal to drive. I shall play with that tomorrow.
Getting long on words and short on pictures now. I’ve dispensed with the battery option for these scales – need 6.5V + for a reliable reference voltage. The reference voltage drives the load cell and will be used to drive the ADC as well – a stable voltage helps, apparently. That makes a lot of sense, even to me.
The Arduino Mini Pro will also take the 6.5V-12V source and drive the display via its on-board regulator. This should all work. And I will document it properly, rather than rambling 🙂