OSH – a discussion

For electronics hardware – most people are used to shopping and then buying – simple transaction – however Open-Source Hardware at the board level can be different.

@ensign published https://www.envirodiy.org/tools-for-collaborative-solutions-to-water-challenges/

So, I’m interested in a discussion of how can OSH is used – Mayfly and other.

Electronics Hardware is thought of as having a life cycle – with a diverse community of people contributing – the concept, use cases  – – > brainstorming/requirements/objectives,- – >   circuit design – – > prototyping – – > characterization & hardware software intergration testing  – – > financing and manufacturing  – – > “use case stories” – – > selling/distribution – – > installation – – > operation (hopefully for many years – – > and of course recycling.

All of which can require very specific skills at different stages – even the though the engineering is often thought of as a key enabler – and can take time and be risky. So, for advanced hardware the model is often thought of as paying ahead,

The Mayfly 1.1 is great for a bundle of hardware features – and since its open source, these can be inspected, appreciated and also characterized/commented.

So while the big picture may be https://baynature.org/article/capturing-the-flood-in-californias-ancient-underground-waterways/  – I’m wondering how people are down scaling to individual projects –  is it monitoring TDS or simple water levels, is the purpose to inform a watershed occupants, ..

Of course, once there is sensoring hardware with the right set of features, then people with software skills get to jump in. The Arduino environment can make it easier to get started, however once there is momentum its worth sharing what it took (especially the detours, the warts, the problems)

I’m wondering what user really find works with the Mayfly (1.1) and what are synergistic enabler?

So, I’ll start, I’ve really appreaciated the ModularSensor classes communicating with diverse instruments and also the internet facing side sending the results to MMW and being able to collect solar power.

A simple EC-monitor, using the ADC and only internal storage is here, and I’ve got software that makes it easy to download – https://github.com/neilh10/ModularSensors/wiki/Stream-Disconnect-EC-Monitor

However, for other instruments I needed the Modbus/RS485 and developed a board to do that

https://github.com/neilh10/SensorModbusMaster/issues/1

Part of the problem was that full SDI-12 electrical standard isn’t implemented on the Mayfly – so if I have to do a separate board, I might as well do a standard RS485 which interfaces to a wider range of instruments.

Hardware powering can be a very specific challenge, sometimes an artform – and of course if you get it wrong – and it smokes and is bricked, then its rather frustrating. So, a key is always to attempt to make it forgiving of simple mistakes – like on installation getting the instrument wires crossed.

I helped characterize the Mayfly 1.1 +11.3V, which can provide power at 1.1W. (Watts=Volts x Amps, 11.3V * 0.1A) and of course it doesn’t smoke.  The Mayfly 1.1 has many new components, and right now it’s a pretty tough world to get them all together and manufacture the boards in volume.

Rather frustrating, the first version of the wing board I created, had a boost part that has become obsolete, so the new wing board will use a new boost part that has 12V/6W of power – pretty much the max that can be derived from a 4.2V LiIon battery. With that extra power, the hard part is the e-fusing to stop it from smoking.  Effectively I’m testing and proving circuits on a wing board. I should say I’m an individual contributor, no institutional backing, and have no connection with the Stroud Water Research Center, but I am using the OSH philosophy of sharing the inner workings. I haven’t quite had the bandwidth to document it as this point, but will be doing so soon from this link

https://github.com/EnviroDIY/Mayfly-Modbus-Wing/issues/3

Another area that I’m interested in is the ability to do crowd source testing. (Ie that I can test a ModularSensors release easly).  I’m a great admirer of the ModularSensors Classy package (to use a pun 🙂 – one traditional area that is missing is crowd source testing.  However, a crowd source test framework would be something on a release that can be “git clone” locally, one button build, and then download to a Mayfly with simple comms and sensor interface, and then the synergy of crowd source testing of 90% of core functions.