goodbyegti

Sadly the article appears to overlook that Munich is switching back to Windows:

https://www.theregister.com/2017/11/13/munich_committee_says...

Indeed, and when it kills someone, nothing to do with Google.

Absolutely. I will stick with my 80USD shutter precision hub dynamo. 5W of output is plenty for charging in the day and lights at night, and I don't even notice the drag.

https://github.com/dougszumski/KalmanFilter

^ A very simple Python implementation of a Kalman Filter with Matplotlib for visualisation.

Further to this, here's a very simple application of a Kalman filter which plots a graph (in Python):

https://github.com/dougszumski/KalmanFilter

I had a similar problem with a DLP chipset controller (BGA) in an expensive home projector. Instead of attempting to reflow the solder joints, i made a clamp which held the chip in place. Four years later, I've not had a problem since.

For C fans I ported the Python example from this excellent book here:

https://github.com/dougszumski/NNet

A really nice intro is this classic paper on numerical integration of the equation for barotropic vorticity:

http://mathsci.ucd.ie/~plynch/eniac/CFvN-1950.pdf

It's something you can solve in a few lines of python and run with minimal compute resources. A nice thing to try is to feed in some data from somewhere like below and watch what happens as you integrate it forwards in time.

http://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalys...

I work in this sector, and what you say is correct, however it is important to note that in TMS the fields are pulsed over millisecond time intervals. In doing so they can induce an electric field which can cause neurons to fire. The time varience is crucial. A 3T static field from the magnet of an MR scanner is far less dangerous than a 1T field pulsed in the vicinity of your brain. That is providing that there aren't any ferromagnetic objects close to the MR scanner!

If you want to run the AVR examples and don't have a physical target checkout simavr:

https://github.com/buserror/simavr

Another explanation which may be of interest:

http://www.michaelnielsen.org/ddi/how-the-bitcoin-protocol-a...

If it's been done nicely 100% would seem fair to me for a low volume project like this. That's only my opinion however.

For an EE it's pretty easy to put these parts together and make them 'work', but it's fair to say that making the most out of a 24bit ADC is no easy task. That is unless you want 8+bits of noise on the line. You need to take great care decoupling the digital side from the analogue side, and also with the routing of the ADC inputs - guarding / shielding etc. If that's all been done properly it would certainly be worth paying a premium for the time that's gone into the design. I guess the question is has it?

It's mind blowingly expensive. A very quick breakdown of the main parts:

ADS1299 ~38USD for 1k+

ATMEGA328P ~ 2USD for 2k+

RFDuino ~ 20USD for 10 +

Accelerometer ~ 2USD for 1k+

Board ~ 5USD assembled for 1k+

I'm also not sure I'd want to have an RF module near a 24 bit ADC.

Total is probably around 75USD for 2k+. Simple layout as well.

It's absurdly expensive, even more so the modules. Looking at the BT module, the BLE112 can be purchased for $15 in quantities of 1. So you pay $35 for a small PCB, some pin headers and probably a wrapper around a UART.

The electric field for stimulation threshold is typically quoted as 100V/m. There's no chance of achieving that within the brain using the 9V battery and skin contact approach!