The Raspberry Pi sense hat uses an accelerometer. Grace’s programmers we using that to measure vibration rather successfully.
Turns out it’s actually an IMU. More details here
Environmental Monitoring
3 Likes
Or detect the magnetic field in the arch . 600A of traction current might be in the range of a hall device plus it will be rectified phase so a switched capsiter notch shood have no problems picking it out from the noise
Or a loosely coupled coils to detect the track circuit current
2 Likes
I pizo element with a block of heavy glued to it will do the job
That’s a good idea too - I’ve got some v sensitive accelerometers from my drone flying somewhere… i2c interface and low current…
will dig them out…
For piezo, how about a few of these ?
Courty
1 Like
One lump of heavy and a raw bolt and we are in business
All sounds great see you Thursday hopefully 
Random interjection: Berth step data might be quite accurate from Network Rail’s data feeds (the Train Describer specifically)- as it actually identifies a particular piece of track. If there is a signal very close by, those are at the boundary of berths, so we might get a signal just before a train goes overhead.
I have this which (amongst other things) takes in the TD feeds: https://github.com/naxxfish/nrod-funnel
4 Likes
Wow great! Pretty relevant You around on Thursday?
Mmmaybe not this coming Thursday - possibly next though!
If I get a moment, I’ll see if I can knock something together that’ll monitor berths around Herne Hill and we can see if any steps correlate with Makerspace shaking
2 Likes
That would be easy all we need to know is witch blocks are above us
Therein lies the challenge 
It’s reasonably easy to look through the CORPUS dataset and see what berths are going into and our of Herne Hill - the question will be which are the ones which are above us. It’s made a tad tricky by the fact we appear to have a crossover directly above us (probably why the noise is quite as bad as it is).
Ridear the trane read the numbers
Ha, I wish! Berth IDs don’t match up with signal numbers 
Some berths have signal numbers in them, but not all.
Block I’d is what you want .
Will have a look at it on Thursday and see if I can remember anything from my years on the railway
Could you look on the staff screens in the station, sometimes they are visible?
Here’s the docs on the data I’ve got:
http://nrodwiki.rockshore.net/index.php/SmartBerthDetail
http://nrodwiki.rockshore.net/index.php/Identifying_Locations
Here’s some likely candiate berth steps:
FROM:
STNAME:
FROM 0129
TO 0137
STANOX 87249
NLCDESC: HERNE HILL
FROMLINE:
TOLINE: M
ROUTE: 1
TD: VC
Type: Between
Event: Depart Down
Platform 4
Comment: Migrated on 8/6/2005
STNAME:
FROM 0129
TO 0757
STANOX 87249
NLCDESC: HERNE HILL
FROMLINE:
TOLINE: C
ROUTE: 3
TD: VC
Type: Between
Event: Depart Down
Platform 4
Comment: 02/03/2008
TO:
STNAME:
FROM 0125
TO 0129
STANOX 87249
NLCDESC: HERNE HILL
FROMLINE:
TOLINE:
ROUTE:
Type: Between
Event: Arrive Down
Platform 4
Comment: Migrated on 8/6/2005
STNAME:
FROM 0123
TO 0129
STANOX 87249
NLCDESC: HERNE HILL
FROMLINE:
TOLINE:
ROUTE:
Type: Between
Event: Arrive Down
Platform 4
Comment: 02/03/2008
I think the ones we’re mainly interested in will be the Arrive Up and Depart Down ones from HNH.
1 Like
I get that line and there are a number of fast trains that don’t stop at Herne Hill. Those are most likely the noisy ones
Courty
Not going to make it tonight but will buy those bits for next week
Providing they all go through Herne hill, we can detect them.in fact even if they don’t tell probably go through a signal at some point and that’s detectable. During the governance meeting i was watching thee Td feeds for events around HNH, and they did appear to correlate to trains passing overhead.
tsssk!