Magnetometer Project

[Wah]

A swimming pool should have lots of heat capacity, you may put it under water~

[swisswalter]

Hi Mike

I'm so sad to read about the death rattle of the fridge

Some of the best insulation material is aerogel, it may be worthwhile to try it

[grimble_cornet]

Interesting idea Wah

.....and water would certainly 'damp-down' the thermal oscillations BUT I think I have enough problems at the moment without introducing 'water + electronics = BANG' into the equation

Thanks for the aerogel suggestion Walter.
It does look like a possible solution. The 5mm blanket is not too expensive and is available in small quantities via the internet.

At the moment the cool box is wrapped, rather inelegantly, in a duvet and the temperature does seem quite stable with no significant change as the ambient temperature rises and falls by 6-8 degrees.

Heat mat plus duvet.jpg (278.29 KiB) Viewed 2072 times

However, the trace does still seem to be drifting upwards - the basic frequency from the Bat detector has climbed steadily by about 50Hz since the temperature stabilised last night. This upward drift changes the overall shape of the trace when compared to the Intermagnet data although the individual 'micro-magnetic-events' are still captured.
Stuart has stressed several times that it does take a significant amount of time for things to settle down so........ think I will just leave it for a couple of days and see what happens before I explore a slightly more 'elegant' thermal insulation solution

[swisswalter]

Yes Mike

you are on a state of the project where you have to take your time. No rapid changes anymore (maybe a beer more )

[grimble_cornet]

Beer always helps - and now I can get it from the fridge without fear of disturbing the Bats

[grimble_cornet]

Well this is interesting.................

After transferring the Bat detector from 8 degrees C in the fridge to 27 degrees C in the heated cool box I expected it to take some time to settle down but this is getting silly

During the first few hours the output of the Bat detector increased from about 2700Hz to about 3300Hz before is started to flatten out. So far, so good - this is exactly what has happened before

However, the increase continued over the next 12 hours although at a slower rate. As the temperature was still changing a little too much for comfort, I wrapped the cool box in a duvet and found that the heat mat and pulse thermostat was able to maintain the temperature within less than half a degree. This now pretty much matches Stuart's set up so surely the trace will now settle down

Nope

After 60 hours of very well controlled temperatures - the frequency is still climbing at about 150Hz per day with no sign of it stopping or even slowing down
It is now at about 3760Hz - over 1000Hz higher than when it was in the fridge with nothing having changed other than the temperature of the Bat detector.

The good news is that it is picking up very small changes in the Earth's magnetic field and - if we ignore the relentless uphill climb - matches Intermagnet data very well.

Heat mat plus duvet.jpg (278.29 KiB) Viewed 2043 times

The latest 24 hour record looks like this:

29th Jan 2015.jpg (268.58 KiB) Viewed 2043 times

I'm a bit baffled now and not sure what to do next
I have cracked the temperature stability issue (albeit in a rather inelegant way via the duvet wrap) but can't work out why the frequency is still rising?
One part of me wants to just leave it and see just how high it will go...........
Another part wants to get in there and re-tune the Bat detector to bring the base frequency back down to 2800Hz or so and see if it will stabilise there.

Decisions, decisions....................

Good job that I enjoy a challenge

[Carbon60]

Hi Mike,

I would just leave it to settle down. It will eventually peak. Don't forget that the thermometer will be reading slightly differently to the true temperature of the detector (i.e. it is not actually sitting inside the plastic housing and measuring the temperature of the circuit board). I agree that after 60hrs it should be stable, but let's see how far up the scale it goes. There's still a lot of headroom on the spectrum

Be aware that if it drifts too much then you will have to re-set the measurement frequency range selected for data acquisition.

Cheers

Stu

[Montana]

8 degrees to 27 degrees is quite a jump, especially when just 0.5 degree makes so much difference, it might take a while

Great data though

Alexandra

[grimble_cornet]

Thanks folks, I think you are right and I will just see what happens

I agree Alexandra that the data is pretty impressive - better than I would have dared to expect from a 'hobby electronics project'

It is the need to re-set the frequency selection which makes me want to 'fiddle' but I'll let it settle first.

One question Stuart - the tone coming from the Bat detector seems very 'rough' with lots of quite harsh 'noise' rather than the fairly pure tone I would expect.
I could be wrong, but I think a lot of the 'noise' started when I moved over to a fairly simple 5v power supply (I think it powered a usb hub originally)?
Spectrum Lab seems to be coping with it and is able to identify the 'peak frequency' so, doesn't seem to be preventing me from picking up detailed data which matches Intermagnet data very well but I wondered if you got a 'cleaner' response by using the regulated supply shared with the FGM-3?

[Carbon60]

Hi Mike,

It sounds like the issue is going to be the power supply. Probably best to use something stable and regulated to avoid noise.

Cheers.

Stu.

[grimble_cornet]

Thought it was flattening out last night but no........ still rising slowly!

It did pick up a minor 'event' very well yesterday evening though so I'm not going to change anything for a while

The duvet seems to be working as the heat-mat plus pulse thermostat is now keeping the Bat detector stable to within about 0.3 degrees C

30th January 2015.jpg (280.76 KiB) Viewed 2003 times

[Montana]

Nice peak

Alexandra

[marktownley]

Just thinking here, the circuit will naturally generate 'some' (small amount) heat by the nature it is turned on, if it is wrapped in the duvet all that heat will be trapped and the temperature will rise, gradually. maybe?

[grimble_cornet]

Mmmmm. I had considered that but the box, inside the duvet, is thermostatically controlled with the heater running at 0-100% as required. Once the heating mat is switched off by the thermostat, the temperature in the box drops quite quickly before the heating is switched on again. The temperature probe is inside the foam shell surrounding the Bat detector so is measuring the temperature within about 2cm of the circuit board and remains stready to within 0.3 degrees.
Weird

5 days and counting...............

5 days.jpg (225.48 KiB) Viewed 1996 times

[Carbon60]

Hi Mike,

How's the voltage? If the temperature is stable, then the other variable is voltage (both on the sensor and the magenta 5).

Mark's correct about internal heating from the bat detector, but as you say, this should have stabilised by now, which brings us back to voltage. There's no substitute for a good electrical supply!

Stu.

[grimble_cornet]

That's an interesting point Stu

The sensor is powered from a 12v supply dropped to 5v and stabilised via the circuit you show in your project guide

The Bat 5 - as I said before - uses an old 5v hub power supply which I could replace by tapping another feed from the 12v supply - I have the components but didn't bother when the hub supply seemed to be working

I agree that the Bat 5 supply is the weakest link and it probably makes sense to upgrade BUT...... when the temperature was stable in the fridge, it was giving an almost perfect match to Intermagnet

Maybe things are different at 27 degrees?

Anyway, it seems to be almost flat now so I'll leave it for another 24 hours before considering a power supply upgrade, insulation tidy and Bat 5 re-tune.

[swisswalter]

Hi Mike

+/- 0.2 °C is very good. can you really do better ?

[marktownley]

Are you using a 'wall wart' style power supply?

[Carbon60]

Hi Mike,

For reference, my data over the same period.

Coronal hole 20150131.jpg (98.08 KiB) Viewed 1969 times

Stu.

[swisswalter]

Hi Mike and Stuart


wonderful, you both are almost twins, to be identical twins however, there has to go more work into the Project

[grimble_cornet]

Walter: no I don't think I can improve the temperature control. I would like to 'tidy up' the insulation by replacing the duvet with something more appropriate - maybe the aerogel you suggested?
Mark: yep, its a plug/transformer combo type. I know it's noisy and I can just tap into the supply to the sensor but I'm not convinced that it is the problem as I can get almost perfect data for hours at a time?
Stuart: thanks for the comparison:

31st January 2015 comparisons.jpg (259.48 KiB) Viewed 1955 times

Looking at the last 12-24 hours, it seems that the difference between my data and that from Intermagnet is little more than the difference between individual magnetometers contributing to Intermagnet


I might strip it down and take the opportunity to re-tune the Bat 5 tomorrow

[swisswalter]

Hi Mike

great results, fantastic results. BTW, don't you have to correct for the location on earth ?

[grimble_cornet]

Perfectly correct Walter - I have been looking at a range of magnetometers in 'Mid-latitude Europe'.

Anyway - very happy with today's trace which finished with a nice 80 nT peak which matches very well the data from the Intermagnet site

February 1st 2015.jpg (310.59 KiB) Viewed 1945 times

Don't think I can ask for much more than that

[swisswalter]

Hi Mike

what a fine catch, great match, congratulations

[Carbon60]

Hi Mike,

I think we can conclude that 'SolarChat' magnetometers 1 and 2 are up and running well!

We just need a few more dotted around the globe to have a network


For anyone else on this forum, if you've read this far then you must have an interest in this sort of thing. As you can see, Mike has set up a fully functional magnetometer using little more than a FGM-3 sensor, an ultrasonic emitter, a bat detector and a lap-top. Key to this project is having a location that is free from magnetic disturbances, a good stable power supply (with some basic home electronics required) and excellent thermal management. There can be some challenges in all these things, but as Mike has shown, persistence wins and with some basic equipment suitable temperature stability can be achieved. I think the total cost is in the region of £250-£300, including sensor, detector and heater/thermostat.

I'd encourage anyone who can meet the 'free from magnetic disturbances' criteria to have a go. More details on how to construct this device can be found in the tutorial section.

Good luck.

Stu.