SolarChat!

That is good!

Thanks Stuart - that looks like good news

I know that you are otherwise occupied for the next couple of days but........ I have a few other experiments to carry out so I'll continue with the commentary and you can catch up later.

So........ on to Phase 5:

While Stuart was playing with my data I decided to try to improve my thermal stability (or at least that of the Bat 5)

First stage was to build a 1" (2.5cm) 'box' out of medium density grey foam (packing from a TV). By cutting and sticking, I managed to encapsulate the Bat detector and Ultrasound transmitter within a 'fitted foam shell'.
This package was then wrapped in several layers of bubble wrap which was then placed in a 12x18x18" (30x45x45cm) plastic cool box.
The cool box was then placed in a room with the central heating radiator switched off - should reduce sudden temperature changes when central heating switches on/off?

I left this new arrangement running overnight: Once again, good news that there is good agreement between my data and the Intermagnet data in terms of up/down 'events' BUT...... once again it shoots upwards at 0800!

OK..... either it is a temperature change at the Bat 5 OR it is a temperature change at the buried sensor OR it is something else happening

Here is my latest toy: This sits in a usb port and records the temperature at the usb port and the temperature of the sensor attached by a lead.
I have now placed the sensor inside the bubble wrap next to the Bat 5 and the temperature is being recorded every 30 seconds - at the same interval used to record the magnetic field. Once this has run for 24 hours, I should be able to combine the two sets of data in Excel and, possible, use the temperature data to 'modify' the magnetic data

Watch this space..................

I admire your dedication Sir, this is interesting to follow

I'm waiting for your better results!

Mike,

As a point of reference, here are my temperature data measured at the Magenta 5. As you can see, temperature variation through the 24hr period (Jan 16th) is to within 0.3 Deg C, although over a longer period of several weeks, with changing weather conditions/ambient temperatures, it can vary to within about 1.5 degrees C.

The thermostat seems to control to within +/- 0.6 degrees, but again this depends on the ambient conditions. I always see a temperature differential between the set temp and the measured temp at the bat detector.
Stu.

Hi Mike

another fine part of the story

That's very interesting Stuart.
I have been monitoring my cool box temperature over night using my new toy and there was not a vast amount of change BUT...... judging by the results, it takes very little to throw off the output from the Bat 5.

It was also helpful that a rather nice 'event' happened at about 0300 ut which my magnetometer picked up beautifully.
Here is a graphic showing my data, the matching Intermagnet data and my temperature log:
I am very happy with the shape of my plot during 'the event' - I seem to be capturing the fluctuations very accurately although I need to 'tweak' the scaling slightly.
It all starts to go pear-shaped at about 0800. Even in a room with the radiator switched off (and inside an insulated box etc.) the temperature starts to climb once my central heating kicks in. The temperature only increases by about 1.5 degrees C but this is clearly enough to cause serious problems as my plot climbs away rapidly - as I have seen it do before.

I can see now why Stuart needs the vacuum flask inside a thermostatically controlled heat box

I don't really want to take my Bat 5 apart and can't find a vacuum flask big enough to take it otherwise so....................

As an alternative approach I might try to build in a correction factor for temperature using the data from my temperature logger.
Chances of finding a nice, simple, linear relationship are not good but................. here is the result of applying a 'first guess' correction:
Not perfect, and I'm not sure it will hold up under larger temperature swings but....... a promising first attempt?

I wonder if moving everything out to my observatory will make things easier........ at least the central heating won't be a problem


Another cliffhanger for Walter.........................

Yes, will be interesting see how the correction factor ties in with data longer term.

Mmmmmmm................. having spent most of the day playing with numbers - it doesn't look good

The thermal response of the Bat 5 is both severe and non-linear
I can build in correction factors which work very well over a few degrees but that will be of little/no use when the temperature is changing by 10 degrees or more during the day and won't be able to cope with a range of 0-25 degrees over the year
Even with the Bat 5 inside an insulated package within an insulated cool-box......... the temperature drift has major consequences

The graphic below says it all really:

On the left, we see that my magnetometer can match almost perfectly the data collected by professional stations - as long as the temperature remains stable to within about 1 degree C. The 'jagged' appearance of my data is a result of me taking readings more frequently (30 seconds) - it will become 'smoother' (for better or worse?) once I change over to the 150s interval used by Stuart.

However, once the temperature of the Bat 5 Detector is allowed to change by just a few degrees........ the whole thing falls apart with the accurate data being swamped by the 'thermal noise'.
If I try to compensate for more than a few degrees, it becomes quite complex and the corrections screw up the calculations used to extract field strength from the frequency. I guess that a better mathematician (Stuart?) could sort something out but it looks as if my long-term solution will be the same as Stuart's which is a thermostatically controlled box.

How do the 'professionals' keep their equipment thermally stable?

How about using a daystar temperature control system? Hehehe…

Hi Mike

very interesting results. and congratulations on the good match. Have thought about putting the whole Bat setup in a fridge

DayStar temperature control system.......... nice low cost solution, maybe I could cannibalise the Quark?

Now, Walter........ a fridge is an interesting idea
I had already considered an Electrically Powered Coolbox but the cost is fast approaching that of Stuart's thermostatically controlled heating mat and I doubt if the temperature control would be good enough.
My experiments suggest that I need to be able to keep the temperature stable to within 1 degree C or better to get good results.
I tried putting the Bat 5 in my observatory over night just to see how well that would work (no central heating to worry about):
As you can see, a change of just 3 degrees is enough to totally distort/swamp the true signal.

Only between about 0600 and 1100 when the temperature is very stable, do we see the true magnetic data emerge


If I can re-route the cable........................ it just might be long enough to reach my milk/beer fridge so well worth a try. I will stick my temperature logger in that fridge and see how stable the temperature is.

Now you're talking beer it will all sort itself out

It's funny how we're thinking along the same lines. I considered temperature correction, which is why I bought the digital thermometer, but then realised that this can only be applied at the data analysis stage. At the time of data capture, the Spectrum Lab chart would be severely distorted by the temperature excursions throughout the day, making it impossible to properly view any true geomagnetic effects on the screen at the time the data were being recorded. I also thought of using a fridge, but I was concerned about possible condensation inside the Bat 5 and I'm not sure how tight the temperature control is on a cheap small fridge. It will be interesting to see what measurements you come back with, Mike.


Stu.

Good point about the condensation Stuart.
I have a pier on my patio with my HEQ5 Pro permanently installed. I place a cloth bag over the mount, a large polythene bag over/around that and a chimea cover over the whole lot................. and I still get condensation on the mount when the temperature drops below freezing as it did last night.
I have my temperature probe in my milk/beer fridge which lives in my un-heated laundry room - we will see how good the temperature control is when its close to freezing outside. Unless the temperature change is less than 1 degree C then I think I would be wasting my time - especially as I assume the fridge will struggle to maintain temperature in the Summer.
If I could see a solution to the 'getting the Bat 5 inside a vacuum flask without taking it apart' problem, I would just bite the bullet and get the heat mat / thermostat / cool-box that clearly works for you.

Beer, beer, beer from the fridge

Hi Stuart，

I am very interest to do this project,I live in tall concrete appartment building,about 25m metres height of my house, my windows is oriented north to north east open area,but I can't see the sun or moon all years long on my window view (blocked by another building).may I do this project too?,The sensor needs to be installed under hidden area? Can I make the shelter or waterproof box to install this sensor near the wall?

The temperature probe spent the night in the fridge with the beer:
I find it difficult to believe it is that good

After a bit of re-routing of wires and a lot of cursing.......... I have the Bat 5 set up in my beer fridge and the trace is dropping with the temperature.
We will see if it stabilises and stays that way for 24 hours

Hi Mike

Beers, beers, beers

I don't think you have to go for deep temperatures, just set it to something around your average temperatures, so it can handle getting colder and warmer. Do you need my adress to send the beers to

ham wrote:Hi Stuart，

I am very interest to do this project,I live in tall concrete appartment building,about 25m metres height of my house, my windows is oriented north to north east open area,but I can't see the sun or moon all years long on my window view (blocked by another building).may I do this project too?,The sensor needs to be installed under hidden area? Can I make the shelter or waterproof box to install this sensor near the wall?

Hi Ham


I'm sure you can do that. Not sure about the wires in the near surrounding

Wow....that looks to be very stable indeed, Mike. Nice one. It will be interesting to see the next dataset. If the temperature really is so stable you'll have no problem detecting every detail.

Ham, In contrast, regrettably, I think you'll have significant difficulties with magnetic interference in your environment. These things are really very sensitive and solar effects are usually very weak in comparison with everyday magnetic fields generated by wiring around the home.

Stu.

Well, following Walter's inspired idea (or alcohol fixation) - 21 hours in the beer fridge and things are looking rather good.......
The temperature changed by about 0.5 degrees C overnight but the data plot looks pretty good when compared to the data from Intermagnet
The different Intermagnet stations all show slight variations so I will try to compare mine with a number of different ones at some point.

This data set seems to suggest that the buried magnetic sensor is working well: especially as we had -4C overnight so a 7 degree swing during the observation period.
It also suggests that if I can maintain the temperature of the Bat 5 to within 0.5 degrees C ............. I will have a pretty accurate and (to my simple mind anyway) impressive magnetometer

The beer fridge saves the day!

Perfect, Mike.

The beer fridge looks like a great alternative to the thermostatically controlled 'cool box'. Let's hope it all holds out long-term. Of course, now it means that you can't open the fridge to get your beer out without risking corrupting your data, but sometimes sacrifices have to be made.

You might want to reduce the capture rate to smooth the curves a bit (2 minutes, or so, works well), or do what I do and capture one set at 1 second intervals and another at 2.5 minute intervals (you can run more than one 'Spectrum Lab' at any one time). This gives you the ability to really focus in on any interesting transitions, such as the point of impact when a CME passes by, but present smoother curves without too much data for a more general view.

Anyhow, it proves the point that the system can be set up by others and that it really works when temperature and voltage regulation measures are implemented. These, as you have seen, are really critical.

Cheers

Stu.

Thanks guys - it certainly proves that the project is realistic

The 'beer fridge' is just a 'cheapo' old fridge which lives in my extension and is used to store milk and beer. Your point about opening the door is a good one and the 25m lead to the buried sensor will only just reach it. I am also less than convinced that the fridge will maintain the temperature as accurately in the summer so ...................... its not really a long term solution.

If I was convinced that a small 'beer fridge' would work, I would install one in the observatory (yeah ) but I suspect that it would show more variation than a full size fridge.

I have pretty much decided to follow Stuart's solution using a heat mat plus a 'pulse' thermostat in a cool-box to maintain the temperature. This could live in the observatory or in the house within easy reach of the sensor cable.

Let's see what a 24 hour run produces once I change over to Stuart's 150 second sample routine.

Hi Mike

great results with the fridge. I'm sad to see you leaving that simple method though, but I do understand the need of a beer or two a day

Don't worry Walter, I have other stores of Beer (and wine) in other fridges



Well, here we go - a full 24 hour cycle of The Beer Fridge Magnetometer

I have 'smoothed' my data in the top right graph to simulate Stuart's 150 second sample rate - the top left graph shows my original 30 second data.

The agreement with the Niemegk data is not perfect - there is some variation from about 11am but each station seems to show a slightly different curve during times of low magnetic activity so........ not sure how concerned I should be about that?
The fridge temperature has been very consistent if you believe what the probe is reporting and the outside temperature has only varied by about 3 degrees C over the last 24 hours so I don't think the buried sensor will have been influenced very much?

Hi Mike

but that looks very promissing, almost perfect

I agree looks good. Longer term testing and evaluation required!

Your wish is my command:

I intend to run it from the fridge for a few days just to make sure that other aspects of the gear are behaving. I will then decide on a permanent set-up for the Bat 5 end of the cable.

In the meantime, here we have 42 hours of data from The Beer Fridge:
Here's to Walter for his inspiring idea:

And finally, here is this morning's data in more detail using Stuart's 150 second sampling rate - seems very 'frisky' with lots of 'pulses' when compared to Intermagnet?

Well done Mike we just need a CME now

Alexandra

YES

That would be very nice

Is it worth putting an RF supressor on the fridge? They are 'electrically noisy' things as they turn on an off, and can play havoc with expensive hifi's - maybe they could have an adverse effect on the bat detector unit?

It wouldn't surprise me if the fridge was generating noise Mark

I have had the washing machine and spin dryer, which live in the same room, working hard this afternoon so it will be interesting to look at tonight's trace

Hi Mike,

I don't know how 'remote' your sensor is in relation to other sources of interference, but these things are so sensitive that you could also be detecting movement of cars in the street, or other such things. From 20m away I can detect vehicles turning in our Cul-de-sac. The deviation only needs only to be a few nT to appear significant.

I think your system is now up and running really well for a home made system operating in a garden environment. Professional systems will be located in a remote field quite some distance from any sources of interference.

Wait for a CME event and I'm confident that you'll be more than happy with the magnetometer response.

Stu.

Congratulations! Good to see such amateur professional results and I am now trying to introduce this to some of my friends from mainland China~

Thank you Wah and I am pleased to hear that you are encouraging others to try this project - I have had a lot of fun getting to work.

Thanks Stuart; I am pretty happy with the results I'm getting since moving into the fridge

I had the washing machine and tumble dryer working yesterday and they live next to the fridge
I got a temperature swing in the fridge (only a degree or so) but it had only a small impact on the results:

Here is 54 hours of Bat in the Fridge data compared with Intermagnet:
I'm pretty impressed with the correlation - especially now I realise how much the 'scaling' changes the appearance of the published graphs and makes it very difficult to combine 2 or 3 days worth of data.




I moved over to 150s samples for temperature and magnetic data last night and am very happy with the detail recorded during a minor event early this morning

Anyway: I have decided to bite the bullet and have ordered a heat mat and thermostat - same models as you use - to create a permanent recording station without being concerned by fridge access, washing machine interactions etc.

Congratulations we now have an aurora alert base in the midlands too

Alexandra

Hi Mike

just wonderful results and many thanks for the beer

Very funky!

Hi Mike,

As a point of reference, here are my data from the same period. Looks like we're in pretty good agreement.
Stu.

Thanks for all of your support and encouragement folks

This thread seems to have generated a lot of interest and I hope other people will be motivated to have a go.
For me, it has provided precisely the complex, multi-variable, problem solving challenge which attracted me to science in the first place

With any luck, the 'thermostatically controlled heat mat in a cool box' solution will provide a reliable long-term monitoring tool - watch this space

Hi Mike and Stuart

you've done it, congratulations

Well, the weather has warmed up a little over the last few days and what I was afraid of has happened
The thermostat in the fridge cuts in as the temperature rises and we get a lovely 'spike' as the cooling cycle responds.

As can be seen from the image shown below, even a 2.5 degree change totally distorts the output from the Bat detector.
As the outside temperature continued to rise it became obvious that this 'cooling cycle' really sounded the death rattle for the 'fridge solution'

In anticipation of this problem I ordered the same heating mat and thermostat which Stuart uses. It arrived yesterday and I set up my 'foam wrapped' Bat detector in an old cool box along with the mat. The thermostat was set to 27 degrees C, my temperature probe inserted into the wrapping around the Bat detector and the whole lot left to equilibrate.

As the temperature of the Bat detector increased from about 8 to 27 degrees C, the output frequency changed dramatically but eventually stabilised and I left it to run overnight.
There is good news and bad news; as always

The thermostat has reduced the 'overnight temperature drop' and the 'central heating 8am boost' to less than 1 degree C.
The bad news is that even this 'oscillation' is enough to distort the trace

Next step is to try to improve the positioning and insulation around the cool box in an attempt to 'damp out' the temperature fluctuation to something close to the 0.3 degrees which is required for excellent results - watch this space

watching with interest!

Hi Mike,

It still surprises me just how temperature sensitive the Magenta 5 actually is, considering it's supposed to be 'digital quartz' tuned. More insulation around the box should help, of course. Warmer weather will also help


Stu.

Hi Mike,

Another option to consider, maybe, is using a high heat capacity filler within the box, surrounding the Magenta 5. Maybe cobbles or small stones to fill the space, since these will absorb and release heat, helping to smooth out the variations (like a brick block storage heater). This will be much more efficient than air as a way of storing heat.

Just a thought.

Stu.

Good point Stuart - I had considered going down that route with the FGM-3 after reading a link from the Speake website.
I was slightly worried about what filling or insulation would be safe to use with the heat mat but I guess it generates so little heat that there is no real fire risk.
I would like to avoid taking the Bat detector out of its case as I would have to do to get it inside a vacuum jar as you have done so have been considering alternatives.
My next move (if additional padding around the cool box doesn't work) involves putting the Bat detector in a polythene bag surrounded by sand in a large tin or glass jar - exactly the same logic as your 'storage heater' suggestion.
Watch this space...........