Cumulus Support

Hi All
A note of caution - I only used 4.5V for a very short time - half and hour or so. The 3.2 V supply has been going some time and that does seem to improve the transmission/reception. I don't have a very good line of sight - through slate, blockwork and somehow the signal manages to get round or through odd bits of granite wall so I need all the power I can get. Oh ! I forgot the 11000V overhead cables and the neighbour with a similar unit !!
Regards
Alan

I'm still interested in using a small solar panel to charge 3 NIMH batteries in series with a diode in series with the output to the fine offset. This ought if the batteries are fully charged to reduce the fully charged voltage of three in series from 4.2 volts to 3.25 volts, which would be fine I think.

Then we have the possible problem of a small solar panel overcharging the three batteries. In summer, this would be a very likely, since the weather station uses so little power and the panel I have would be putting out six volts. To solve that, I think we could make a very simple voltage regulator with the LM317 which can be set up in a current limiting fashion so that the batteries would not be spoiled by over-charging. The circuit is REALLY simple and can be found here with a chart to show what value of control resistor would be needed for different maximum currents.

http://www.reuk.co.uk/Solar-Battery-Cha ... LM317T.htm





Maybe Gina could comment on whether a six volt panel could be limited to a satisfactory extent by using this circuit into three series connected nIMH batteries and then feeding their output to the Fine Offset through a 1n4001 diode to drop the voltage by 0.8v? I think that if things were set up as I described, the circuit voltage at the batteries could never rise above the fully charged voltage of the batteries at 1.35 to 1.4 volts each (x3 = 4.05v to 4.2v). A series diode to the station power connector should drop this by 0.8v to 3.25v to 3.6v.

Thanks


Tony

Hi Tony,

You don't need to worry about overcharging on these small systems, as long as you use relatively small area solar cells. Basically, any energy not stored by the batteries is dumped as heat, so a cell putting out a few 10's of mA will cause a few mW to be dissipated as heat in the batteries. This is why the garden lights have no regulation system. As a general rule, you can trickle charge NiCd or NiMH batteries at under 10% of their rated capacity forever without damage. For example, if you purchase AA's rated 2000 mAH, you could put 200 mA constantly into them without damage or overheating. Most hobby class solar panels supply far less than this - mine put out about 5 mA / sq. mm. If you are salvaging cells from garden lights, for instance, they seem to supply at about 15 - 25 mA, or about 1% of the rated capacity above. You might find some debate about the 10% figure online, but nobody will dispute 1%.

If you are going to use scrap garden light cell segments, will need 9 or 10 segments as each segment puts out about 0.5V in full sunlight. My plan involves 9 segments = 4.5V @ 20 mA in full sun. This flows through a diode, to make ~3.9 V available to my NimH cells. (These are just what are in my bin - Nicad would be cheaper and as good or better in this application). The 3 cells in series will charge up to a bit over 3.6 V, then feed through a diode to provide ~3.0 V to the weather station. Diodes are 1N4001 that are also in the junk bin. The tenth segment would give you a little extra margin for cloudy winter days if you are concerned (I'm not)

I have only built this on paper so far. This should last for a few years, at which time the station will likely be obsolete anyway as we'll all be forecasting the weather by telepathy, or the professional services will be offering customized reports down to the square meter!

One further thought: I've used the 317 as a current regulator several times. The circuit in the link will work fine if you really want to use the oversized solar panel. The excess energy will be dumped as heat in the 317 so you might want to add a heat sink to it. You should also add a series diode from the solar panel to the 317 to protect both when it gets dark.
Also, your 400x diodes will have different forward volt drops with current, ranging from ~0.6 V @ 10 mA to ~0.85 V @ 200 mA.

Thanks for that info Charlie. I don't have scrap garden lights around, but I do have a solar charger unit that I bought off Ebay last year. It is a pretty good charger and in summer would fully charge flat AA batteries in just about a day. It has 28 cells. This panel is attached to a box with contacts for charging AAA, AA and D batteries. It is marked '150 - 7 volts' so the cells must be connected in pairs in parallel and then each pair must be in series with the rest to get the seven volts. I am thinking it can put out maybe more than three AA batteries can handle, even with a voltage dropping diode or two in series. For this reason, I figured the control circuit would be a good idea.

I also have about five of the LM217 voltage regulators somewhere. I bought them off ebay for a project I never got around to. Now I just have to find them.

Anyway - the project can wait a while yet. Not much sun around here lately.

I don't think there's much I can add to what Charlie has said. If you need to reduce the current, that circuit would be fine. I'm not familiar with charging NiCd or NiMh cells/batteries but I have made a lead-acid battery charger. This has both current and voltage limiting based on the LM317T (plus power transistors) and switches to a low top-up current once the battery has reached full voltage. This was designed for car and tractor battery charging and would be overkill for this little job. Cheap garden solar lights can be had for ten or twenty pounds for ten or similar. The panels give about 3v in full sunlight but very little current as Charlie said. I was thinking of using 6 for charging 3 NiCd batteries in series with the panels in series/parallel arrangement (2 in parallel, 3 in series) but maybe 3 would do. They're certainly the cheapest source of solar panels that I've found

Thanks for you thoughts on the subject Gina.

christer wrote:

Charlie wrote:Yes, please let us know if the batteries are the problem, and please measure the batteries before you throw them out. If the lithium batteries have died after only 5 months, this is very bad news. Based on what others are seeing, I would have expected more than 5 years! Were they new when you put them in place?

Yes! I will report. But it seems to be too cold for at least one week more. So I will wait until it´s more conveniant to change the batteries.
On the other hand I think of install a battery eleminator during the summer. The fact is that the steel tube that holds the outdoor unit 5 merer above ground is bolted to incoming mains supply fuse box at our RC airfield.

//Christer
ps minimum temp = -28.6C http://mfkjupiter.se/vader/record.htm /ds

Well, yesterday my outdoor sensor started to deliver data without changing the batteries! But. The temp har gone up from -20C to only -5C. Maybee the RF power is dependent of the temp? I will check the batteries later.

Regards
Christer

Electronic devices can certain perform poorly (or cease altogether) at low temperatures. Components have to be specifically certified to work at these low temperatures. eg. commercial ICs are specified for 0-70 C - a wider range requires military spec components.

Charlie wrote:...If you are going to use scrap garden light cell segments, will need 9 or 10 segments as each segment puts out about 0.5V in full sunlight. My plan involves 9 segments = 4.5V @ 20 mA in full sun...

Charlie, from my 'ancient' days as being an electronic technician, putting 9 solar cells in series would produce the total voltage of approximately 4.5 (as you claimed ie: 9 * 0.5v = 4.5v).

But there is also the question of the current - voltage is not enough.

By putting the solar cells in series would produce a total current of only 15 - 25 mA.

In order to increase the current, you would need to have the same number of solar cells but this time wired in parallel.

To produce a decent current (such as something around 60 - 100 mA) you would need a total 36 solar cells, 4 groups of 9 series-connected solar cells.

And the use of a voltage/current regulator is highly recommended - series-connected diodes would not be safe enough.

gemini06720 wrote:

Charlie wrote: And the use of a voltage/current regulator is highly recommended - series-connected diodes would not be safe enough.

I was wondering what would happen if say the output of an LM317 at 4.5volts was fed into 3 series connected nimh cells. I expect that the fully charged combined battery potential of around 4.2v would not be exceeded until the batteries stopped taking any more current,but given that the Fine Offset stations draw so little current themselves, is it possible that the no load voltage on the system could creep higher?

I ask this because I have noticed by sticking a volt meter across them that some cheaper regulated supplies deliver higher voltages than the rated one, until a load is applied. Maybe this wouldn't make any difference - I don't know, but for sure a weather station that can last a year or more on a pair of alkaline batteries, must be drawing next to no current, and might appear as virtually a 'no load' situation to a voltage regulator. I'd not be pleased if I blew the sensor / transmitter unit by fiddling about.

I think I read somewhere that 'modern' units of this type effectively go into 'stand-by' mode inbetween transmit pulses in order to acheive the battery saving.

But that might be gibberish - anyone confirm

beteljuice wrote:I think I read somewhere that 'modern' units of this type effectively go into 'stand-by' mode inbetween transmit pulses in order to acheive the battery saving.

But that might be gibberish - anyone confirm

I guess that's possible but then how would the unit for example, know the highest gust speed since the last transmit. I would imagine the energy saving comes from only sending the data every 48 secs whereas more frequent would mean more power.

I meant that the transmit 'carrier' is not left open. This being the greatest current consumer.

hmmm ...... but then how would they know if they lose contact, unless it's a time-out thing.

I was trying to figure out when I said that, because I thought I said the exact opposite. It turns out the quote was edited a bit too agressively, and attributed a response to me.

It's true that the open circuit voltage could rise to 4.5 V as a static potential. Some devices could be damaged by just a few electrons at this level of overvoltage. Fortunately, none of those devices are in this station, which is lucky or they would never survive the first thunderstorm of the season! Someone else posted running their system at 4.5 V for 30 minutes with no ill effects, so I don't think this is an issue. Also, for any meaningful current to flow, the diodes will need to conduct, which means their forward volt drop is in play, dropping voltage to the ~3V as outlined.

As to the earlier post about current from garder lights, the garden warts do indeed only put out on the order of 20 ma, and this is tons for this application. There is no need for parallel segments.

I'm planning to put a meter on this beast and see how much current it takes, but I suspect it is pulsed (at least the transmitter) and I won't be able to get a reading.