New(ish) Gadget: Eva Mist Ultrasonic humidifier

A little while ago, I ordered and received an Eva Mist Ultrasonic humidifier from H20, as the Russel Hobbs mist fan has 1) a tiny reservoir 2) no humidistat. This rectifies those faults with a more generous 6 liter reservoir and a humidistat which can be set in 5% steps from 40% upwards. Mine is set to 65%, and does fairly well, although with all the windows open on a hot day, it sometimes struggles to keep up - but I'm convinced it's helping. Given the summer heat, raising humidity is a must; some people think that putting your plants over trays of water (with pebbles or some other "wick" in) helps, whilst others say it's virtually worthless. Something that can pump over 6l a day into the air is definitely making a difference! When it's really hot, I turn the mist on the fan on too.

It has some random features like lighting up with various colours, which I keep off!

It's basically a rebadged Eva-Mist EDV2500i.

New Gadget: Light Meter

Tenmars TM-202

I've been trying to learn about the light levels needed for various orchids to grow happily. "Quantities" like "50% shade" don't help very much when you don't have a greenhouse or shade-house. However, many orchid light preferences are available in a unit called a "foot candle", which of course the human eye is ill equipped to assess.

Fortunately, there are toys that can help with that... I bought a Tenmars TM-202 from instruments.co.za, and have been happily wandering around the house thrusting it in the general direction of windows everywhere...!

I knew that Phalaenopsis were fairly "low light", whilst Cattleyas were fairly "high light", however, equipped with the knowledge that foot candle requirements of many plants were known, I set out to figure out what was actually needed. A bit of judicious googling yields this handy reference for the genera and intergeneric hybrids I'm growing:

• Angraecum: Species dependent; sesquipedale: 2,500-4,000, apparently preferring very bright light, perhaps with morning sun.
• Ansellia: 2,500-5,000, apparently preferring very bright light, perhaps with morning sun.
• Ascocentrum: 3,000-3,500, apparently preferring very bright light, perhaps with morning sun.
• Beallara:
• Cattleya: 3,000-3,500
• Cymbidium: Standard: 2,000-4,000 Miniature: 1,000-3,500
• Dendrobium: 1,500-4,000 (kingianum is probably at the upper end of this)
• Disa: 1,500-2,000
• Laelia: 2,000-3,500
• Laeliocattleya: Presumably takes into account their ancestry, so 2,000-3,500
• Lycaste: (deciduous spp) 2,000-4,000
• Oncidium: 2,000-4,000
• Paphiopedilum: 2,000-3,000
• Phalaenopsis: 1,000-1,500
• Phragmipedium: 2,500-3,500
• Rhynchostylis: ~3,000
• Sophrolaeliocattleya: Presumably takes into account their ancestry, so 2,000-3,500
• Stenoglottis: 2,500-3,500
• Zygopetalum: Similar to Cymbidiums, so 2,000-4,000, probably at the upper end

So, all in all, it looks like I need to get more light in here...!

Incidentally, you can get an idea of the light intensity using a camera's built in light meter - at least if it displays the exposure it thinks is required. Using a piece of white paper, with the camera lens at a distance of about 30cm from the paper, at f2.8 with a film speed of 200, the shutter speed is (very...) approximately equal to foot candles. If that sounds like too much maths to you, First Rays have a handy form you can fill in that works it out for you.

You can also work this out at any settings with the following formula:

6*F^2/(I*E)

where F = F stop (aperture)
I = ISO setting
E = Shutter Speed (exposure)

I did this in my office, and I got:

6 * 2.6*2.6 /(80*1/33) = 16.731 FC (wow, that is dim).

If you don't have a camera with an obvious light meter, even your cellphone camera will give you this information if you can read the EXIM data (on my Android phone, I can see these details for each picture by tapping the menu button, then tapping Details and scrolling down); again, take a picture of a plain white sheet of paper with the lens about 30cm away and filling the field of view.  

Graphing the interSeptor

After a considerable amount of mucking around, I persuaded various things to work together, and I now have Cacti making nice web-accessible graphs for me. My next project is to get a small "widget" display for OOAB that shows Current/Max/Min/Av values for temperature and RH.


If anyone ever needs them, the Jacarta interSeptor SNMP OIDs you need to read the Temp and Humidity are:

.1.3.6.1.4.1.19011.1.3.1.1.3.2.1.3 = current temperature
.1.3.6.1.4.1.19011.1.3.1.1.3.2.1.5 = current RH 

The values need to be multiplied by 0.1 to get a real temperature, as it stores the value as an integer with .1 of a degree increments.

Basically, the steps are to:

1. Get a dyndns account or similar and make your router use it (if your machine/router/ has a dynamic IP address)  
2. Use the port forward/firewall function of your router to forward any SNMP requests (port 161) to your interSeptor's internal IP address
3. Install Cacti
4. Use the SNMP OIDs to create graphs. 
5. Because cacti uses dynamic references to its graphs, you're going to need to publish the graphs somewhere using the built in FTP function, or publish them to another web-accessible directory on the server running cacti. 

Temperature
Relative Humidity
Graphs
(these are still a work in progress)

New Gadget: Jacarta interSeptor

Today I received a Jacarta interSeptor from RS Components, which I plan to use for monitoring the balcony's temperature and humidity levels. Essentially, this (rather pricey) toy sticks a thermometer and hygrometer into a box that speaks http. I could theoretically add another sensor to monitor temperatures and humidity elsewhere (like in my wine rack!) but the optional sensors are also $$$.

The little sensors can be on the end of up to 20m of straight through Ethernet cabling, and the interSeptor unit itself can be anywhere with a network point and some power (why the heck don't they support PoE?!). You can also use the sensors to monitor high/low or open/closed circuits, so you can interface them with other things, along with a host of "official" sensors.

The sensor specifications recommend operating temperatures of 0-80ºC and 10-90% RH. You can set alarms ("thresholds") from -15 to 65ºC and from 5-95% RH. The accuracy is stated as +/- 1ºC and +/- 3% RH, which seems adequate for my purposes (although the RH could do with being a little higher).

Several somewhat minor but quite frustrating things immediately hit me:

1. It has a Fat Plug (aka "wall wart")
2. It's an English-style Fat Plug, which doesn't plug into sockets in South Africa
3. It doesn't work like they say it does.

I hate Fat Plugs with a passion. Fat plugs are generally a major pain in the rear end to fit into multiplug adaptors. Having a UK plug on a product sold in South Africa is somewhat annoying - fortunately, I had an adaptor lying around I could use. That was a minor inconvenience compared to trying to talk to the thing.

In terms of "not working", theoretically, the interSeptor should use DHCP to pick up an IP address. However, for some reason, v2.03 of the firmware for these devices doesn't actually seem to work with DHCP servers (I tried two different ones, one on a Mikrotik router, and one on a Netgear router). With the Mikrotik, I could see the DHCP server offering a lease, but the interSeptor wasn't accepting it (or rejecting it) - it was point blank refusing it.

I wrote to their technical support, explaining the DHCP problem. They sent me the latest version of the firmware, v2.08. Of course, there's a bit of a Catch 22 here, in that if it doesn't DHCP, you can't upgrade the firmware, because you can't actually connect to the device... To their credit, they respond pretty fast!

Solving intractable problems like this often takes up chunks of my time at work, so I thought I'd try a different approach. I have a Sunix USB to Serial (RS232) adapter, and as the interSeptor comes with a serial cable and instructions on how to configure it through there, I thought I'd give it a try - after downloading drivers for Windows 7, and TeraTerm, as HyperTerm is no longer part of Windows. No joy at all on either COM port, or plugged into either of the 2 serial ports on the Sunix adapter. Irksome.

Being a network device, these little gadgets come pre-configured with an IP address, so as a last ditch attempt to connect to the blasted thing, I decided to try changing my computer's IP address manually to something within the same subnet range as the pre-configured IP on the interSeptor, and simply plugging a straight Ethernet cable between my computer and the device. This, of course, worked. I upgraded the firmware, and now the device quite happily talks DHCP.

Now comes the frustrating bit...!

The whole point of getting something like this is to let me keep a virtual eye on the plants when I'm not at home. What I'd like to do would be to have the device accessible over the Internet to Those In The Know (i.e. me). It would also be quite cool to have a little applet that sits on the blog here showing the live and historical temperature and humidity levels.

Whilst I can make it work oh-so-happily on the LAN (from either of the two subnets) I *cannot* persuade it to talk to the broader "internet". I have yet to pin down whether it's an issue with the Netgear router not doing it's firewall rules correctly (tried updating that to the latest firmware!) or that the interSeptor only responds to local (RFC1918 addresses), or those within the same subnet - given that I have NAT between the two subnets here, no matter which one I'm actually in, the devices on the other side will still "think" I'm in their subnet.

I've spent pretty much my whole evening on this now and am feeling like it's time for bed. There *must* be a sneaky way I can get information out across the Internet that doesn't rely on me setting up email alerts - but that's a job for another day... SNMP might be the way (and the included CD has the MIB).

The little Java viewer applets are quite cool, but only available within the LAN so far :/

These gadgets have a lot of promise, but so far, GRRR!!! To be more useful to gardeners, they should make the sensors weatherproof, make them work with DHCP (which they have done - IF yours has the right firmware) and I think the sensors could do with measuring up to 100% RH.

This "review" might seem a little negative at the moment, but I'm sure as soon as I have it worked out, I'll be singing its praises...!

Under (Water) Pressure

Given the somewhat terrible water around these parts, and rather than spending a small fortune buying distilled water from a pharmacy (or begging it from people who have lab grade distillers or RO machines), I've been considering getting a fairly hefty Reverse Osmosis (RO) unit, not only for the benefit of the orchids, but also the fish (who currently get bottled water, as do the rats), and of course, the humans around OoaB.

When I asked for a quote and about the need (or otherwise) for a pressure booster pump, "The Waterboy", Marco Justino, based in Durban, said that this was not required as long as the incoming water pressure was over 3 bar.

I said I'd try and figure out a way of measuring the water pressure here and get back to him.

I spent a frustrating 30 minutes or so going around all the places that might conceivably have a pool filter pressure gauge. None did, not even the pool shop. Eventually, I thought laterally and wondered if I might persuade a car oil pressure gauge to work (at least briefly); a quick trip to Midas and R90 later, I had a gauge that could read up to about 7 bar. A few Gardena hose fittings on top, and I had what I hoped would be a workable water pressure gauge.

Oil pressure gauge and Gardena fitting

Conveniently, one of the fittings that came with the oil pressure gauge happened to be just about the right size to self-tap into the plastic of the Gardena fitting. Yay!

Ooer, it actually works!

Somewhat to my surprise, the jury-rigged gadget worked, and showed the pressure on the cold tap in the bathroom to be about 4 Bar (coincidentally enough to just pump a column of water about 40m into the air). Why the hot water pressure in our house is therefore so totally pathetic (taking a hundred years to fill the kitchen sink and giving us showers worthy of the worst in the UK, legendary home of pathetic water pressure), I don't know, but I suspect is has something to do with the geyser (immersion water heater) that looks like it was wrought out of lead, by gnomes at work deep in a mine, in about 1824.

So it looks like there's enough pressure for RO, yay, which saves about R800 on a booster pump. Still, given the fairly high TDS/EC, I wonder if a higher pressure might not be needed to force that water across a membrane?

Incidentally, I tested (pre calibration) the Oasis ozonated water available from the Oasis filtered water shop here, and it has an EC of about 8 - which will do for now. And at R4.50 for 5 litres, is a whole lot more affordable than R37.15 for 5l, or R13.75 for 1l that pharmaceutical grade distilled water costs!

Of course, I imagine that water and oil pressure gauges are not happy buddies, and I doubt that gauge will work again once its insides corrode...

Calibrating a Hanna HI 98129

A few posts ago, I talked about a combined pH and EC meter (Hanna HI 98129) I recently acquired from RS Components (who have an excellent overnight service on anything they have in stock - even to a backend Dorpie like Grahamstown)!

Of course, like any scientific instrument, it needs a baseline from which to work in order to ensure accuracy. For this, you need to calibrate your instrumentation. Some instruments drift a *lot* over time - so much so that they can become dangerously misleading if you rely on them and don't realise that the values can drift! So, in the interests of accurate readings, I ordered some calibration and storage solutions:

To calibrate pH, you need two solutions:
Hanna HI-7007/1L, a pH 7.01 buffer
Hanna HI-7004/1L a pH 4.01 buffer

To calibrate EC, you need one solution:
Hanna HI-7031, a solution with an EC of 1,413μS

Hanna recommend storing their pH electrodes in a particular storage solution,
Hanna HI-70300

I'm still waiting for the electrode cleaning solution,
Hanna HI-7061

Interestingly, these solutions cost about the same as the instrument itself... Science works, but the toys can get pricey, fast!

Here's a random shot of the calibration and storage solutions around the instrument which is sitting in pH 4.01 buffer. Note the pH had already drifted out by .01 in the time it took me to assemble the shot!

A few tips:

• Make sure you rinse out the container you're doing the measurements a couple of times with the reagents you want to use for calibration (and ideally the instrument too) - this should remove any contamination which might affect the readings. This includes previous buffers and solutions!
• When you do a two point pH calibration, don't rinse the electrode with pH 4.01 buffer, it assumes this is immersing it in the buffer - instead, add slightly more to the container you're doing the calibrations in and swish the instrument back and forth a couple of times. If you don't, I find the calibration goes waaaay out. 
• When you're trying to do the EC calibration, you have to start in EC measurement mode (NOT pH) - the manual isn't explicit here, so I then ended up having to redo my pH calibration. Again.
• Storing a pH sensor in the recommended solution is usually a good idea; the manual specifically warns against storing it dry (requiring a one hour rehydration in storage solution) or in distilled/tap water, which (I assume) causes the delicate ionic balance inside the electrode to go awry.  I'm storing mine with a few drops of the solution in the pH electrode "cup" inside the lid, standing upright in a glass, hopefully preventing leaks. The EC of this solution is higher than the measurement range of this instrument, so I'm not sure flooding the entire cap is a good idea. 
• Ideally, calibrations should be performed at STP (Sea level, 25ºC), although these meters can theoretically cope with deviations from this as they have a temperature correction factor built in. Temperature affects both the pH of buffers and EC values of salt solutions (in fact the labels have the correct values at a range of temperatures printed on them). This evening, it was about 27 degrees in my house when I did this (and very humid!).

New Gadget: 40cm "Mist Fan"

A lot has been written about the importance of air movement and humidity for orchid culture. Our little balcony can get stiflingly hot during the day (indeed, by 8am it's often over 30ºC in there) and in January and particularly February, Grahamstown air temperatures can regularly reach into the 40s.

As much as I am tempted to demolish a wall/replace windows and install a massive fan in the other side or the balcony to get a wet wall evaporative greenhouse cooler up and running 1) it's very expensive 2) my landlord would have a fit 3) I suspect the Grahamstown Aesthetics Committee would object to the rather grievous alteration to a building from the 1800s! And a misting system would result in very wet floors (not to mention there's no water supply out there)...

A few weeks ago, I saw some "ultrasonic mist fans" in a shop. I decided that these might well help out, in that fans provide air movement, mist provides humidity and evaporating mist cools air. I ordered a Russel Hobbs 41cm Mist Fan from Makro online and it was delivered the other day.

After setting it up, I played with the various modes; it even has a "natural wind" function that varies the fan speed constantly. A very rudimentary trial suggests it cools the air by about one degree Celcius whilst the mister is on. The 1.5l water reservoir has to be refilled about three times a day, which is a bit of a pain. Another slight annoyance is that if there is a power cut (or you turn it off at the wall or via a timer etc) it resets itself to an "off" condition, which pretty much eliminates any hope for automation. :(

Until I receive my rather fancy LAN-enabled thermohygrometer (usually used for monitoring computer datacentres!), I'm not sure what it's doing to the humidity; I haven't been able to find even a basic hygrometer in town.

It's a bit too early to tell what the plants think of their new more "buoyant" environment, but I hope they like it...

New Gadget: Combo pH & EC meter

Yesterday I received a Hanna Instruments HI 98129 combo pH and EC meter. Grahamstown has a chequered history of water quality; debate has been raging over the last few years whether it's even fit for human consumption, let alone raising fish or fairly delicate plants like orchids.

Most orchids like their water to be slightly on the acidic side; a pH meter lets you see just how acidic or alkaline your water is.

Electrical conductivity of water gives you an approximation of the amount of salts dissolved in the water; pure water hardly conducts electricity at all, whilst water with a lot of salts and minerals in it conducts electricity quite well (which means you can use distilled water to put out electrical fires in server rooms, but spilling salt water on electronics usually results in them not ever working again). Most orchids are only ever watered by rainfall, which is pretty close to distilled water in its composition - hardly any mineral salts and a slightly acidic pH.

One generally expects that replicating the conditions orchids usually grow in in the wild will result in success at home; long experience has shown people that this is indeed the case.

There is of course another source of salts which we generally consider "good" for our plants - fertilisers, which are made up of various salts and minerals, which of course affect the composition of the water. Years of advice has suggested that fairly weak concentrations of fertilisers on a fairly frequent basis ("weakly, weekly") are the best bet for orchid cultivation. Too much fertiliser can be worse that too little (just as more orchids are killed by over-watering than drying out!).

I hauled the meter out of its box this morning and quickly did some tests. As I expected, Grahamstown water did not fare well:

Grahamstown Tapwater: pH 8.20, EC 669μS/cm  - quite alkaline and rather salty (for orchids).
Aquelle spring water: pH 7.2, EC 72  μS/cm - not too bad!
Distilled water: pH 7.05, EC 3μS/cm - pretty good, as you'd expect. 

Experience has shown that different kinds of orchids will tolerate different levels of dissolved salts in their water before it essentially has the same effect as you drinking seawater - i.e. it dehydrates you and is no good at all, despite your best intentions.

I found a reference on the Eastern Province Orchid Society's website which suggests the following thresholds:

400μS/cm: Cattleya, Paphiopedilum
600μS/cm: Odontoglossum, Coelogyne, Lycaste, Oncidium, Masdevallia, Miltonia
800μS/cm: Phalaenopsis, Cymbidium

That means that even without fertiliser added, Grahamstown tapwater is unsuitable for all but the most commonly and widely grown orchids (Phalaenopsis and Cymbidium). And I suspect Disa are real pansies and need a lot lower than that! Buying distilled water and/or spring water to water your plants seems a little excessive, so I'm going to test some other possible sources, particularly spring water. Given that we live in a rented second story flat, there isn't really any feasible way of harvesting rainwater, and Reverse Osmosis units require plumbing modifications, unfortunately!

Next I'll have to test the various fertilisers we use to see how they affect the conductivity and pH and just how much we should be using... I've also ordered some calibration and storage solutions for the meter to make sure it's accurate and stays that way!