Hey,

I’m new to these forums but I thought I’d write about my mission to get a decent environmental control system and what I’ve decided to do. I'll update this thread as I go along so feel free to make your comments.

I’m going to make a box that will do the following stuff:


Intelligent temperature control
I want my box to drop the temperature slightly at night using the fans or turning on an air conditioning unit. I want to be able to set a min and max temperature for both time zones. I want to be able to see my incoming air temperature so that I can make best use of an airconditioning unit if I have one to use.

Automatic fan control
My plan is to use a variac to control the van speed, and use a stepper motor to control the variac. The fan speed could then be set on a completely linear basis so you never end up switching to different speeds. This will keep the temperature very nearly constant and noise to a minimum.

Humidity control
I’d like my dehumidifier turning on when it needs to be on. I’d also like it to be off if the fans are turning over above a set level since there would be very little point trying to dry out constantly changing air.

Wireless interface
Yes, you heard it right I want to be able to control my controller from my browser. This is very optimistic but what the hell, these are just plans

Ill write some code for a micro controller to handle all of this and plug the sensors in to do the monitoring. Simple eh? Well maybe not but it might be an interesting journey

ToMMY2k

Sound ambitious. What micro controller you thinking of using - PICS?

Sounds perfect.

When it's made and tested I will buy one please. sounds great. Good luck

Yeah probably Although I will be using something like MatchPort for the wifi - this would get expensive with the wifi option



You couldn't afford it haha

For fun Ill probably make one and see whats involved although at this stage the idea of selling them would remove the fun from the project.

ToMMY2k

Ok some bits arrived.

I got the variac, two pullys, timing belt, grub screws, drill bits for making the holes for the grub screws and a set of taps for threading the tiny holes.

Click to view attachment

More images to follow

ToMMY2k

about 12 years ago, i used an old dos 286 pc with an i/o card, think it was a 1512, to control lights fans a heater and take input from temp senor. wrote it in basic, it kinda worked, but my instrumentation/control/software skills aren't really up to what you're gonna do. i never ran it for a whole grow, just an experiment at the time. i would have thought some company would have came up with a commercial solution to this by now. maybe some piece of hardware controlled by a dedicated iphone app?

Its does seem obvious although this does exist. There are commercial controllers that do this sort of thing but they cost a lot of money and are for much larger scale operations. There isnt enough market for a niche product like this for a company to devote and real resources to.

Anyway, Im currently hunting for an enclosure the right size and it getting quite hard to find something just right. Everything is either massive or too small

Ill keep hunting.

Cheers,
Tom

enclosures? usually maplin or electrical wholesalers is the way to go, but i expect you've already tried there.

Try looking into...........

Honeywell Spyder

There are lots of different BMS type controllers..............

Priva systems are quite "cheap".

A quick search and I couldnt actually find a price or a uk stockist...any ideas?

ToMMY2k

I think Ive found the enclosure I need at farnell, it aint cheap tho. £40

http://uk.farnell.com/jsp/search/productde...jsp?sku=1004142

And Ill also need:

http://uk.farnell.com/jsp/search/productde....jsp?sku=208619

ToMMY2k

Priva

here's my temp contoller,works an absolute treat an only cost about £60. http://www.uk420.com/boards/index.php?show...c=73144&hl= instead of a stepper motor 2 control speed you could just use a electronic variac an have switched resistances to change speeds.

Thanks for the info, I did consider using an electronic variac, then I considered using fixed transformers and switching between 5 fixed speeds, then I finally landed on a proper variac controlled automatically by a stepper and microcontroller.

You could say this was overkill but I like to make things and the real work isnt the in mechanics and electronic circuit, its in the programming of the controller to take various sensor inputs, and control multiple outputs and all the logic in between.

An "electronic variac" aka dimmer is nothing more than a PWM controller which does nothing more then modify a normal sine wave by chopping bits off. This damages fans and can cause humming, which is actually the sharp voltage spikes of the PWM modified sine wave causing huge magnetic forces on fan rotor and these forces then inducing vibration which you can hear as the hum. Using resistance to control an inductive load is also not a good idea as it will basically create lots of heat.

The part that I want to achieve is a constant temperature. Any sort of switching between one speed or another would not let me achive this.

The variac option while much more expensive is much kinder to your fans, should be more efficient than resistors, will provide the ultimate control of the fan speed and for the sake of an extra £40 for the stepper, pullies and variac to protect my £150 fan I know which one Id choose

That said your controller is far better value than mine and if your happy with it then thats what counts

ToMMY2k

i see, you are trying to go for a full analogue output control, adjustable to any setting, rather than a set number of different output steps for fan speed control. a stepper motor might not be the best choice though, since it also rotates in steps, i guess you know that though. i used to maintain hvac plant, we used a belimo damper actuator, which was in effect a geared shaft driven by a small dc motor, perfect to get the kind of exact angular positioning you would need. its an enclosed 3 wire geared motor, 2 wires dc supply to the motor, 3rd wire a direction wire. think they were 24v motor supply, 5v control for direction. slow enough speed to adjust airflow dampers on a duct, so would be ok to position the variac control slowly to exact position. would take a bit of experimenting and setting up though.

An idea i was tinkering with a few years ago was a signal generator and power amplifier,varying the frequency changes the motor speed, unfortunately the amp had to go back to its owner so i sort of gave up on it.

If the stepper motor is geared correctly you should get accurate control anyway. Nice work by the way Tom, keeping an eye on this one.

There's a German outfit, Cleware, that do usb temp/humidity sensors plus mains relays. Let's you turn things on/off in relation to the temp or humidity.

Very impressive Tommy, interesting project you got going mate will be superb to see it up and running, I see many an hour being tinkered away should be great fun
best of luck getting it all set up
Rj

Thanks for all the comments peeps

The UPS guy (a alyways think UPS seem the most professional delivery service!) just delivered a big box of stuff to my door which included my enclosure which at £40 is perfect while being expensive. Its IP 67 rated so I could probably submerge the unit without a heart stopping electric shock!

The gear ratio of the pully is 4:1, and the stepper has 200 turns per revolution. The windings on the variac are perhaps 1mm apart so Im sure I will definately get 100% resolution from the stepper configuration. The problem way well be that I actually end up stopping between windings so in terms of control I might need to map each winding position out so I can jump between windings instead of just using steps.

mokum777: You are essentially talking about a servo system (I think). A DC motor with encoding. The variac would provide the encoding and feedback to the microcontroller which would in turn control the dc motor. That is more advanced than a stepper and yes it would be been a better choice since I could directly control the voltage output. If the voltage supply dropped slightly I could compensate instead of waiting for a temperature change.

I might just add some sort of output voltage monitoring into the system! Thanks for the idea!

ToMMY2k

i only mentioned it because it was a fully enclosed off the shelf unit, which could do what you want, but its much more fun building it yourself

Ok girls and boys are you sitting down? The next bit to my controller.

Mounting the variac

In my new enclosure I mounted the variac to the chassis plate. A nice easy way to transfer mounting hole patterns to a metal plate is to use a bit of cardboard, and puch holes through the cardboard into the bolt holes in the variac. Then using the holes mark the metal plate. Easy eh?

To drill the holes you need to make sure you alyways drill at least one size up. So for M5 holes I drilled 6.5mm (1.5 sizes up I know). This gives me a bit of room for error in positioning the holes, drilling etc. After that I countersunk the holes to make the plate all health and safely

The next image is a blury shot of the mounting post or spacer as they are known. I had to chop a bit off the bolt as it was way to long for the hole in the varac. I used my angle grinder for this task although you might opt for something a bit safer. An angle grinder takes 5 secs whereas a hacksaw takes a minute and you have to clean up the end of the bolt anyway. I screwed the spacers into the variac and then bolted the spacers into the chassis plate.

And there we have it a mounted variac

One thing I didnt get an image of is me cutting the end of the shaft off. I miscalculated the depth of the spacer off I needed (I guesed off the top of my head). In reality I needed to only keep about 10mm of the shaft visible so the pully could attach on.

And finally...I had to test the variac. It works like a charm!!! Perfect, what seems to be nearly infinte control I noticed that the voltage would drop to about 7.3 volts before the fan wouldnt move but this voltage stayed constant through a range of movement. I need to check the current draw since there will be a cut off point to this too. I wonder if I can automatically detect the best point to turn off the variac? Another feature added perhaps!

Click to view attachment

ToMMY2k

Another installment....two in one day! I must have nothing to do! I wish!!!!

Drilling the pullys

This one was a bit of a pain. I hate drilling aluminimum since it tends to grab the work piece and this when using small drills tends to break the drill bit leaving a section of it stuck inside the hole

First we drill out the hole in the center of the pully. Im of the opinion that you should do this in a single step since you want your hole to be straight and centered. Unfortunately I didnt have the right sized drill bit (I am waiting for a 10.5mm cobolt drill bit to arrive) so I drilled it out anyway using some old tatty blunt probably shapened, maybe my grandads old drill bit.

While not the right size it certainly did produce a smooth hole. No photo of that sorry.

The next bit is the fun bit where we drill out the tiny hole for the set screw (or grub screw as they are sometimes called). I had to do this three times on the large pully. The first time I snapped the drill bit due to the vice that holds the piece moving with vibration.

The second time was due to me pushing down to hard.

The third time was lucky and produced a perfect hole with nice swarf and everything

The moral of this story is that if you have a really shitty old pillar drill you need to spend some money and buy a new one! Im gonna get a meddings.

Anyway the smaller pully (the one that I actually couldnt afford a mistake with) went perfectly

Click to view attachment

ToMMY2k

Tapping the Pullys

We need to get a set screw into the hole so we tap the inside of the hole which makes a thread.

Ive never done anything as small as this so I did it in quarter turn steps...so a quarter turn clockwise (in to the hole) and then backed off a bit, then another quarter turn etc.

It went smooth as you like, no problems and as easy as 123.

The tiny set screws (2.5mm) requre an allen key that is really tiny. I think they need a 1mm key but Ill have to order one and hope it is the right size.

Click to view attachment

ToMMY2k

Im gonna get one of these to add to the controller Being able to see whats going on is alyways useful!

http://uk.farnell.com/batron/bt120032avb-c...phic/dp/1220415

Click to view attachment

ToMMY2k

Dude if that works it will be a quality piece of kit

Best of luck with it

I like the "if that works" bit. hehe

It better had do!!!! I know for a fact Im going to have to do two boards, one to control the system and another to take data from the system and display it. This might be a nice start to make everything modular for when I have enough time to get it working via wifi

Ok on with the show......

ToMMY2k

Circuit

This should be farily straight forward. For anyone whos interested ill post the diagram once it works! For now since I cant progress anywhere else I might as well get this bit done. Time to find my PIC programmer.

I hate working with verbo board

I love making circuits...its like sweet and sour!

Click to view attachment

ToMMY2k

da daaar! Done...well...it needs links making and lots of tracks breaking underneath, and of course soldering, programming, testing, redeveloping and then finally it might be considered done

I cant find the 12v regulator for the stepper side of the supply line ...hmmm those pesky voltage regulators!

Anyway I need to order some more bits for it since I lost my brain while ordering first time around

Click to view attachment

ToMMY2k

Ok Ive not forgotten

Ive had a real nightmare with my old programmer and after lots of wrangling with Microchip Ive finally got a new programmer which has cost me a small fortune. Anyway, Ill keep thinking this is for the fun of it and Im sure eventually I will justify it to myself

A few updates..why not its been a while!

Ive got all of the bits to make this work now and Ive decided to make this all modular for those who want to attept to have a go assembling one for themselves once Ive got mine working!

Im going to make the main controller box accept connections (probably from a 3.5mm jack [think headphones]) for all of the sensors. The current chip Im using will take about 10 inputs so you could in theory plug in 10 sensors.

Im also going to seperate all of the outputs...so you could have a variac controller in a seperate box and somewhere else have a relay controller for turning a heater on for example. Again you could plug these controls into the main controller box located somewhere else.

Im working on the code to take a a sensors output, to control the variac. Id like to be able to customise the controls but that will come later. I have lots of details to fix like calibration of the temp sensor for example..the list is endless. My modular design might sound complicated but its actually simpler than making an all in one box that does everything.

ToMMY2k

youll need to get a step by step guide up how to make one of them mate
with as many pictures and a shopping list for us tokers

thats sounds like it will be the bollocks when its done mate

Ill probably just sell the one I have and start again

There have been loads of times Ive thought ... "if only I could turn this off OR on when it get this warm" All of the controllers Ive seen dont really cater for small operations with large scale customisation in mind.

Imagine being able to have 10 sensors all being able to control all of the outputs based on certain conditions...time of day...what another sensor is reading etc.

Im getting all excited hehe

ToMMY2k

Mains socket boxes plus blank top plates make ok project cases.

as do biscuit tins and ice cream tubs!

loving this idea , don't forget to update! electronics has never agreed with me but always interests me

you'll need to set the cut off point at around 35% output, or you'll burn your fan out

The motors are brushless AC (well at least the Ruck fan I have is ) ...so as long as the stator is rotating there shouldnt be any problems. You wouldnt want to be pushing current through a static fan I will accept that.

By 35% output Ill assume you mean 35% of the rated air flow. Is this because of overheating?

ToMMY2k

35% nominal voltage, and yes, because of overheating, you wouldn't be able to control the motors via voltage at all if it weren't for the fact they're designed with external rotors, this is so the air flow cools the excess current caused by the rotor slip, you should never control any other type of ac motor via voltage alone, as the motor slows because the voltage can't produce a strong enough flux to hold the rotor at speed, and slip causes high inductance which in turn causes high current (hence large start up currents with ac motors). rvk list the lowest voltage at 105v which is actually more than 35% and that's without a cf, so maybe even 35% is too low

Check out 'cleware'.


They do usb sensors and mains relays, free software for it too. Among other things you could run a Planet Ito lighting schedule. (7 days and nights for every 5 of our Earth days.)

Only temp/humidity on the sensors but if we could get out heads together maybe we could add pH, ec etc.

Looks good but unfortunately you need a PC to make that system work. I wouldnt trust USB to feed my cat reliably either It might be fine for getting your photos off your camera or printing to but 24/7 operation over a number of months? Ive done a few projects using USB and it really can throw up some unusual problems!

It does look good tho...its a shame you need a PC to connect to those devices because if the sensors were I2C based things could be really portable!

Ill send them an email and see what they say

As for pH and ec...I think the easiest thing there would be to hack a cheap meter and probe the display for the reading.

ToMMY2k

Hmmm...Im with you there but a variac being an auto transformer means that it has an output impedance which increases as you wind the transformer down.

As soon as you short the leads of the transformer (when wound down) the voltage drops to 0 and the current flowing through the circuit would also drop to the rated output of the transformer. Given that the transformer might only be rated for 3A, at a 5% winding you might see 150mA and unloaded the transformer might only present 12v. That is (in the scale of full load), virtually nothing and these calculations are firstly very rough and sencodly based on a complete short of the transformer. Add the fans impendance into the mix and Im sure it would be much much less.

I suppose you could reduce the frequency of the sinewave so that a smaller proportion of slip was produced but you would defeat the point of using a variac in the first place if you could provide a powerful adjustable sine wave source

I think it might be an idea to measure the current being drawn at specific voltages and also measure the temp of the rotor to see exactly what is going on. I think the point here is that a variac is far less damaging to the fan than a PWM based approach and I can bet that RVK would probably rate their fans according to the majority of controllers providing an RMS of 105v due to their PWM approach which we know would definately show you conditions you talk about.

ToMMY2k

your transformer will obviously have a higher current rating than your fan. the output impedance of the transformer wont affect the current draw of the fan, it works the same as a standard transformer, the input and outputs just happen to share the same winding. people missunderstand autotransformers and believe them to work by basically introducing an inline inductive impedance to the circuit, but the secondary output is actually induced into the secondary circuit, the transformer isn't a current limiting device because the transformer has a neutral connection on the primary, if you had no permenant neutral to the transformer it would act how you describe with the only current path to neutral through the load. measure the current in the input to the transformer and then in the output, and the output current is higher than input, this can only happen by induction, if it was merely an inline impedance then the output current would be less than the total current, also if it was a series impedance then loads powered by the transformer wouldn't pull their rated current at whatever the voltage you were applying to them and would thus not produce the right amount of power. all of this is irrelevant however because your fan will still pull a current at below 30% and there will still be a reduced airflow and it will overheat.

when the transformer is wound right down, the output leads are both at neutral anyway, so of course the voltage would be zero, you'd get zero current through the output leads, and just the normal no load current through the coil of the transformer. but if you apply only for example 50volts to the fan, and your transformer can supply say 3 amps, the fan will draw whatever amperage the fan would draw if it was connected directly to any other 50volt supply, the impedance of the transformer makes no odds, your airflow will be massively reduced, and your fan will overheat, there's no point in dropping the supply voltage to the fan below 105 volts anyway, as if you look at the airflow/pressure charts for rvks, at 105volts the airflow is something ridiculously low, for example the rvk 125mm L1 at 105v and around 200ps flows only 2l/s yet still pulls 0.153A when at full speed it pulls only 0.1A more, at 50v the rotor will still turn but it will be so slowly it will produce pretty much 0l/s airflow, yet still be pulling roughly rated current.

for fan flow/pressure curves:

http://catalogue.systemair.com/COM/index.aspx

a good article on autotransformers:

http://ecmweb.com/mag/electric_basics_autotransformers/

as you can see in the first few paragraphs, the volt/current relationships between the supply to the transformer and output of the transformer, are the same as with a standard transformer. the impedance of the winding does not effect the load impedance nor the magnitude of the current drawn by the load. if you short circuit the output of an autotransformer, you reduce the input impedance of the transformer, the same as with a standard transformer, the current will be limited only the maximum the transformer can supply at short circuit, as your fan will (hopefully!! ) have a much lower max current than your trannie can supply, you'll burn out the windings


"As shown in the diagram, the line current is 10A for a 2000VA output (10A x 200V). The load current is 20A at 100V or 2000VA output."

Ok this is interesting stuff...but probably a bit too much for this thread

I see what you are saying about the impedance of the transformer but I have a few questions about the variac vs PWM supply that you may well be able to answer.

Im going to write what I see and how I understand this issue...this might be wrong but please feel free to correct me as understanding this is important to me for obvious reasons

Ok here goes...

A PWM supply provides an average voltage over time, but the spikes, peak to peak are probably close to 240v. In terms of an inexpensive controller this is almost certainly true.

If the PWM voltage spikes are like this, then of course while the PWM supply is wound down, lets say to 24v average voltage for arguments sake the fan is still seeing 240v voltage spikes. The fan would obviously run at a specific speed for this average voltage and current draw. At this specific speed the 240v spikes would result in large currents being drawn while the fan is in slip. This is partly what causes the hum in some fans (the other cause is the switching). Any generation of audiable sound in a transformer (or fan coil), because of vibration induced by changing magnetic fields, usually also results in a large amount of heat generation due to the large currents being passed.

So... for a fraction of the fans rotation due to the fan not turning quickly enough (this is slip btw) the fan is offering virtually no resistance to the potential of the supply and would therefore pass those currents generated by the 240v spikes to potentially whatever the maximum rating of the controller is (maybe 2A?). This current draw is not generating movement so as we reduce the average supply voltage we are generate more heat for more time. This is incredibly bad for the fan because of the stress both in terms of flash heat generation and mechanical/magnetic stresses. Our fan coil becomes fuse wire in this sence and the short bursts of high voltage and large current will do the obvious.

Now if we use a variac it seems like a different story.

Firstly if we wind the variac to 24v RMS the spikes peak to peak are 31v. The current that the transformer can provide at this winding position is limited from 3A to something much smaller (Id have to measure to be sure so I wont post a figure).

Using the same process as above, the voltage peaks while the fan is in slip are much less and the maximum current the variac can provide at those voltages is also limited. This would obviously equate to a much lower heat generation factor.

Im completely happy with the 35% figure with a PWM supply causing problems. It makes perfect sence. However in terms of the variac approach Im having a hard time applying my understanding of the process of how heat is generated in the fan to the voltage and current that a variac could supply when wound to less than 35%.

There is also another point to all this. My controller will remember what output it needed, to keep a temperature inside constant for a given outside air temperature. It would be reasonable then to assume that if there was no temperature difference the fan would need to be off anyway. If there was a temperature difference the controller would remember that the fan needed to be on more than 35% to make any temeprature change due to the low air flow rates at voltages less than 35%.

I am really looking forward to your reply Any chance to learn something like this should alyways be used! Thanks for getting involved!

ToMMY2k

PS: Ill do some measurements from my variac now just so I can understand what it is capable of.

Ok I have done some quick measurements.

The variac didnt seem to limit the current as much as I was expecting. It did however limit it to 1.8A at 2.8v and then to well over 9A as I wound the transformer up. Its a 3A transformer!

I didnt try that for too long so I didnt get to record any voltages. I managed to melt my test leads tho

The fan however showed signs of behaving the way I expected at slow speeds. For example at 2.8v the fan only drew 33mA and showed a fairly liner response as I wound the transformer up. At higher speeds the effects of the slip were apparant with increased current draw until the fan reached its maximum speed for a given voltage.

I dont have a PWM controller to compare against and my tests were only quick.

ToMMY2k

cheap fan controllers aren't true PWM devices, they just cut out a portion of the sinewave, which isn't the same as PWM from say an inverter, which pulses DC at specific intervals, you do get buzzing from the fan because of the switching, but it's not the same as the buzzing caused by an ac drive. but yes, it does cause vibration and a bit more heat and mechanical stress that you get from using a variac.

however! this isn't the same as the excess current draw from the rotor being in slip. basically imagine that a stationary fan is a transformer with the rotor being the secondary and a dead short, hence why you get a massive current if you lock the rotor of a motor. this is how power transfer works in a transformer. if the secondary of the transformer is open circuit, then the magnetic field of the secondary induces an opposite voltage into the primary and the two voltages cancel out and no current flows (or very little current, due to losses, but for the sake of conversation this is a perfect transformer) if you short out the secondary, high current is induced in the secondary now it has a complete circuit to flow through and the power is dissipated across the length of the wire, the voltage drops across the secondary circuit and the magnetic field in the secondary is destroyed and there's no voltage reflected back to the primary, and a high current is allowed to flow.

now current can only be induced by a changing magnetic field, if the rotor is stationary to the rotating field of the stator, then the magnetic field is changing on a high magnitude in comparison to the rotor, and as the rotor "winding" (the squirrel cage) is shorted out, if the rotor is stationary to the stator then it's acting like a shorted secondary of a transformer, if the rotor is spinning then the magnetic field isn't changing very much in relation to the stator, so the magnetic field inducted into the rotor tries to induce an equal but opposite voltage as that being fed into the statorand the two cancel eachother out if the rotor spins at exactly the same speed as the rotating field. if however they cancel eachother out, then the rotor would be no longer attracted to the rotating field and the motor would produce no torque, so the rotor always sits just (edit)-behind synchronous speed. there is some slip and so there is some power transfer between rotor and stator.

if you drop the voltage supplying the motor, the magnetic field still rotates at the same speed because you haven't changed frequency, so even though the stator is only being fed by say 30v (assume this is slow enough to stop the rotor from spinning) then you're essentially supplying a transformer 30v and then shorting out the secondary, because the magnetic field induced into the rotor isn't spinning fast enough to reflect the voltage in the primary = high current in both the primary and the rotor "windings" (or the bars of the squirrel cage) this is why ac drives control voltage and frequency, to eliminate the slip, by changing the frequency, you change the speed of the rotating field in the stator, AC drives always drop voltage and frequency in proportion to eachother. if you supplied full voltage and just changed the frequency, then the stator would pull an unneccessarily large magnetizing current, which would i imagine increase the torque produced by the motor, but too high a magnetizing current saturates the iron in the core and causes excessive heating and harmonics etc etc

the current caused by slip from low voltage is not as high as if you supplied 240v to the motor and slowed the rotor to the same speed by force, but because you now have almost 0 airflow, there's nothing cooling the windings anymore. and although having a totally stationary rotor is the absolute extreme, that's just to demonstrate the principle, slowing the rotor of a motor by applying force (i.e. holding the rotor by hand or seized bearings whatever) is essentially the same as loading up a transformer, slowing the speed of the fan by dropping the voltage, is still loading it up, it's just now rather than the load being the fan rotor rotating it's just current circulating in the windings of the rotor. there is less voltage on the input, so it's not as massive a current as loading the fan up by force, but now less of the current in the stator is being turned into rotating energy and more of it is being turned into heat because of the high current induced into the rotor. this is why these fan are made so that the stator is in the centre, and the rotor rotates around the stator, i.e. external rotor motor, it's too allow the airflow cool the high currents induced into the rotor when a motor is run under high slip conditions, if there is low airflow the (totally wasted by the way) current heats up the rotor, which in turn makes the stator windings heat up, and your fan dies

sorry, got a little distracted.

so what kind of current was the fan pulling at say 80v? the fan rotor may still be rotating, but the airflow is almost non existent, i don't really think current at 2.8v is really relevant, as the voltage is so small, only minute currents will flow anyway, but when you get to say 40v+ the current will be getting to more damaging levels if there is no cooling

also to add, the current isn't limited by the variac (only by the resistance of the winding wire which is very small) the variac acts like any other transformer, 3A is the LIMIT that transformer can supply without overheating, not the maximum it can supply, short the output of a variac and a massive current will be induced into the output, same as with any other transformer, it's not the same as just sticking an inductor in series, the top half of the winding acts as a primary and the bottom half as a secondary, and can deliver as much current as the secondary circuit demands. it's hard to explain without a diagram showing the flow of current, but you'll just have to take my word for it

here's a good article on voltage effects on motors, if you look at the graph on the right, you can see "full load" current goes up the further away you get from nominal voltage, overvoltage causes excess inductive current and lower power factor, undervoltage causes higher resistive current and better power factor, but efficiency goes down either way you go. the difference with our application however is we have a speed dependant load, so we see a different behaviour, because obviously by dropping the speed we flow less air and put less of a strain on the motor, so by dropping input voltage we still don't go over full load current as the case with a fixed load, but you're still inducing a high proportion of your input energy into slip current into the rotor rather than rotational energy, (especially if the rotor isn't moving, as 100% of the input energy is being turned into heat and 0% turned into kinetic) and because fans are designed to be run at no less than a minimum airflow/speed, buy running at less than minimum voltage you are running outside of design parameters and will (depending on environment of course, if it's 0'c it'll probably be ok!!) experience overheating, especially on less well designed/cheaper fans

anywho, i'll shut up now, i should put the spliff down and go out!! :


http://www.motorsanddrives.com/cowern/motorterms12.html