 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : Windmill Controller/Charger
  Page 2 of 2 |
|||||
| Author | Message | ||||
Gizmo Admin Group  Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
Hi Chris The controller just switches a couple of mosfets on or off. But it could use PMW. The X chips ( 18X, 28X ) have a PWMOUT command, and its multitasked, meaning it runs in the background. So you set a PMW frequence/duty cycle, and then step through other code untill you need to change the PWM settings. I think you need to set the PWM settings to zero to stop it, dont have the manual on me. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| RossW Guru  Joined: 25/02/2006 Location: AustraliaPosts: 495 |
Hi folks. Just come across this site and been having a look around and reading through the forums. This message is *mainly* for Gizmo... I like your controller - but I'm looking for something different, or ideas, (valid) reasons, arguments etc against what I want to do! I've just put up a 1000W/1500W mill here in Southern NSW, Australia. As in keeping with the "rule of thumb" about days of nil wind after erecting a mill, it was blowing a gale the morning I put it up, but by the time we had it up the wind had gone and hasn't been above a gentle puff since! I've long thought about how I want to do this, and some of the problems I perceive with renewable energy sources (PV, Wind particularly) in that often there IS power there, but just not enough to operate the (mostly very simple) charging systems available. Sure, MPPT overcomes this, but we're in a problem situation then with having to have massive controllers for larger installations, and losing significant amounts of power (even if a converter is 90% efficient, it's wasting 100W in a 1KW system and more as power goes up!) A wee bit of background. My home is underground. Built in the side of a hill, covered with a minimum of 1m of dirt (and about 4m at the back). Hydronic heating coils in the slab (will come back to that if anyone is interested). We have a lot of computer equipment and sophisticated home automation (C-Bus), and our current energy demand is about 18 KWH/Day. I have a 5KW continuous (10KW peak) inverter and moderate 48V battery bank. PV contributes some energy to the system but the bulk of it comes form a 4-cylinder automotive engine (converter to LPG) coupled to a 14KVA/3ph machine, which operates with the inverter/charger to power larger loads (washing machine, dishwasher, welder etc) and charge batteries. The engine also provides heat (hot water from the cooling jacket, and a heat exchanger on the exhaust) for domestic hot water and hydronics etc. My thoughts are (and I may be shot down over this) - we see a lot of activity from the wind turbine where it hasn't made the 50+ volts required to charge the batteries. Even if it's only 30, 40, 50 volts, and perhaps only 10 watts at the low end, perhaps up to 50 watts of "potential, available" power - If it's there for 10 hours, it makes a non-trivial contribution. So how to "use" this otherwise waste power without wasting more in a lossy conversion at full power? How about a hybrid system? The converter need only up the voltage while it's below what is required to convince power to flow into the batteries. My windgen is 3-phase AC, so a 3-ph bridge. Your controller could have one additional output - one for "load", one for "charge" and one for "inverter". When you sense the machines output is too low to charge the battery directly, turn OFF that FET, but turn ON the one which runs the DC-DC converter. This now only needs to be rated to (say) 50 watts or so (the otherwise wasted part). Once the turbine is generating enough to charge batteries directly, do so and save on the conversion (in)efficiency! On a different matter, I like your home-made current probe, but am curious - given your claimed inaccuracy (an amp or two at 10A), wouldn't it be far simpler/easier to use the op-amp to buffer (and increase) the voltage drop across the battery negative, and feed that into the A/D? A simple gain adjustment to calibrate for whatever (arbitary) cable resistance there is, and use the cable itself as the sense resistor? Cheers, RossW |
||||
| Gizmo Admin Group Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
Hi RossW All good points. I would like to revisit the controller in the near future, and maybe consider some sort of intelligent charging. One of the problems with the current design is most of the pins, code and variables are used just to drive the LCD display. This leaves little room for anything else. I would like to have two chips, one to drive the display, and one to control the charger. This will give a charger with much more processing power, and more output pins as well. Yeah I'm not happy with the hall sensor. A couple of op amps across a shunt resistor would be better. The cirsuit would need to measure plus and minus amps, as I intended it to display current drawn as well as charging current. I was working on this circuit...
The first op-amp ampifies the +/- voltage from the shunt to a range -2 to +2 volts, the 2nd op amp moves this voltage up to the 0 to 4 volt range, usable on the PICAXE. So 2 volts = 0 amps. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
Bryan1 Guru Joined: 22/02/2006 Location: AustraliaPosts: 1344 |
Hiya Glenn, Eh mate I've noticed your using the picaxe for your charger project. The picaxe is a great chip for small projects but as you've said the code space just isn't there on one chip. But if you used a pic chip then the space would be there as the picaxe interpretor takes up a heap of space on the chip which is a pic chip to start with. There is a basic compiler out there called oshonsoft and it's pretty cheap to register so the transition shouldn't be hard and using a 16f877a would not only give you 40 pins but also 8K of code space etc. That above chip can use a bootloader so all you need is a serial port connection and simple terminal program to talk to and program the chip. If you like I can makeup a simple board with the rs232 output and the pic all mounted, tested and ready to play with. As I have all the gear here it will only cost me a bit of time and the postage and I'll be glad to do it for you as a token of gratitude for providing such a great forum. Cheers Bryan |
||||
| RossW Guru Joined: 25/02/2006 Location: AustraliaPosts: 495 |
I was playing about a little this afternoon (when I *FINALLY* got
some more breeze!). For nearly 3 hours we had enough wind to turn the
turbine at a moderate speed but it was only making 30 to 50 volts (I
did say this is in a 48V system, didn't I?). Anyhow, I ran the output
of the 3-phase-full-wave-bridge to a pump I had lying about, and it was
moving a hell of a lot of water about (out of a 30 litre bucket, up a
pipe and back in, just circulating) - and making at least 20 litres/min
if not more, and at a reasonable pressure (it was a positive
displacement pressure pump - the standby pump for the household water).
All this while the commercial converter wasn't making ANY contribution
to the batteries. Hell, even if I only got 10W out of it, it's sure
better than nothing! I've been looking at the UC3845B (about $4), which
lends itself (I think) perfectly to the application. Pull the inverter
transformer from an old PC power supply, re-wind it with a few more
turns on the secondary (and use something heavier than 30AWG) and
there's one mighty fine, simple, current-limited high-efficiency dc-dc
converter.
Have you considered using one of the ATMEL chips? I built my generator-controller which runs a 20 char x 4 line display, RS232 comms to the rest of the house, monitors battery voltage, oil pressure, water temperature in, water temperature out, heat exchanger temperature, O2 level and a couple of other things I can't remember offhand, has outputs for ignition, electronic governor, starter (crank), water pump, AC output to house, generator available state to the inverter and loud warning alarm, inputs from inverter, mode switch and a couple of other interlocks, and opto-isolated inputs from the "points" (electronic ignition) and zero-crossing from the alternator output so it can display in real time the engine RPM and output frequency in Hz, and STILL has some bits spare. I used the 90S8535, but if I was doing it today I'd use one of the MEGA chips with heaps more FLASH. The 90S8535 were costing me about $10 which wasn't bad for an 8MIPS machine with 8 x 10-bit A/D converters and 24 IO lines, 8K FLASH, plenty of EEPROM to store all the config and run-time stuff (engine hours, starts, all the time constants etc, warm up, cool-down, throttle-up etc delays).
What's with the love-afair with 741s? Just thinking out loud here, looking for an easier way to do it. The rest of your circuit doesn't use a -Ve rail anywhere from memory, be a shame to have to make one just for this. How about, if you used a high-impeadance divider from a known reference but not to *GROUND*, but to the other side of the shunt. RossW |
||||
| Gizmo Admin Group Joined: 05/06/2004 Location: AustraliaPosts: 5078 |
Hi Ross I havn't considered a ATMEL chip, or even a blank pic as per Bryan's message. Most of the little projects on this web site were built keeping in mind that other people may want to copy the idea and build their own. So I tend to go for common components, steel sizes, bearings, etc. Thats why I went for the PICAXE, easy to come by, cheap and the simple BASIC language is easy to learn. But I would like to try a PIC or ATMEL one day. I would like to start againg from scratch, and this time build it into a bigger box, or even a rack mount unit. I'm keen to try some intelligent charging and have more switched outputs. Even a Mains powered battery charger that can be switched in when the batt voltage drops to a dangerous level. I have a drawer full of 741's so I tend to use them for everything, and the circuit simulation software I use has them listed. But I'm open to suggestions, I would like to get rid of that -9 volts. Glenn The best time to plant a tree was twenty years ago, the second best time is right now. JAQ |
||||
| RossW Guru Joined: 25/02/2006 Location: AustraliaPosts: 495 |
There are free programmers for the ATMEL, there are things like BASCOM for those who want to use BASIC (or a basic-like environment anyway). Or of course, you could distribute a hex file people could just write in. The "interface" is just a D25 connector from the parallel printer port and 3 bits of wire to the chip :) My generator controller has had a few upgrades to the code as I've thought of other things I wanted it to do (for example, beep the warning alarm a few times when the temperature gets high, but before it gets to the automatic shutdown level) - so I left a small hole in the case just big enough for the connector to slip in. I'll e-mail you seperately a userid/login to have a look at the beasty :)
Alright for those who have mains power :) Well, that's a valid reason I suppose. Still, I reckon the convenience of having them is outweighed by
their need for bipolar supplies, noise, compensation problems etc. Still, thinking about it a little more... you can probably use a 741 in the sensing circuit with a little creative design work. Again, without thinking too hard.... Using 0V/Ground as your reference and +V to power your opamp and
other devices, as long as the input from the shunt doesn't get too
close to the 0V rail you're ok. Imagine, if you will, the sense input
being part of a voltage divider. Lets just say it's a 10K from sense to
opamp input, and another 10K from there to +V. Lets assume +V is 10V
for now because it makes it easy. If you A/D is 10 bits from 0 to 4.096V thats what, 4mV/bit. RossW |
||||
| dwyer Guru Joined: 19/09/2005 Location: AustraliaPosts: 574 |
|
||||
| dwyer Guru Joined: 19/09/2005 Location: AustraliaPosts: 574 |
Hi Ross sometime ago l saw your idea about your 4-cylinder automotive engine coupled to a 14KVA/3ph machine which l have done same as you when l brought 10KVA/3ph last year and just finish yesterday and had 4 cylinder 1200cc corolla motor run 3000 rpm however when the load from my lathe corolla motor die due lack of torgue so l do not what to do as do you have any problem with your 4 cylinder motor ? lathe motor is 6kva Can you able give me any idea thank Dwyer |
||||
| RossW Guru Joined: 25/02/2006 Location: AustraliaPosts: 495 |
First off.... the RAPS5 inverter/charger I use is the "primary" power source for the whole house. It can supply 5KVA continuously (obviously, subject to battery capacity!), 6KVA for 30 minutes and 10KVA for a few seconds. That said however, when the genset is running, the inverter output and the generator output are directly parallel connected, and the PAIR of them can now supply the combined output. So with say 10KVA from the genset, and fully charged batteries, I could run loads up to 15KVA, or 20KVA peak. I've had no trouble at all using large welders or a MIG off this system. As to your problem - a 1200cc corolla engine should be able to produce that sort of power easily I'd have thought. What are you using for a governor? Is it possible to use a star/delta start for your lathe? (Ie, start the thing moving in star where it'll draw less current) and then switch to delta when it's up to speed? Or, using a VFD/Soft-starter? Is your generator 2 pole or 4 pole? I guess it's a 2 pole if you're running 3000 RPM, unless you've got a reduction in there somewhere. I found that the 4-pole machines are much heavier (mine is about 100Kg just for the machine, no pulleys or anything), and seem to have a much better "transient" capacity. Also, that I can run the engine slightly faster (1900 RPM) and get better speed regulation as a result. Have you tried spinning the lathe up before bringing it on-line? Or is that not practical? (6KVA motor is quite a beast, I guess you can't turn the bugger by hand hey?!) RossW |
||||
| dwyer Guru Joined: 19/09/2005 Location: AustraliaPosts: 574 |
Hi Ross Well last night l had the engine running and notice that 2 cylinder out of 4 arent firing at 3000rpm and l just brought compresser tester to-day and will look into it My Aternator is meccalte 10Kva 2 pole 3000 rpm 50 hz and my industrail lathe only take 415 3ph because tranformer 2 phase and cooling pump take 3ph all rely on 415 volts Aternator is driven straight out of flywheel with modify cutting welding gearbox bellhousing so aternator bolt on to bell housing and l will send picture later on and l dont have governor on l relied on voltmeter Anyway where l able to buy governor as last time l brought the gen from the dearer and did mention about the governor they dearer dont know where l get one Regard Dwyer |
||||
| RossW Guru Joined: 25/02/2006 Location: AustraliaPosts: 495 |
Erghh! Yeah, that won't help much! *nod* I guessed 2 pole because of the 3000 RPM. I chose a Sincro 4 pole with electronic AVR. As long as I get the frequency right, it'll take care of the voltage itself. Can be connected in star, delta, double-delta, and a bunch of others I've never even seen! I have mine in double-delta to give me single-phase output. Driving it straight off the crankshaft certainly reduces losses. I used a Woodward APECS-500 governor, details here: http://www.woodward.com/pdf/ic/36710.pdf A small inductive pickup sits about 0.02" from the flywheel (ring gear) and counts teeth as they go past - so it gets 112 pulses per revolution - this lets it monitor the speed and respond very quickly. The engine I used was from an old ford laser (1800cc, made by mazda). Originally a fuel-injected engine, I pulled off all the injection stuff and have converted it to run on Propane (LPG). The linear actuator that the governor controls directly drives (opens and closes) the butterfly behind the gas mixer, so as the engine RPM tries to drop, the actuator opens the throttle and vice versa. I also have an oxygen sensor in the exhaust manifold which feeds back to a Lambda controller to open and close a stepper in the gas line between the regulator and the mixer, so regardless of load, the fuel mixture is kept at optimum. Can post you photos if you want 'em. Regards, RossW |
||||
| dwyer Guru Joined: 19/09/2005 Location: AustraliaPosts: 574 |
Hi Ross This late afternoon l put the compressor tester and found two cylinder are weak and sqiud some oil into the spark plug hole and compressor went up hell alot so l took the head off check valves seat seem to be ok l guess rings are worn out so l might get motor rebuilt thank for letting me know about the governer and will look into it There is the photo and this is first time l have done on the internet
|
||||
| Page 2 of 2 |
|||||
 Print this page |
