Hi, EE's I could use some ideas on a domestic project that has hit a glitch - one that, as an EE, I should have seen coming!
We have installed a wood burning stove insert in our family room to provide a third fuel source in the event of an electrical power failure (we have electricity, a gas furnace (needs electric power, of course), and now wood.) The wood stove insert has a 50 watt turbo-fan to push air around the firebox - this is needed to get the up to
35,000 BTU/hour heat out (it is not a free-standing stove, it is in the old fireplace.)To run this fan when the power is out , I acquired a 12 VDC deep cycle marine battery and a 300 watt ,12 VDC to 120 VAC converter. Well, it runs after a fashion but with a nasty buzzing noise. Reason: the motor (a shaded pole AC induction type) does not like the massive harmonics in the convertor's AC output.
Waveform and harmonics: I put a 'scope on the AC waveform. It is virtually a 60 Hz square wave... but not quite. While a true square wave goes from + to - instantly, this waveform stops at zero and holds for about 4 mS, it also holds + and - for about 4 mS (all aproximate times.) The promotion for this unit calls it a "modified sine wave" - I'd call it a "brutalised sine wave"! It's OK for resistance heating, tungsten lamps and 120 VAC power supplies with a FW or HW rectifier, but not for AC motors. Of course, this waveform allows the converter output devices to be either fully on or fully off for negligable internal power loss using only two high current DC voltage rails (+ and -). My understanding is that a better, stepwise (but lossless) synthesis of a sinewave needs multiple switched high current DC lines - much more expensive than this little converter.
I presume the harmonics are all odd since a true square wave is sine (wt) + (1/3)sine(3wt) + (1/5)sine(5wt) ... etc. I've not done a Fourrier analysis on this "brutalised sine wave" but it may be all odd harminics, too (anyone done this?) Anyway, I tried using a 780 mH series choke to reduce the 180 Hz, 300 Hz, ... etc. No go. With a resistive load (75 watt lamp, not the motor) this choke dropped the voltage to about 40 VAC rms with a nearly triangular waveform across the lamp. Tried a smaller, 21 mH choke and it took out the higher harmonics but left a very significant 180 Hz component such that you could hardly recognise the sinewave (it was in there somewhere!) So, series chokes don't work.
Next I tried a 5 MFD AC capacitor across the lamp with the series 21 mH choke in. Result: 120 VAC RMS across the load but huge increase in AC current, presumably mostly leading VARS - but not measured. I did not measure the DC current to check this. I could do more measurements but I think the problem is clear.
Musings...
- Is there any work around that still uses the 12 VDC/120VAC converter?
- A large part of the AC energy is in the higher harmonics. Attentuating them on the AC side, even if possible, would seem to cut efficiency.
- I can't replace the fan motor with a 12 VDC type - it's mechanically infeasible (but it's the best solution as a 12 VDC power supply for regular 120 VAC use would be very easy.)
- Any scheme that delivers a true sine wave at 120 VAC is very lossy (low efficiency), like a class A amplifier.
- Running a simple rotary fan (from a microwave oven) directly from the crude AC is actually much quieter, no buzz. Perhaps the stove fan has more vibrating metal parts.
- The stove fan does not need to run at full speed to be effective. Indeed, it's OK down to 70 VAC from a variac, or with the rheostat speed control quite a way down to, say, 1/4 speed.
- I've run out of ideas... does anyone have any more?
Here's the fan voltage/current curve on pure AC: Fan speed control rheostat on full. Supply by variac (pure sine wave, not converter)
VAC Current (AC amps)
60 0.2 (this is minimum for self start) 70 0.25 80 0.3 90 0.34 100 0.38 110 0.43 120 0.46 125 0.48 No PF info - assume ~0.9) 70 VAC seems to be the practical minimum for lowest speed operation.Feel free to email me directly at "analogdino at(taboy) rogers dot(ty as they) com" (decoded) Thanks for all replies. Cheers, Roger