|
Posted by Don Klipstein on May 3, 2008, 2:13 am
Nebenzahl wrote:
>On 5/1/2008 6:29 PM Don Klipstein spake thus:
>
>> I've blown a modified sine inverter in the past with current spikes
>> drawn by capacitors.
>>
>> There are LC filters, but those tend to be homebrew and experimental and
>> things can go wrong. For example, you have to really know what you're
>> doing to not have them resonate (whether at the fundamental frequency or a
>> harmonic), or get close enough to resonating to deliver excessive voltage.
>> I'm sure one not professionally made and rated for use with inverters
>> would void warranties and safety ratings all over the place. And I don't
>> know where to get any that are rated for this.
>
>So are you saying that an LC filter (say, a pi configuration, two caps
>and an inductor) *could* work, if one were to select the correct values
>of L and C? I understand how one would want to avoid resonance (@60 &
>120 Hz) here. Should be simple to do, one would think; not exactly
>rocket science.
A 2-C, 1-L pi has capacitive input - bad for the inverter.
Better would be just one L and one C. Watch for output voltage varying
with load even far from resonance. With no load or very light load,
output voltage exceeds input voltage for every frequency below resonance,
though the gain is minor to negligible for frequencies far below
resonance.
Squarewaves have all of the odd harmonics. Modified sine waves
(squarewave modified to 1/8 cycle at zero, 3/8 cycle positive, 1/8 cycle
zero, 3/8 cycle negative) also have all of them, though notably somewhat
less of the 3rd and 5th harmonics than with a squarewave.
You can make the behavior more predictable and damp down resonances by
adding a resistive load to the output. However, the resistor load will
waste power. You also have to prepare for the load or connections to it
failing, or be able to tolerate such failure.
Making an inductor that will handle your load current without saturating
requires knowing how to design gapped iron core inductors.
This is usually done with a laminated E-I core, and will probably be at
least a little heavy and bulky.
Otherwise, hope you can find a "reactor" type lamp ballast with the
inductance and current capability that you need. (Probably heavy and a
little bulky)
Air core ones will be huge and probably scramble the color of operating
color CRTs a few feet away unless shielded. Ferrite core ones will
saturate more easily than iron core ones.
If a setup is going to be something for everyday use in your home or
frequent or regular use in your camper, or if it has to work at times
when technically qualified personnel are not around and awake and with
a fire extinguisher handy, then I recommend something
"more-tried-and-true" than something that I consider "experimental" or
"science project quality" or even "senior design project quality".
Electrical equipment for everyday use should have such reliability and
"fale-safeness" (adequately unable to fail or adequately unable to have
failures start a fire) to pass the sort of testing that UL does upon
electrical equipment of kinds that normally get (or should) have
listing/approval of UL or similar safety testing laboratory organization.
- Don Klipstein (don@misty.com)
| 
Yahoo! 
Google 
Windows Live 
del.icio.us 
digg 
Netscape 
XML Sitemap