Re: Power Inverters

M.I.5¾ wrote:
"spamme9" <spamme9@xxxxxxxxx> wrote in message news:VkSWk.803$us6.591@xxxxxxxxxxxxxxxxxxxxxxx
M.I.5¾ wrote:
"spamme9" <spamme9@xxxxxxxxx> wrote in message news:EYHWk.690$us6.331@xxxxxxxxxxxxxxxxxxxxxxx
M.I.5¾ wrote:
"spamme9" <spamme9@xxxxxxxxx> wrote in message news:1ZuWk.503$us6.7@xxxxxxxxxxxxxxxxxxxxxxx
M.I.5¾ wrote:
"spamme9" <spamme9@xxxxxxxxx> wrote in message news:1CLVk.154$QX3.23@xxxxxxxxxxxxxxxxxxxxxxx
nospam@xxxxxxxxxxxxxx wrote:
We're looking for a power inverter to recharge our Compaq notebook
in the car. Is it better to buy one with a larger rating (400W) rather than
a smaller one? Black & Decker has a 100W model ($13) and it gets
good reviews:

http://tinyurl.com/5fstl9

But the notebook uses something like 90W, so we'll be very close to
this inverter's capacity. The alternative is something larger, e.g.

http://tinyurl.com/6nfbrp

This one's 400W, but we're wondering if the extra cost is worth it.

Thanks for any info.




The 100W one isn't big enough.

Here's the ugliness of all this.

Any inverter you can afford will be a Chinese knock-off
of a knock-off of another knock-off of a maybe-good original
design.
The "designer" copied the circuit (not to be confused with the design)
and used the cheapest available
parts and the cheapest labor. They had not a clue to the original
design tradeoffs. They probably spent more money on the design of
the attractive display box than the design of the inverter.

It's tested to run a light-bulb in very infrequent use for as long
as the warranty...assuming you can find the vendor.

But your laptop is NOT a light bulb. It's a diode and a capacitor.
When the fast rise of the modified sinewave slams into the cap, you
get MUCH higher peak currents than you would with a light bulb.
Just twice the current for half the time is still twice the heat in
the output bridge of the inverter. And that assumes the transistors
can reliably take twice or 4-times or more the peak current.

When using a square wave or modified sinewave (not so square wave)
converter to drive a switching power supply, you need much more
"rated watts" than you'd think.

Typing this got me thinking...
Newer power supplies have built-in power factor correction to make
life easier for the power companies. What happens when you try to
run one of those from a square wave???

First, you have assumed a light bulb has no inrush current. Wrong - it most certainly does.
I love it when people tell me what I'm assuming...then tell me I'm WRONG
for assuming it.

FWIW, the 100W light bulb I'm holding in my hand is 10.4 ohms.
Peak inrush current
could be considerably less than slamming into a capacitor.
But the issue pertinent to this thread is that it only happens
for a few cycles at turn-on
and doesn't contribute to long-term stress and heat in the output bridge
of the inverter.

The 100watt light bulb that I'm holding is 6.4 ohms (different type of bulb probably). Thats 37.5 Amps of inrush current (at 240 volts) - fairly close to the 40 Amps quoted for most laptop power supplies. I don't understand the point in your last sentence because exactly the same conditions apply to the power supply as to the light bulb.

Second, if the inverter can't supply the inrush current demanded by the power supply, then it just provides what it can until the reservoir capacitor is adequately charged. It is seldom a problem.
That's a good point. If you plug in the light before turning on the
inverter, the inverter's primary current limit should take care of it.
But if you start the inverter then turn on the light, the energy
stored in the output cap at 160V can supply more peak inrush current.
Assuming an AC output, what output capacitor would that be? AFAIK there are no capacitors that store AC.


I find your posts extremely entertaining. You speak with
such authority.
I think 3/4 may be an over-estimation.
I laugh my ass off.
Too bad that people who come here for information can't tell
good info from bad. I don't normally get involved unless there's
a safety issue that could hurt the person being misinformed.
I was bored...my bad.
I'm so caught up in this comedy that I can't put it down.

Please your self sunshine.

If you wanna discuss inrush current, start your own thread.
It's of minor importance in a discussion of, "how many watts should
my inverter be to run my laptop."

Why? I can post what I like where I like just as everone else does. If you want to play newsnet nazi then do it elsewhere.

Everybody who knows how the typical modern radio-shack/chinese 12vdc>>120vac
inverter works please click your next button now.

It's just you and me now...
Pay attention.

I doubt that many people know how an inverter works old or modern.

If you can't lift your inverter comfortably with one hand, it may
be an REALLY OLD design that produces REAL square waves and has a HUGE
AC transformer that runs at 50/60 Hz. and is connected directly to
the output socket. Different discussion entirely.

The typical modern, cheapo inverter works by using a high frequency 12vdc to 160vdc switching supply. I'm in the USA. If you live where the standard
is not 120V 60 Hz, scale appropriately.

If the inverter was to produce a DC output, then this would be a feasible way to construct such a device (and as I have pointed out elswhere, there are several advantages to doing so). But for an inverter to power a standard computer power supply using an AC output, the output will be 50 or 60 Hz depending on which side of the great divide you live on. That invariably requires a transformer to step up the switched 12 volt input that it gets. Although switch mode supplies will work off higher frequencies than 60 Hz (we regularly use them from aircraft supplies (400 Hz), most usr would be reluctant to use them at anything other that the frequencies stated on the tally plate (50-60 Hz).

The capacitor has the 160VDC source voltage for the output.
After the cap is a full bridge switch that alternates polarity of
the DC voltage to the output socket. The waveform is a symmetrical
pulse, positive off negative off positive.....
the amplitude of the pulse train is such that the peak voltage of the
output is approximately the peak voltage of a sinewave.
The pulse width is adjusted so that the RMS value of the output
is approximately the RMS value of a sinewave.

Now it is you that is providing the amusement. There would be no advantage to building an AC output inverter that way. Examples abound here in both straight inverter and in UPS form (essentialy the same thing). All use an output transformer, though granted the modified square wave output requires a smaller transformer than a true sine wave would. I have never encountered an inverter that employs a variable pulse width to adjust the RMS output. Indeed the modified square wave (often wrongly called modified sine wave) is defined as the sum of two true square waves with one shifted 90 degrees with respect to the other.

The peak voltage is correct, so rectifier/diodes in the load produce the
proper rectified DC voltage. The RMS value is correct so that light bulbs have the proper heat/light values.
The problem happens when your load wants more current for less
than the full on-time of the pulse. Peak current and the width
of the current pulse are inversely proportional. But the I^2*R
losses in the FETs have a quadratic relationship.
Twice the peak current for half the time is the same power,
but 4X the heat for half the time...net effect is the FETs
dissipate TWICE the power. Not a problem if the FETs and heat
sink are designed
to do that. Whatchawannabet they're not.
Actual results will depend on the design of both the source
inverter and the load laptop supply. If there's enough
resistance/inductance in the path, it's not a problem...IF...

The types of FETs encountered in these devices are MOSFETs whose forward resistance when turned on is measured in milliohms. This is not going to contribute to any extent to the heat dissipation of these devices. Where the heat is dissipated is while the FETs are in the proces of turning on and turning off. They have to got from high resistance to low resistance passing through all the resistances in between. It is during this time that the heat dissipation is at its highest. The use of a modified sine wave output removes the problem of shoot through currents which would exacerbate the problem no end.

My advice is that the OP buy more capacity than the load number
printed on the laptop supply label. And I've shown why.

No one has disagreed with this point.
So, back to your comment...
If your inverter is running and you short the output, the energy of the
160VDC
on the cap dumps through the FET bridge.

For an AC output inverter thus would be the 160vDC capacitor that doesn't exist except in your imagination. I rather fancy, that you have assumed that inverters are built the same way as switch mode power supplies. Although DC output types may be, 50 and 60 Hz AC output types most certainly are not. Just opened one up that turned up that I aquired a coupe,k of months ago. What do you know, the AC output comes direct from a moderate sized transformer with moderate filtering. The input goes via two (presumably) MOSFETs that go pretty well straight to the 12 volt DC input (though there is some filtering on the route). There does appear to be a feed back winding, presumably providing some form of regulation. Oh, and it's made in China and cost a tenner.

If you short the output, then start the inverter, the INPUT side of the
dc/dc converter limits and there's no energy in the cap to dump.

This is the DC/DC converter that doesn't exist exept in your imagination.
Thanks for telling me what I'm assuming.

Congratulations, you found an old/obsolete style inverter.
There are lots of those around...I have several in my junk heap.

If you want to hear it from someone else, you have but to use
your search engine.
http://www.smps.us/power-inverter.html
quoting:
Most commercial DC-AC inverters circuits use the same basic concept: a low dc voltage from the input source is first stepped-up to a higher-voltage dc link corresponding to the peak value of the desired ac voltage. A second power stage then generates an ac voltage by using full-bridge or half bridge configuration
end quote

and
http://en.wikipedia.org/wiki/Inverter_(electrical)
which shows a bridged output in the "advanced designs" section.

The first site won't be useful to you because it's a commercial site.
Wikipedia, however, will let you edit the page. You can "correct"
the content. I'm sure the original authors would like to hear that
the design concept exists only in MY head. I'll check the page later
after your "corrections" are entered.

Incompetence is pervasive, but with significant effort, can be mitigated.
Unwillingness to learn when bludgeoned with knowledge
results in much entertainment for me.

I have a very new inveter.

So you quote a website that is specifically referring to what is obviously large industrial UPSs and assume the same applies to small domestic inverters.

I'm not assuming anything. I am holding one in my hand. I know exactly
how it works 'cause I repaired it.

You just can't bring yourself to admit there's something you don't know
everything about. Not a problem for me. But I do worry about all the people
who might read the newsgroup archive and take your advice.

The larger type of industrial UPS do indeed operate on the
principle you state because there are considerable advantages in doing so. Reducing the amount of ironmongery being a pretty good incentive This entire building that I am in is run from such an example (except lighting). Since it is 3 phase it has to have a sinusoidal output to avoid large triplun currents and this can be achieved in the manner described. Indeed many aircraft power supplies operate this way but only because they have to produce a relatively pure sine wave 400 Hz output from an engine mounted generator whose output is frequency wild.

But for a small single phase inverter of a a few hundred watts to even a kilowatt or so, there is no advantage and a consideable disadvantage in doing so. One of the main advantages of using a design that has a final 50 or 60 Hz step up transformer is that it is possible to exploit a convenient magnetic phonomenon in the transformer that regulates the output. This can only be done with a pulse or square wave output because it requires the core to saturate. The principle first appeared on the early Sorensen AC stabiliser units but these were not popular with users because the magnetostriction of the core generated a lot of noise.

I don't know whether to laugh or be very sad.
Much like your transformer, I'm saturated.
I'm gonna give up.
Have a nice day.
.

Loading