Pure sin Vs Modified Sin

[NiHaoMike]

Why they don't build inverters like class D amplifiers with a triangle wave to drive the pwm and LC to smooth it to a real sine wave?

Delta Sigma is superior to traditional PWM when it comes to output quality at a given switching frequency.

doesn't work for a push-pull circuit, which is how the majority of 60hz iron transformer based msw inverters operate.

The majority of the inductive kickback gets clamped by the other side and the little remainder (due to leakage inductance) is taken care of by the snubber. In such a design, one trick I like to use is to use a fan as the load in the (diode RC clamp) snubber so the energy dissipated is used to cool the circuit instead of being wasted. If isolation is not needed and the output is DC, it's even possible to use the primary side of the push pull as a forward converter that doubles the input voltage, with the secondary stage connected in series.

[TarHeelTom]

There is a third option....DC.  I have a 12V camp and for a while I was buying 2000/4000W broken inverters to run the fridge etc.  These were bundled in sales with smaller inverters.  Usual problem was blown FET in H bridge.  The smaller ones I just pulled out the H bridge FETs and ran pure 140V DC out to the outlet.  Chargers, electronic CFL lamps, CD/DVD players etc work just fine on DC.  Anything with a switching power supply.  A little more efficient and less noise.

Built a camper in a regular sized passenger van (but without most of the seats) back before everyone and their brother were producing these.  Ran across a small refridgerator which ran on either 12 volt DC or 120 AC.  Had a battery charger under it and two extra 12 volts batteries which hooked to the alternator anytime the engine was running.  Was young and stupid and hooked the two batteries in parallel.  Next time it'll be two huge 6 volt batteries in series.

The battery charger failed one day and took out the transistors in the fridge.

In troubleshooting it, we discovered that the fridge ran on 40 volts AC.  There was a transformer which stepped down the 120 volts to 40 volts, or stepped up the inverted 12 volts to 40 volts.

Drove it with the fridge cold for about 6 or 7 years, then sold it.

Tom

[hamdi.tn]

Some people here suggest to use a High voltage DC in case all equipment is powered by an SMPS, am not sure but since this boost is for domestic use , is there any kind of regulation and special protection for equipment with high voltage DC output ?

[NiHaoMike]

Some new data centers are using HVDC power distribution. Main problem with HVDC is arcing. A good compromise between AC and DC is a trapezoid wave, which solves the arcing issue, avoids the EMI issues with modified sine, yet is about as efficient as a modified sine inverter.

[Jeroen3]

Newer smps supplies tend to only draw mains current on the top and bottom of the sine.
Those mains LED lights (with an ic for driving it) do this the worst way, creating noises on the line most meters cannot work with.

I can imagine trouble when you apply square waves, that is actually a high frequency square if poorly filtered.
If you run it through a transformer, it becomes much better. But most smps do not use an 50/60 Hz transformer anymore.

Nice images: http://www.vfds.com/blog/what-is-a-vfd

[nitro2k01]

Some new data centers are using HVDC power distribution. Main problem with HVDC is arcing.

What does HVDC mean in this context? The common 48 VDC, or something higher?

[DanielS]

Newer smps supplies tend to only draw mains current on the top and bottom of the sine.

Newer? What you are describing here is what a bridge that dumps energy directly in the bulk decoupling cap would do and that topology has been around for 50+ years.

Regulations in many countries are starting to require power factor correction. For PC power supplies, 80+ certification requires a power factor of 0.9 or better, which is only achievable with active PFC. What APFC does is boost the line voltage so the current waveform follows the voltage waveform, which means current is being drawn through most of the cycle instead of only peaks ("top and bottom") where line voltage is higher than the bulk capacitor voltage.

[NiHaoMike]

What does HVDC mean in this context? The common 48 VDC, or something higher?

About 380V DC.

[David Hess]

I have always wondered why modified sine is even called modified sine, it is so much modified that modified square would be more appropriate term

I assume because calling it modified sine indicates that the RMS and peak values are the same as a true sine wave source which is the whole reason to use a modified sine anyway.

Transformers and motors need to be derated and will run hotter on a modified sine source.  Passive lamp ballasts may fail completely.

[Jeroen3]

Maybe you can add "stepped approximation sine" to the title, since the ups I recently bought uses that.
Although it's probably just a filtered modified sine.

[hamdi.tn]

well i finally chosen to use boost due to an update that even motor are powered through an SMPS and not directly from the main so no need for a true sine and the additional cost of a 50Hz heavy transformer , and am selecting a stupid push pull transformer , and now am incapable of thinking , i calculated that i need a 36 turn ratio transformer , so if i have 2 turn on both Low voltage output i should have around 72 turn on the high voltage one ? am totally incapable of finding a transformer with that turn ratio in any catalogue of any manufacture " i know", and what the hell this notation " turn ratio 35:2:2:3 " mean  i guess it's a turn number so the Nps turn ratio is in fact 17 or am totally wrong . SOS   

[DanielS]

I assume because calling it modified sine indicates that the RMS and peak values are the same as a true sine wave source which is the whole reason to use a modified sine anyway.

There are several variants of "modified sine" output stages out there. The main reason for it that I remember hearing was to reduce the amount of harmonic content and give diode bridges in SMPS time to recover. You can match the RMS value with a square inverter by simply making its output ±120V for 120V_RMS.

I have an APC BX1000 and its output stage uses an H-bridge with a chunky output inductor (around 50 turns of #14 wire on a 4-5cm long ferrite or powdered iron bar) combined with an additional switched inductor to snub transitions. I do not have an oscilloscope at home (yet) to take a look at what the waveforms are like but the inductors should soften the waveforms by a fair amount when under load.

[hamdi.tn]

So guys any help with the transformer