Modern quality. It is HF (according to Victron).
This LF fixation is like audiophiles raving about tubes. Meaningless babble.
Inverters High or Low Frequency ?
- Thread starter Noel
- Start date
This LF fixation is like audiophiles raving about tubes. Meaningless babble.
Maybe I'm not understanding your point, but the Victron ones at least do seem to take a high current 20 kHz PWM-switched 8-10V and transform it to the needed line voltage whether 120V or 230V.
Take a look at this Victron video starting at about 5:50. He draws a schematic showing a FET bridge driven by a PWM controller switching direct off the battery, then driving the big transformer. Later at about 11:50 he talks about the FETs and 20 kHz switching frequency. He doesn't show any filtering or output capacitors in this quick sketch but I assume that the transformer and output caps handle the filtering needed to separate PWM generated 60 Hz from the 20 kHz PWM switching.
OK I guess we are at the point in the discussion where we insult each other.
I know you have expertise in this area and I've asked you a couple times to comment on the topology differences, not just throw out vague references to quality. I'd still like to hear your views on that. I don't care if you call it LF or HF or tube based.
I did look at the paper that @Dzl referenced earlier, and it does speak about transformer-based LF and also about HF. Differences they show are where switching happens. LF switches directly off the battery and drives a transformer. And among LF configurations, those using H bridge are stated as suitable for sine wave generation. Victron uses the topology that they call LF with H-bridge switching driving a transformer.
If you'd care to speak to that, I'd be interested to hear what you have to say about Victron's use of this topology and how that relates to the heavy magnetics it uses, vs. what is used in the light cheap "HF" inverters.
After you said that I looked again at Victron and learned something so I'll eat part of a crow on this one. They claim to do both together.
They actually call it Hybrid HF which they say means " Combined high frequency and line frequency technologies ensure the best of both worlds. " What does that mean? That they use what is traditionally a LF transformer-based topology, switching directly from the battery, but combine that with PWM controlling the bridge?
Again I couldn't care less about the specific names other than to have a shared terminology. I'd just like to understand the topology differences between the heavy iron units like the Victron Multiplus, and how that relates to long duration surge capability. I do think Victron engineering and quality is part of the equation, but I think there are also topology differences between the Victron/Magnum/Samlex heavy iron and the cheap light "HF" inverters.
However, if they actually do all have the exact same topology, and the cheap light inverters just don't use devices or heat sinks big enough to allow surge beyond milliseconds, I'd be surprised but also quite interested to understand that too.
Could either of you cite where you are getting this info from (regarding Victron). I am having a really hard time finding specific information from Victron about their inverters and would like to learn more about this.
And I completely agree on both points as previously stated in post #45.
gnubie
Solar Wizard
- Joined
- Sep 20, 2019
- Messages
- 3,847
I'm having to wonder at this point if you have properly read this thread where the basics of 'HF' and 'LF' inverters has already been explained.
I already posted one good Victron video earlier, which talks about the design. Another good Victron webinar video talks about overload and surge specs. Starting about 9:45 he talks specifically about surge duration and explains that the 2X max power number in the datasheet is rated for 2 minutes (for example the 6kW max power for a Victron 3000VA which is rated for 2400W continuous)
Thanks. Wish Victron would put some of this stuff in writing, much easier to reference a datasheet or even manual than a half hour long video. I have been meaning to watch the video though, so I'll check it out when I get a chance.
Thank you sir may I have another. Yes I've read several opinions. Apparently I missed the clear summary of the HF/LF basics in this thread, my apologies.
You said HF and LF both use PWM at 10+ kHz, and the difference is that what you call HF does PWM directly on the output terminals. But @HaldorEE says all PWM is HF, which seems to conflict with that. @BiduleOhm says HF = small light ferrite working at HF (a bit recursive) and LF uses big heavy transformer working at line frequency.
I asked about the Victron topology, which does PWM switching on input not output, has a big heavy iron transformer, but does not switch at line frequency. So it is not an "HF" or "HF Direct" topology by your definition. @HaldorEE says it is HF because it uses PWM. It does not fit into either of @BiduleOhm's two boxes.
Lacking any other references other than conflicting opinions, I fall back to the the reference @Dzl usefully provided, following are some excerpts. Based on this, I would call the Victron unit an LF transformer-based topology using an H-bridge and PWM for pure sine generation.
"Low Frequency Transformer Based Inverters
"The following topologies are based on low frequency switching of the low voltage DC side, applying
the resulting DC pulses to a step-up transformer. Two common topologies are the push-pull, and
the H-Bridge. The push-pull topology is suitable for production of square and modified square output
waveforms, while the H-Bridge is useful for producing modified square wave and sine wave outputs."
... (later in the part on H-Bridge topology)
"After a period of time (variable according to pulse width modulation for voltage regulation) the
switches that were closed open, and the bottom two transistor switches close providing off-time
shorting (Figure 9B). The length of the on and off-time is determined according to the PWM
controller."
"High Frequency Inverters
"High frequency inverters are another approach to creating higher power AC from low voltage DC.
The name, high frequency, refers to the speed at which the transistors switch on and off. This type of
inverter creates low voltage AC from battery power, and applies it to a high frequency transformer,
which creates high voltage AC. The high voltage AC is then rectified (changed back to DC) to high
voltage DC and then a low frequency switcher (an H-Bridge) creates utility power AC."
...
"The main advantage behind high frequency switchers is the very lightweight and physical size. Most
HF inverters are low cost for the smaller sized units (less than about 300W).
Disadvantages of HF inverters are poor surge ability due to the characteristics of the switching power
section supplying the bridge (limits their usage to motor loads). Lack of isolation between the
transistors and AC loads makes them very vulnerable to transients caused by reactive loads since
there is no transformer to isolate and act as a “flywheel” to oppose fast changes in output current. HF
inverters exhibit high idle current because the high voltage switcher runs constantly, and this also
often causes interference with TV’s, radios, etc."
60 Hz transformers have some high frequency filtering capability. Not enough to fix what is wrong with modified sine wave inverters, but some.
Pretty much all modern inverters are high frequency as in they use high frequency PWM modulation to turn DC into AC.
The cheapskates do it by stepping the battery voltage up to 200 VDC and directly converting that to 120 VAC. They then add some inductors and capacitors to filter out the high frequency noise. That could result in an acceptable inverter (basically that is how a class D audio amp works).
The problem is they cut too many corners until the inverter is working as hard as it can to marginally meet specs and there is nothing left over for surge capability.
Higher end inverters convert battery voltage to some low voltage AC. In Victron it is 8 VAC, then use a transformer to convert this to 120 VAC. The transformer is safer than direct DC to AC conversion because it ensures that high voltage can't end up on your batteries even if things go wrong inside the inverter.
Because these high end manufactures aren't trying to squeeze out every single penny possible from the designs, their products are not working as hard as they possibly can to meet specs which means they have reserve left to deal with surge requirements. It really is as simple as that. There is no magical low frequency fairy that sprinkles pixy dust over the inverter and makes it work better.
Power transformers are expensive so when you see one in an inverter you know their focus is probably more on meeting specs instead of making the cheapest possible product. The transformer in my Victron costs over $300 (I looked it up). There is a reason why good 2000W inverters cost more than $600.
Pretty much all modern inverters are high frequency as in they use high frequency PWM modulation to turn DC into AC.
The cheapskates do it by stepping the battery voltage up to 200 VDC and directly converting that to 120 VAC. They then add some inductors and capacitors to filter out the high frequency noise. That could result in an acceptable inverter (basically that is how a class D audio amp works).
The problem is they cut too many corners until the inverter is working as hard as it can to marginally meet specs and there is nothing left over for surge capability.
Higher end inverters convert battery voltage to some low voltage AC. In Victron it is 8 VAC, then use a transformer to convert this to 120 VAC. The transformer is safer than direct DC to AC conversion because it ensures that high voltage can't end up on your batteries even if things go wrong inside the inverter.
Because these high end manufactures aren't trying to squeeze out every single penny possible from the designs, their products are not working as hard as they possibly can to meet specs which means they have reserve left to deal with surge requirements. It really is as simple as that. There is no magical low frequency fairy that sprinkles pixy dust over the inverter and makes it work better.
Power transformers are expensive so when you see one in an inverter you know their focus is probably more on meeting specs instead of making the cheapest possible product. The transformer in my Victron costs over $300 (I looked it up). There is a reason why good 2000W inverters cost more than $600.
Last edited:
The explanation you are quoting from is not accurate.
Low frequency when talking about AC power inverters are circuits that are switching at some low multiple of 60 Hz (180 Hz is about as high as I have seen so far). This is how a modified sine wave inverter works. Really old pure sine wave inverters used square wave or modified sine wave inverters with the output passing through a ferroresonant transformer that both cleans the noise up and adds magnetic surge capability. Normal power transformers don't act as magnetic flywheels to store energy. That is a property of ferroresonant transformers. However ferroresonant transformers are even more expensive than normal ac power transformers and create problems of their own that modern inverters don't have. That is why modern inverters don't use ferroresonant transformers.
10 KHz is high frequency! Is it 60 Hz? No. it is 167 times higher frequency than 60 Hz. That is what I mean by high frequency.
Could an inverter use even higher switching frequency (like in the MHz) and get away with using smaller, cheaper magnetics. Yes, this is commonly done in low power DC-DC converters like wall wart power supplies. The problem with doing this in a high power AC inverter is that it could cause massive radio interference. As in the FCC will hunt you down and putting you in jail kind of interference.
PWM at 10KHz is generally considered on the high end for AC power circuits. Most motor drives (VFDs) use switching frequency between 4000 to 10000 Hz and inverter technology borrows heavily from VFD drives (basically an inverter is a simplified VFD that operates at a constant frequency and voltage level).
Last edited:
One more thing. Talking about motor drives as though doing this is easy is misinformed. I don't know of anything more electrically demanding than starting a motor with a locked rotor. The startup surge requirement can easily be 10 times normal runtime current. And it won't go away until the motor starts spinning or the circuit breaker trips.
gnubie
Solar Wizard
- Joined
- Sep 20, 2019
- Messages
- 3,847
Part of the HF vs LF (and to be clear here these are both the PWM driven types) problem is how they regulate the output, explanation by gifting the cheapies with that ability which they don't always have. HF inverters generally, but not always, maintain a fixed PWM ratio (poorly worded) to construct the sine wave and use soft regulation on the HVDC side to try to keep it at the right voltage vs load. LF inverters generally, but not always, have a feedback path from the AC side into the PWM board which adjusts the PWM duty cycle and regulates the AC output that way.
Thanks for that explanation and continuing this dialog, it is useful for me.
OK let's set PWM aside as I believe that all modern inverters use PWM whether they have large transformers and switching at input or small and switching at output. Your position is that anything PWM is HF. Fine, let's assume that all the suppliers that produce what they call "LF" inverters (and many also produce what they call "HF" inverters) just use LF as a marketing term to describe a category that has a large, heavy iron, low frequency transformer at the core of the design.
Names aside, I see mainly two classes of designs for power inverters on the market, and I believe that most or all use PWM controllers:
1) "LF" designs that are heavy, roughly cubic in form factor, contain large heavy low frequency transformers, and have better surge.
2) "HF" designs that are lighter, long narrow form factor, smaller/higher frequency magnetics, and poor surge
These are two different topologies. The first historically switched at line frequency, but modern designs use a PWM controller mated to a large low frequency transformer. I have seen these referred to as "high/low" inverters, Victron calls it "Hybrid HF", other "LF" inverter manufacturers speak vaguely about new techonologies in their "LF" inverters and I'll guess that is just a similar use of PWM with large low frequency iron.
The second type, if I understand correctly, typically switches at the output not input and does not use a large low frequency transformer.
Your position seems to be that there is no inherent advantage for either of these topologies with regards to surge performance. The only difference driving the observable performance difference between the two classes for products on the market is that the "LF" models are built by quality manufacturers who build in sufficient margin, and the "HF" models are built as light and cheap as possible, cutting corners and eliminating margin in favor of low cost.
I understand your position, but I'm not sure I buy it yet. It seems at odds with the vast majority of what I can find on this topic. Perhaps you are right and the entire industry is deluded in thinking that large transformer-based designs are better for surge performance. But I'd like to read some papers from authoritative sources about that, I hope you can point me to some references.
In the interim, theory aside, what I see actually on the market are large heavy expensive cubes that have good surge, and long narrow cheap lightweight inverters that have poor surge. Different topologies and different form factors. Some manufacturers carry both categories. Presumably they build quality into their "LF" line and skimp and cut corners for their "HF" line?
What do you call those two classes of products?
I didn't follow the part about "explanation by gifting the cheapies", was that an autocorrect error? Not sure what you meant.
On the regulation differences--is that somehow coupled to the cheaper "HF Direct" topology you describe? Or is it just another separate area of the design where corners may be cut? It seems like you could vary PWM to provide voltage regulation for switching at output just as well as switching at input. And since the PWM control likely comes from an inexpensive IC I would think the "HF" guys would use it as it should be cheap to do that?
OK I found an interesting and relevant document, a TI inverter reference design. Please see section 3 "Comparison of Low-Frequency Inverter vs. High-Frequency Inverter." They define LF and HF inverters in the way I had understood, what they call a LF Inverter includes the Victron topology.
https://www.ti.com/lit/an/slaa602a/slaa602a.pdf
Some excerpts:
"In the LF inverter,the battery voltage is first chopped with the full bridge (using high-frequency PWM, generally 3 kHz to 20 kHz) to an AC waveform. The iron core transformer then boosts the 12-V chopped waveform to 220-VRMS output waveform at 50 Hz. At the output of the transformer, a capacitor helps filters the waveform to make a clean 50-Hz AC Sine Wave. Although this inversion method is widespread today, the iron core transformer is quite bulky and increases the cost of the overall solution."
...
"The next advancement in inversion technology is the use of high frequency inverters, or HF inverters. This technology involves more processing complexity but can significantly increase overall system efficiency and eliminate the use of the bulky iron core transformer. In a high-frequency inverter, the battery voltage is converted to an intermediate high DC voltage before it’s converted to an AC waveform using Pulse Width modulation."
https://www.ti.com/lit/an/slaa602a/slaa602a.pdf
Some excerpts:
"In the LF inverter,the battery voltage is first chopped with the full bridge (using high-frequency PWM, generally 3 kHz to 20 kHz) to an AC waveform. The iron core transformer then boosts the 12-V chopped waveform to 220-VRMS output waveform at 50 Hz. At the output of the transformer, a capacitor helps filters the waveform to make a clean 50-Hz AC Sine Wave. Although this inversion method is widespread today, the iron core transformer is quite bulky and increases the cost of the overall solution."
...
"The next advancement in inversion technology is the use of high frequency inverters, or HF inverters. This technology involves more processing complexity but can significantly increase overall system efficiency and eliminate the use of the bulky iron core transformer. In a high-frequency inverter, the battery voltage is converted to an intermediate high DC voltage before it’s converted to an AC waveform using Pulse Width modulation."
