diy solar

diy solar

Victron Quattro + Autotransformer for split phase (which option is best?)

Agreed. And the fact that it's open source and modifiable with an active community developing stuff like NodeRed ... yeah. Winner.
With the Venus OS being open source and pre-compiled to run on an RPI or a Beaglebone black, it can be a very cost-effective solution to buying a Venus GX or CCGX. An interesting solution for the DIY/Tinker/Programmer community.

The only other significant thing I noticed was that they seem to be concentrating on the European 50hz market. Most of their products that catch the eye are all 230v 50hz.. pretty useless in the USA.

Look closely at the bottom of the datasheets for pretty much any series of Victron inverter. You'll see they are 60hz switchable. What not too many people know is that this is true of a lot of medium to high end 50hz inverter manufacturers. And you'll notice for the most part the 230V/50hz models are much cheaper than their 120V or 120/240V counterparts stateside.

Also buying toroid AT's is not that expensive either IMHO. I, for example, have them custom made at a factory in China rather than buying Victron AT's. Then it's sized to the client's installation and lower cost as well.

What's the downside of high frequency architectures? I've seen claims of not being able to handle power surges but that's clearly not true of the Victron's...

So I have not tested them but I have heard the smaller Multiplus inverters do appear to have a little trouble starting larger loads as compared to a comparable inverter with a different bridge design and larger transformer it seems. But again I have not tested them myself.

The larger MP and Quattro inverters use hybrid design with accompanying isolation transfomer as far as I know and I have not heard of any issues with startup loads.

Check out this video from Victron:

Inside the Victron MultiPlus & a Detailed Explanation of How It Works - Victron Energy

 
Last edited:
What's the downside of high frequency architectures? I've seen claims of not being able to handle power surges but that's clearly not true of the Victron's...

It is possible to build a quality high frequency inverter.. seems like Outback and Victron are the only two that seem to do it. I'm not sure why this is however, they are all built using the same basic foundation. Maybe its component selection? Over building key parts? I really don't know and am only guessing.

But ya, they don't handle power surges very well at all. This isn't an issue if you're running a fridge or a sump pump, but its a big problem if you go to start a well water pump, air conditioner, or an air compressor. Big HF Inverter + Small well water pump = Fine.. but if that well water pump pulls significant juice, that's an issue for HF inverters when they're already running near capacity.

EDITED: Meant to say Outback and Victron
 
Last edited:
Do you mean a 120V 48V MP?

For US market yes that would be ideal.

Are you suggesting their 48v 230v 50Hz can work at 240v 60 Hz with the added cost of the auto transformer or some other add on ?

A bit pricey but would that be an option??

I run a bunch of Victron gear with my Schneider CSW 4048 which was more cost effective even after adding the ComBox.
 
Last edited:
For US market yes that would be ideal.

Are you suggesting their 48v 230v 50Hz can work at 240v 60 Hz with the added cost of the auto transformer or some other add on ?

A bit pricey but would that be an option??

I run a bunch of Victron gear with my Schneider CSW 4048 which was more cost effective even after adding the ComBox.
The short answer is yes, you can use pretty much any (I have not found a model that is not capable of this) 230V 50Hz Victron inverter (Phoenix) or inverter/charger in 120/240V split phase environments using any appropriately installed 120/240 auto-transformer including units such as the PSX-240 from Outback or the 5K from Solaredge. The one from Solaredge is the best value in my opinion unless a person has them custom ordered like I do. That is generally cheaper yet and they are more efficient toroid transformers.

I believe the retail pricing of the Multiplus II 48/5K/230 is in the ~$1600 range. With that and the cost of an AT, it'll likely be more expensive than the CSW by a bit. As far as how they compare, I would say they each have thier pros and cons. For me, the deciding factor is BMS compatibility and AC-Coupling/PV Inverter support. I prefer not to use DC-coupling unless I need to or the best solution is typically AC+DC coupling for medium to large systems. I'm mainly referring to off-grid but other applications can be served well by an AC+DC Coupled installation.

Again I would warn against using the Victron for anything but off-grid due to certification issues. You can probably charge with it from the grid but I would check the local code on that before doing anything.

One final note, I would recommend if possible always make sure a person has a qualified sparky help with the wiring, particularly with installing AT's. :cool: It can get tricky and electrocution can definitely happen if it's wired incorrectly.
 
The short answer is yes, you can use pretty much any (I have not found a model that is not capable of this) 230V 50Hz Victron inverter (Phoenix) or inverter/charger in 120/240V split phase environments using any appropriately installed 120/240 auto-transformer including units such as the PSX-240 from Outback or the 5K from Solaredge. The one from Solaredge is the best value in my opinion unless a person has them custom ordered like I do. That is generally cheaper yet and they are more efficient toroid transformers.
If a person is highly adventurous, a person might want to investigate the Fangpusun Xtender XTM and XTH inverters. Thier tech support so far has been great and although I have not had the opportunity to install one yet, I am comfortable with thier efforts to perfectly "clone" Studer that they appear IMHO to be one of the best bangs for your buck right now for a high-end inverter/charger.

I should make a post about these perhaps. In the meantime, here is a pricesheet and a link to the specs along with some photos. Look at the size of that inductor. These are comparable I would say to Victron and SMA. Studer OEM inverters are extremely good and many consider them superior. The Fangpusun models are very, very good, part-for-part clones down to/including the firmware from what I can tell so far. They are so accurately identical that they work seamlessly with other Studer equipment from what I understand.

One of the major selling points for me is actually the firmware compatibility with Studer. That means I can update to new features as Studer rolls them out. If a person reads the Studer manual, you will find they are in many ways much more advanced than almost any inverter on the market, including SMA. And with the XCOM-CAN for BMS integration and fantastic AC-Coupling support, it appears to be a super deal all the way around I think.

They also let me know there will be 120V models before too long so a person could use 2 for split-phase instead of using an AT if that is preferable.

Again these could end out being a money pit but so far from extensive communications with them they seem to be on the ball. Only a hands-on test will really tell.

I'm wanting to get one of their Victron cloned controllers also to play with. So cheap. I heard that the CEO of Studer told the CEO of Victron that they should be flattered as Fangpusun only clones the very best. That could just be hearsay but fun nonetheless.
 

Attachments

  • XTM_XTH-price_sheet.jpg
    XTM_XTH-price_sheet.jpg
    53 KB · Views: 35
  • VictronMPPT.jpg
    VictronMPPT.jpg
    76 KB · Views: 36
  • MPPT-100_50.png
    MPPT-100_50.png
    377.4 KB · Views: 36
  • Fangpusun_XTM_interior.jpg
    Fangpusun_XTM_interior.jpg
    46.4 KB · Views: 36
  • XTM_Specs.jpg
    XTM_Specs.jpg
    185.7 KB · Views: 35
  • XTH1.png
    XTH1.png
    422.8 KB · Views: 30
  • XTH2.jpg
    XTH2.jpg
    55.7 KB · Views: 33
Last edited:
Thanks for the education.

I’m happy with the value Schneider provides with great hardware. The CSW for instance has an AT built in so although I feed it 240v from the grid and output 240v to a typical US critical loads panel using separate 120v legs I could feed it 120v from a GenSet and still have it output 240v to my AC critical loads panel. That would be only if my solar and grid power went down for days in a power outage without sun.

My application is to make, store and use power for the critical loads on a daily basis. The only way the system will pay for itself quickly is in an extended power outage where it keeps the basement from flooding and saves two freezers of food.

Where Schneider falls down is their klunky software and horrible support. The Victron software is excellent and I have never needed to contact support.

“If I knew then what I know now” I would probably go with a 48v MP and AT setup to go with my 3 Victron MPPT’s, SmartShunt and CerboGX.

Maybe in the future I might sell off the CSW4048 with ComBox and make the Victron change to simplify my monitoring/management life!

Fangpusun seems too good to be true but I’m open to being proven wrong ?
 
Thanks for the education.

I’m happy with the value Schneider provides with great hardware. The CSW for instance has an AT built in so although I feed it 240v from the grid and output 240v to a typical US critical loads panel using separate 120v legs I could feed it 120v from a GenSet and still have it output 240v to my AC critical loads panel. That would be only if my solar and grid power went down for days in a power outage without sun.

My application is to make, store and use power for the critical loads on a daily basis. The only way the system will pay for itself quickly is in an extended power outage where it keeps the basement from flooding and saves two freezers of food.

Where Schneider falls down is their klunky software and horrible support. The Victron software is excellent and I have never needed to contact support.

“If I knew then what I know now” I would probably go with a 48v MP and AT setup to go with my 3 Victron MPPT’s, SmartShunt and CerboGX.

Maybe in the future I might sell off the CSW4048 with ComBox and make the Victron change to simplify my monitoring/management life!

Fangpusun seems too good to be true but I’m open to being proven wrong ?

Have you ever considered Time-Of-Use load shifting? In my installation area, the utilities only offer flat rates so TOU is useless but if you have TOU in your area, you could probably reduce your bill.

I'm hoping to test a Fangpusun this coming year. They recently raised their prices though (the sheet I sent is up-to-date as of a couple of weeks ago). With the new pricing, they aren't really any cheaper than Victron inverters (for installers, we get different pricing) but an interesting option for sure. It does some things that Victron currently doesn't have the capability of. For example, a use case is adding more storage capacity to an existing AC-Coupled off-grid installation. If the system is aging and the existing batteries only hold say 80%, adding more storage might be needed but retiring the existing bank would simply be wasteful. Directly adding additional batteries to the existing bank would simply be .... unwise to say the least. At that point there aren't many options however with the software parameters provided on a Studer/Fangpusun, I can install one of their inverters as a current-source-inverter (rather than a voltage-source-inverter) with frequency-shift power control (pf) into any existing AC-Coupled system and use the new inverter to operate a new battery bank that will charge/discharge in support of the existing system. Essentially adding additional capacity without touching the original system.

I haven't tried this yet (not a use case ATM) but the engineer in me is curious to deploy a system upgrade like this and thoroughly test it. The design works but implementation in the field always brings new and interesting challenges.
 
Have you ever considered Time-Of-Use load shifting? In my installation area, the utilities only offer flat rates so TOU is useless but if you have TOU in your area, you could probably reduce your bill.

Have you analyzed the cost of implementing load-shifting using batteries?
For a proposed battery, considering its purchase price and cycle life, cost of inverter and balance of system components, and the number of cycles per day it will experience for utility TOU schedule, what is the cost per kWh (ignoring time value of money)? The lifetime to amortize over or consider rate-of-return?

My math, utility rates, and battery costs say it load shifting (with battery storage) doesn't benefit me.
I think it is more cost effective to deliver the utility excess kWh while the sun shines in order to pay for consumption during peak times.
Off-grid (during blackouts), it is more cost effective to curtail surplus PV production than to store it in batteries.
In other words, only load-shift by controlling loads, not by storing power in batteries.
 
Have you analyzed the cost of implementing load-shifting using batteries?
For a proposed battery, considering its purchase price and cycle life, cost of inverter and balance of system components, and the number of cycles per day it will experience for utility TOU schedule, what is the cost per kWh (ignoring time value of money)? The lifetime to amortize over or consider rate-of-return?

My math, utility rates, and battery costs say it load shifting (with battery storage) doesn't benefit me.
I think it is more cost effective to deliver the utility excess kWh while the sun shines in order to pay for consumption during peak times.
Off-grid (during blackouts), it is more cost effective to curtail surplus PV production than to store it in batteries.
In other words, only load-shift by controlling loads, not by storing power in batteries.
Actually I have but I have not implemented one yet. You make some very good points. Every use case is different and IMHO I believe there are some use cases where it could make sense possibly.

There are places where TOU is implemented that grid feed is not available or is not paid for providing very little ROI.

We figured it up and using, for example, the Chinese LiFePo4 cells many on this forum have bought or talk about, combined with a single inverter, we took actual data from actual homes and it came out that the system would be paid for in 2.5 years or less in the money saved on the utility bill.

At 2.5 years, 1 cycle per day, comes to ~912-913 cycles. For a battery that is rated to essentially 4K cycles to 80% @ 1C, your barely at 95% (theoretically). From the lab and field tests I have seen and read, assuming properly operated (BMS CAN bus comms, never exceeding 3.5VPC etc), there is a very good chance at the low C rates we see in storage applications they will probably be higher than that.

(This sent before I was done, crazy phone, but I'll probably add more in another post).
 
You saw 2.5 year payback for a system with lifespan several times that? What utility rate per kWh and what battery cost?

My TOU rates are $0.15, $0.30, $0.45/kWh. Used to be (with other rates), peak was Noon to 6:00 PM so I could be a net producer. Now 4:00 to 9:00 PM, not much PV production available. My area of the country is expensive. Elsewhere, electricity rates are a fraction of that.

With $0.30 spread and battery costing $0.20/kWh, battery system costing $0.30, can't win.
Even if no net metering, better to curtail PV production for no export than to store power.

Some people have to pay "demand charge" based on peak draw. If infrequent, battery could help there. But I think better to just shut off loads to reduce peak draw (maybe fill in valley during different hour). Or over-panel more.

Some commercial rates may reach $0.60 or $1.00, so more cost effective.

Standby generators are another option. I looked at a small one, decided the labor for oil changes was worth more than the power produced. Trukinbear posted his costs to run a diesel generator. Cost per kWh was higher than my top-tier rates. So worthwhile for some purposes, but not to offset grid consumption.

I think many people will invest in battery systems thinking it will save them money, but it will cost them more in the long run.

The primary purpose of my batteries is to power the island grid so grid-tie PV inverters have somebody to play with. During blackouts, I can run A/C an everything else direct off PV. At night, battery can barely keep my several refrigerators and other loads running until dawn. I should set up load-shed of those at night when grid is down. Could have 1/4 the battery in that case. But my 20 kWh battery is reasonable to support the entire house during brief times (minutes) when GT inverters drop off line, and to suck down their output for a few seconds during load-dump, while frequency-shift occurs.

In my case none of that is cost-effective, just as luxury and convenience.
A small system could run phone and internet/laptops.
A generator run a few hours per day would chill the refrigerators and freezers.
 
Have you ever considered Time-Of-Use load shifting? In my installation area, the utilities only offer flat rates so TOU is useless but if you have TOU in your area, you could probably reduce your bill.

Flat rate here as well.
 
So I was trying to find the notes where we did the figuring but I can't find it. It was a back-of-the-napkin calculation anyway based on one of the FL power rate structures. I *believe* it was FPL but I'm not sure. In any case, I was looking around at rates in general now and it's definitely more along the lines of what you mentioned, making it less than feasible in many cases.

Up in GMP's coverage area (just to grab a random utility), their rates are:

Usage Peak $0.26114/kWh
Off-peak $0.11131/kWh
Peak hours Monday-Friday 1:00pm – 9:00pm

So at worst 1000kWh would be ~$260 if a person never used off-peak. Or if using load shifting, that would mean ~$111. The difference there is not enough to implement TOU in a reasonable timeframe that would pay itself off like I had thought unless the batteries were dirt cheap. Even then unlikely. I'm inclined to believe we made a mistake in the calculations.

I remember now the batteries in question we were looking at were second life and less than $100 per kWh before installation. Clearly my memory is not as good as I thought it was...

Generators are an interesting thought. Too many factors there to make a blanket statement but I would think in general it wouldn't be cost-effective for demand load shaving just as you mentioned.

As a random note in contrast with that, for off-grid systems, I usually recommend to budget-constrained clients to not fret over installing enough storage to cover extended rainy days or constantly checking the SOC to load shave. Just use the system like you would the grid and let genset pick up the load when you need to because we're going to install one anyway and LP in this area is really cheap. Besides that, it's good for the genset to run anyway.

For limited budgets, even something like a Westinghouse WGen9500DF should work or I think it will. We're going to be testing it soon on a new install. Been looking for a smaller/cheaper genset that we can actually configure with autostart (auto-choke) without spending a mini fortune on a full-size standby. Sure I could put on an actuator for the choke but that is just something else to fail IMHO.

@Hedges out of curiosity, do you have a master load-shedding contactor to prevent black-start? You may not have long enough blackouts to warrant it or have an autostart genset. If you do though, I'm curious where do you have your load-shedding SOC cut-off set? Do you only use around 10-12kWh or do you take it lower than that? Also are you using the RS485 piggyback cards for switching UL1741 to Island-Mode when the grid drops? No rush on a reply, just curious.
 
@Hedges out of curiosity, do you have a master load-shedding contactor to prevent black-start? You may not have long enough blackouts to warrant it or have an autostart genset. If you do though, I'm curious where do you have your load-shedding SOC cut-off set? Do you only use around 10-12kWh or do you take it lower than that? Also are you using the RS485 piggyback cards for switching UL1741 to Island-Mode when the grid drops? No rush on a reply, just curious.
Yes, because my AC coupled Sunny Boys are on output of Sunny Island, they can only operate and provide PV for battery charging if Sunny Island is delivering AC. So they are on a "critical" breaker panel, which has 100A breaker feeding the load-shed relay. I bought the (ABB?) relay which SMA resells. SI data sheet said several amperages available, but I only found 100A.

It is 3pst. 48V coil measures 10 ohms, so it draws 5A (250W!) through the 0.5A PTC fuse in SI, but slams shut in milliseconds and opens an extra relay contact switching the coil to much higher resistance (two coils in series, with relay shorting one.)

I left the default load-shed setting of 70% DoD. SI continues to run after that to at least 80% DoD battery-protect. My first test with grid breaker shut off, house was load-shed at 3:00 AM after consuming 14 kWh. I have a few old refrigerators, a tube amp, several yard lights including LED rope and sodium. With lights and amp turned off, draw appears low enough.

What I need to do is set up a signal for either "on grid" or "on grid" || "< 20% DoD" for those excess loads, possibly also refrigerators. Also electric heat - I select between gas and electric depending on time of use, burning off surplus electric credits before they're reset each year.

Heard from the neighbor there was a brief power failure yesterday. I hadn't even noticed. Electric heat (either 5 kW or 10 kW) was operating. If left on the battery would hit 70% DoD load-shed, reminding me to turn off the heater. Then I'd have to wait for charging back to 50% DoD before power returned.

Was supposed to use RS-485 for UL1741/Island. My SWR2500U aren't supported, so I bough 10000TLUS. Turns out that in "backup" those only respond to frequency shift by remaining online to about 64.5 Hz but don't adjust output. SMA reviewed their code, said I was correct about that, and to just leave them in "island", with Sunny Island taking care of UL1741. I've now bought a pallet of 5000US which I'm going to install instead.

I have a second load-shed relay which I will control around 20% DoD for the garage. I will use other signals (rather than that power relay) for other loads due to where they are located. In particular, A/C thermostat so it runs direct off PV but doesn't run battery down much. Also one leg to the electric dryer so motor and timer keep running but heating element shuts off.

I could add DC coupled PV. I have two SIC-40, the Sunny Island Charger, with SIC-PB so they talk to SI. That could be an alternative to load-shed relay. But putting 20% charge back in would take longer. DC coupled PV should be limited to 0.2C, so as to not charge batteries faster than recommended. Early in the morning it would be slower than 0.2C unless over-paneled (which would be wasted capacity middle of the day of charge controller clipped). But the AC coupled PV I have installed is more like 0.5C, so batteries will reach 0.2C earlier and reconnect loads sooner. Sunny Island would regulate PV output during that period, so battery charging would sit constant at 0.2C.
 
@Hedges Cool, that is what I was wondering. Very nice setup. I'd be curious to know how you approach configuring your signaling when you implement that if you don't mind. Are you thinking about using DigIn for that? I don't think it would work for that BUT I may not understand exactly what you have in mind.

We recently had a system where during the rainy season it would shed and per normal settings would not reconnect the loads immediately once the generator was running (manual start, sometimes the gen wouldn't get started before load shed occurs). For the client's convenience, I used another relay in the SI in parallel with the AutoLod1Soc relay and configured it for GnRn, reconnecting the loads when the gen connects.

It is 3pst. 48V coil measures 10 ohms, so it draws 5A (250W!) through the 0.5A PTC fuse in SI, but slams shut in milliseconds and opens an extra relay contact switching the coil to much higher resistance (two coils in series, with relay shorting one.)

We usually use ABB AF65 or AF95 contactors. For single SI+AT installs, I'll put in an AF52. But yeah it's a crazy surge on the relay. I have been contemplating switching to either 120V or 240V drive for the coils. In fact, I have some I ordered for that purpose. If the SI is off, it's going to drop loads regardless. If you have any thoughts on reasons not to do this I would be glad to hear it. Right now I'm thinking I'd rather not have the high DC A through the relay on a system that has a manual start gen and is likely to trip the load shed a few times or more per year. In the long run it's probably not an issue but another driving factor is contactors with higher voltage coils are generally cheaper to source.

SIC-40. If I may say, I have never met anyone still contemplating installing those these days although it could make sense for your application. I can still source them from a supplier in the UK but they are hard to come by. I was tempted to get one or two for a specific install awhile back but SMA claims they are incompatible with LI batteries. I'm assuming the SI doesn't pass the real-time instructions on from the BMS.

That said, Victron actually has it listed in their documentation DC-coupling backup (small MPPT) as a requirement for AC-Coupling with PV inverters. However it has since been clarified by a rep at the company that it's not required per se, just highly recommended. At this point, it's cheaper to install a load shedding contactor than to install a separate MPPT so I don't usually spec it.

Do you have a backup genset at all then? I got the impression from your previous post that you're not using one ATM? Just curious.
 
DigIn would be signaling to SI. Available functions include indicating whether grid vs. generator is feeding it.

I think one SI relay function indicates whether connected to AC input (could be grid or generator.) Another of course is one of two programmable load-shed levels (wish more than 2 could be set, since I have 8 relays in four SI). These should be sufficient for various load control and shedding.

I'd rather enable heavy loads based on amount of excess PV which would otherwise get shed, rather than based on DoD. Since reconnect is 20% higher than shed, not sure anything higher than 20% DoD can be used for shedding. Somebody did build a switch based on frequency. Still, you can't be sure about available watts based on frequency. 61 Hz has Sunny Boy deliver 100% of available (at the moment), linearly down to 0% at 62 Hz. That's not 100% of Sunny Boy max rating, rather 100% of PV at the moment. So is 61.5 Hz 5000W available? or 100W available late in the evening?

5A DC for the relay coil is a lot, but no big deal. If SMA Germany designed it, it is robust. A PTC fuse has a maximum (peak) current spec, and tripping is based on temperature which of course is I^2 x R x time. AC relays I think reduce their hold current by nature of the magnetic circuit when pole piece is pulled in, so don't need the switching that power DC relays implement. I considered using them but prefer not to route AC through that portion of the SI. If you do, would be good to fuse it with small, fast fuse. Maybe current-limiting resistor or light bulb?

You could be right that configuration of voltage levels for SIC is at boot-up, wake-up or otherwise infrequent rather than real-time. Charge phase would of course be real time. I picked these up on eBay a few years ago, then had a hard time getting SIC-PB. Some sites listed it but showed the wrong picture. I bought direct from MSTE Solar, with wired payment.

Midnight Classic with data interface provides the same function. Trukinbear had a problem due to which side of shunt he connected Midnight, same as SI, I think. You'd think current from SI to battery and current reported by SIC to battery would be accounted for the same, but apparently not and Midnight has to bypass shunt. Not sure if SIC is same as Midnight in that regard or different.

For now my only use of SIC is on a portable testbed, later to be used for a truck-mounted system. The Sunpower 327W I was going to use are too low Vmp as a single panel, too high Voc if two in series in extreme cold. So I'm considering buying Sunpower 427W. They are a good fit voltage wise and will span the width of the truck (stick out about as far as mirrors.)

What you want is for SI to tell SIC or Midnight to reduce current output to match desired battery charge current (0.2C in my case), or deliver more when the extra is being used by SI to generate AC. Charge controllers ought to monitor a battery shunt for that purpose, but I don't know if any do. I think someone said the Victron ecosystem has inverter telling DC coupled charge controllers how much current. I would limit DC coupled PV to 0.2C, use AC coupled for the rest.

No generator presently. With PV production on a good day being several times usable battery capacity, should rarely be an issue. Especially with gas heat available, the idea is to manage loads. I might set up a 12VDC system for Uverse/Ethernet/VOIP that would have multiple days battery capacity, PV, and (island) grid charging.
 
But I'm not crazy about their use of the high frequency architecture in their larger inverter systems. The only saving point is that we don't hear about failures so they must be doing something right.

Wanted to address this point. A high frequency inverter is one that uses pulse width modulated, high voltage DC to directly generate AC line voltage. This concept is the lowest cost solution to get pure sine wave AC since the DC to DC converter uses high frequency switching circuitry to boost battery DC to high voltage DC. Using high frequency switching boost permits the use of small, lightweight and inexpensive transformers.

Victron uses pulse width modulation to convert battery voltage into approximately 8VAC at very high current. This is then stepped up to AC mains voltage using a conventional low frequency power transformer. The power transformer is an excellent noise filter resulting in very low noise, high quality AC power. This also makes it very easy to use essentially the same inverter in every market in the world by just using the right transformer. This design concept is used by virtually all of the high quality inverter manufactures in the world.

Modified sine wave inverters use high frequency DC boost combined with a much older technology to generate a crude approximation of a sign wave. It is hard to understand why anybody would make a modified sine wave inverter any longer. They aren't significantly cheaper than pure sine wave high frequency inverters any more.

A long time ago people added a ferroresonate transformer to a modified sine wave inverter to clean up the output noise. This works for certain definitions of works. One of the other benefits of a ferroresonate transformer is that it provides a regulating capability which greatly enhances the surge capability of the inverter. Thus was born the mythical surge prowess of "low frequency" inverters. It isn't the low frequency nature of these inverters that gave them increased surge capability, it was the ferroresonate transformer. You could add a ferroresonate transformer to ANY inverter and improve surge capability in exactly the same way.

Time passed, technology improved. Turns out that ferroresonate transformers have their problems too. They are very expensive per watt, are heavy and can burn up if you don't load them down sufficiently. Basically nobody uses ferroresonate transformers anymore.

Surge capability in modern inverters is purely a function of the quality of the design and manufacture. If you are designing a "3000W" inverter that will be sold for $400, then you are going to cut corners that means that "3000W" inverter is something that any responsible electrical engineer would probably rate at 1500W or less.

Victron designs their inverters to deliver rated power in real world applications and their surge capability is measured in minutes instead of milliseconds. Their use of low frequency power transformers is their choice of how to design products for different markets by simply selecting a different power transformer. Makes sense to me. If they chose to use the "high frequency" design paradigm I mentioned at the beginning of this post, I am sure it would work equally well. This is because Victron is a quality company that doesn't sell crap.
 
Last edited:
To continue. I don't believe there is any difference in how a Multiplus or a Quattro is designed. The differences between them are feature sets and power rating of the components.

Who knew, an $1800 inverter can deliver more surge capability than a $1200 Inverter from the same company. Imagine that.
 
To continue. I don't believe there is any difference in how a Multiplus or a Quattro is designed. The differences between them are feature sets and power rating of the components.

Who knew, an $1800 inverter can deliver more surge capability than a $1200 Inverter from the same company. Imagine that.

Another difference is that for some reason they don’t make a 48v Multiplus. ?
 
Wanted to address this point. A high frequency inverter is one that uses pulse width modulated, high voltage DC to directly generate AC line voltage. This concept is the lowest cost solution to get pure sine wave AC since the DC to DC converter uses high frequency switching circuitry to boost battery DC to high voltage DC. Using high frequency switching boost permits the use of small, lightweight and inexpensive transformers.

Victron uses pulse width modulation to convert battery voltage into approximately 8VAC at very high current. This is then stepped up to AC mains voltage using a conventional low frequency power transformer. The power transformer is an excellent noise filter resulting in very low noise, high quality AC power. This also makes it very easy to use essentially the same inverter in every market in the world by just using the right transformer. This design concept is used by virtually all of the high quality inverter manufactures in the world.

Modified sine wave inverters use high frequency DC boost combined with a much older technology to generate a crude approximation of a sign wave. It is hard to understand why anybody would make a modified sine wave inverter any longer. They aren't significantly cheaper than pure sine wave high frequency inverters any more.

A long time ago people added a ferroresonate transformer to a modified sine wave inverter to clean up the output noise. This works for certain definitions of works. One of the other benefits of a ferroresonate transformer is that it provides a regulating capability which greatly enhances the surge capability of the inverter. Thus was born the mythical surge prowess of "low frequency" inverters. It isn't the low frequency nature of these inverters that gave them increased surge capability, it was the ferroresonate transformer. You could add a ferroresonate transformer to ANY inverter and improve surge capability in exactly the same way.

Time passed, technology improved. Turns out that ferroresonate transformers have their problems too. They are very expensive per watt, are heavy and can burn up if you don't load them down sufficiently. Basically nobody uses ferroresonate transformers anymore.

Surge capability in modern inverters is purely a function of the quality of the design and manufacture. If you are designing a "3000W" inverter that will be sold for $400, then you are going to cut corners that means that "3000W" inverter is something that any responsible electrical engineer would probably rate at 1500W or less.

Victron designs their inverters to deliver rated power in real world applications and their surge capability is measured in minutes instead of milliseconds. Their use of low frequency power transformers is their choice of how to design products for different markets by simply selecting a different power transformer. Makes sense to me. If they chose to use the "high frequency" design paradigm I mentioned at the beginning of this post, I am sure it would work equally well. This is because Victron is a quality company that doesn't sell crap.

Do you have documentation showing the victron are LF inverters? Judging by the fact that their 6kW unit only weighs about 64 lbs, I find it hard to believe.
My Radian 8048 (8kW 120/240) is a high freq inverter and it weights 120 lbs.
My SMA Sunny Islands are 6kW Low Frequency (120 only) units and weigh 139 lbs each.

So when someone shows me a 5kw or 6kw inverter that's only 50 or 60 lbs, it screams High Freq...
 
Back
Top